WO2014047941A1 - 网络时延测量方法、装置和系统 - Google Patents

网络时延测量方法、装置和系统 Download PDF

Info

Publication number
WO2014047941A1
WO2014047941A1 PCT/CN2012/082490 CN2012082490W WO2014047941A1 WO 2014047941 A1 WO2014047941 A1 WO 2014047941A1 CN 2012082490 W CN2012082490 W CN 2012082490W WO 2014047941 A1 WO2014047941 A1 WO 2014047941A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
tlp
delay measurement
dcp
measurement
Prior art date
Application number
PCT/CN2012/082490
Other languages
English (en)
French (fr)
Inventor
方伟
刘宏明
畅文俊
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to CN201280001472.8A priority Critical patent/CN103109501B/zh
Priority to ES12885357.9T priority patent/ES2688696T3/es
Priority to PCT/CN2012/082490 priority patent/WO2014047941A1/zh
Priority to EP12885357.9A priority patent/EP2903212B1/en
Publication of WO2014047941A1 publication Critical patent/WO2014047941A1/zh
Priority to US14/669,513 priority patent/US10374925B2/en
Priority to US16/524,570 priority patent/US11133998B2/en

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0852—Delays
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
    • H04L43/106—Active monitoring, e.g. heartbeat, ping or trace-route using time related information in packets, e.g. by adding timestamps
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L7/00—Arrangements for synchronising receiver with transmitter
    • H04L7/0016—Arrangements for synchronising receiver with transmitter correction of synchronization errors
    • H04L7/0033—Correction by delay
    • H04L7/0037—Delay of clock signal

Definitions

  • the present invention relates to communications technologies, and in particular, to a network delay measurement method, apparatus, and system.
  • IP Internet Protocol
  • the measurement of the delay of the network service flow is mainly performed by inserting a special delay measurement packet at the measurement end, carrying the time stamp at the transmitting end and the receiving end in the delay measurement packet, and measuring the packet according to the delay.
  • the time-receiving time stamp is used to calculate the delay result of the network traffic flow.
  • Embodiments of the present invention provide a network delay measurement method, apparatus, and system to implement measurement of network traffic flow delay.
  • the embodiment of the present invention provides a network delay measurement method, where the method includes: acquiring time delay measurement information obtained by measuring at least one target logical port TLP, where the time delay measurement information includes: time stamp information , service flow identification and TLP identification;
  • the foregoing network delay measurement method may include: acquiring, by the at least one TLP, the time delay measurement information obtained by measuring the service flow, including:
  • the DCP obtains the sender delay measurement information obtained by measuring at least one upstream TLP for the sent service flow;
  • the DCP that manages the downstream TLP obtains the receiver delay measurement information obtained by measuring, by the at least one downstream TLP, the received service flow;
  • the sending the delay measurement information to the MCP includes:
  • the DCP that manages the upstream TLP sends the sender delay measurement information to the MCP, where the sender delay measurement information includes the sender timestamp information, the service flow identifier, and the TLP identifier;
  • the DCP managing the downstream TLP sends the receiving end delay measurement information to the MCP, where the receiving end delay measurement information includes the receiving end time stamp information, the service flow identifier, and the TLP identifier.
  • the foregoing network delay measurement method may include:
  • the DCP managing the upstream TLP acquires a measurement period identifier at the end of the measurement period, and sends the measurement period identifier to the MCP;
  • the DCP that manages the downstream TLP obtains the start time of the measurement period. If the difference between the start time and the timestamp information is less than or equal to a preset duration, the receiver delay measurement information belongs to the measurement period. Identifying the corresponding measurement information; if the difference between the start time and the time stamp information is greater than the preset duration, the measurement period identifier is incremented by one, and the time stamp information belongs to the next measurement period, and the measurement period identifier is Sent to the MCP.
  • the network delay measurement method where the preset duration is 2/3 duration of the measurement period.
  • the foregoing network delay measurement method may include:
  • the DCP managing the upstream TLP is synchronized with the upstream TLP by using an NTP or IEEE 1588v2 clock, and the DCP managing the downstream TLP uses the NTP or IEEE 1588v2 clock and the lower
  • the DCP also implements time synchronization by using the NTP or IEEE 1588v2 clock.
  • the foregoing network delay measurement method may include:
  • the DCP that manages the downstream TLP acquires the measurement packet sent by the upstream TLP received by the at least one downstream TLP, and the arrival timestamp information of a measurement packet generated when the measurement packet arrives at the downstream TLP,
  • the measurement message includes: a sender time stamp information
  • the network delay measurement method where the sending end delay measurement information further includes: a sending end service flow characteristic information and a sending end fragment recombination identifier, where the receiving end delay measurement information further includes: receiving End service flow feature information and receiver end fragment reassembly identifier to make the MCP Determining, according to the traffic information characteristic information of the sending end, the sending end fragment reassembly identifier, the receiving end service flow characteristic information, and the receiving end fragment reassembly identifier, that the sending end time stamp information and the receiving end time stamp information correspond to the same service. Time stamp information of the stream.
  • an embodiment of the present invention provides a network delay measurement method, which includes: identifying a service flow according to service flow feature information, and determining whether the service flow is a target service flow;
  • the delay measurement information includes: a time stamp information, a service flow identifier, and a TLP identifier, so that the DCP sends the delay measurement information to the MCP after acquiring the delay measurement information.
  • the network delay measurement method where a delay measurement flag is added to the data packet of the service flow, and the packet delay measurement information corresponding to the delay measurement flag is obtained, where: the upstream TLP is in the target service. Adding a delay measurement flag to the data packet of the stream, and acquiring packet delay measurement information corresponding to the delay measurement flag, where the delay information of the sender includes time stamp information, a service flow identifier, and a TLP identifier, After the DCP that manages the upstream TLP obtains the sender delay measurement information, sends the sender delay measurement information to the MCP;
  • the packet receiving end delay measurement information corresponding to the delay measurement flag is obtained.
  • the receiving end delay measurement information includes the time stamp information, the service flow identifier, and the TLP identifier, so that the DCP that manages the downstream TLP obtains the receiving end delay measurement information, and then sends the receiving end delay measurement information to the MCP.
  • the network delay measurement method where the method further includes:
  • the upstream TLP Before the upstream TLP adds the delay measurement flag to the data packet of the target service flow, use NTP to identify the data packet with the added delay measurement flag, and use the NTP or IEEE 1588v2 clock to manage the downstream TLP. Time synchronization of the DCP;
  • Adding a delay measurement flag to the data packet of the service flow, and acquiring data packet delay measurement information corresponding to the delay measurement flag and the following:
  • the upstream TLP adds a delay measurement flag to the data packet of the target service flow, and obtains a measurement period identifier corresponding to the delay measurement flag, so that the DCP managing the upstream TLP acquires the measurement. After the period identification, the measurement period identification information is sent to the MCP;
  • the network delay measurement method where the method further includes:
  • the receiving module of the downstream TLP receives the measurement packet, generates an arrival timestamp information of the measurement packet, and sends the measurement message and the arrival timestamp information to the DCP of the management downstream TLP. And determining, by the DCP, whether the time stamp information and the receiving time stamp information belong to a preset duration range, and if yes, determining that the sender time stamp information and the receiving end time stamp information belong to the same data. Packet, and send the determination result to the MCP.
  • the network delay measurement method where the sending end delay measurement information further includes: a sending end service flow characteristic information and a sending end fragment recombination identifier; the receiving end delay measurement information further includes: receiving End service flow feature information and receiver end fragment reassembly identifier;
  • the DCP that manages the upstream TLP acquires the sender delay measurement information and sends the measurement information to the MCP, and the DCP that manages the downstream TLP acquires the receiver delay measurement information and sends the information to the MCP, so that Determining, by the MCP, that the sender timestamp information and the receiving end timestamp information are corresponding according to the sending end service flow characteristic information, the sending end fragment reassembly identifier, the receiving end service flow characteristic information, and the receiving end fragment reassembly identifier. Time stamp information for the same packet.
  • the network delay measurement method where the upstream TLP adds a delay measurement flag to the data packet of the target service flow, including:
  • a delay measurement flag is added to the reserved bits of the TOS or the reserved bits of the Flags in the IP header of the packet.
  • the foregoing network delay measurement method where the identifying the service flow according to the service flow characteristic information includes:
  • the service flow is identified according to at least two meta information in the quintuple.
  • the embodiment of the present invention provides a network delay measurement method, which includes: receiving, by a DCP corresponding to an upstream TLP, delay measurement information of a transmitting end and a pair of downstream TLPs.
  • Receiver delay measurement information sent by the DCP, the sender delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier, and the receiver delay measurement information includes time stamp information, a service flow identifier, and a TLP.
  • the network delay measurement method where the method further includes:
  • the MCP receives the measurement period identifier sent by the DCP that manages the upstream TLP, and the MCP receives the measurement period identifier sent by the DCP that manages the downstream TLP, and the measurement period identifier sent by the MCP according to the DCP that manages the upstream TLP. And determining, by the DCP, the measurement period identifier sent by the DCP of the downstream TLP, determining whether the delay information of the sending end and the measurement information of the delay of the receiving end belong to the same measurement period, and if yes, the MCP is delayed according to the sending end.
  • the measurement information and the measurement information of the delay of the receiving end determine the single delay of the network.
  • the network delay measurement method where the method further includes:
  • the information belongs to the receiving end delay measurement information of the same data packet, and the MCP determines the network single delay condition according to the sending end delay measurement information and the receiving end delay measurement information.
  • the network delay measurement method where the method further includes:
  • the MCP receives the sender delay measurement information sent by the DCP corresponding to the upstream TLP, where the sender delay measurement information includes timestamp information, a service flow identifier, a TLP identifier, a sender service flow characteristic information, and a sender fragment reassembly.
  • the receiving end delay measurement information includes time stamp information, a service flow identifier, a TLP identifier, a receiving end service flow characteristic information, and a receiving end Tablet reorganization logo;
  • the MCP Determining, by the MCP, whether the timestamp information of the sending end and the timestamp information of the receiving end correspond to the information of the service flow of the sending end, the fragmentation reassembly identifier of the sending end, the characteristic information of the receiving end service stream, and the fragmentation reassembly identifier of the receiving end.
  • the time stamp information of the same data packet if yes, the MCP determines the network single delay condition according to the sender delay measurement information and the receiver delay measurement information.
  • an embodiment of the present invention provides a DCP, where
  • An obtaining module configured to obtain a delay measurement signal obtained by measuring at least one TLP to the service flow
  • the time delay measurement information includes: time stamp information, a service flow identifier, and a TLP identifier sending module, configured to send the delay measurement information to the measurement control point MCP, so that the MCP according to the time stamp information
  • the service flow identifier and the TLP identifier determine the network delay.
  • the DCP where the DCP is a DCP that manages an upstream TLP;
  • the acquiring module is specifically configured to obtain, by using at least one upstream TLP, measurement delay delay measurement information obtained by measuring the sent service flow;
  • the DCP is a DCP that manages a downstream TLP
  • the acquiring module is specifically configured to acquire, by the at least one downstream TLP, the received end delay measurement information obtained by measuring the received service flow;
  • the DCP sends the delay measurement information to the MCP, including:
  • the DCP is a DCP that manages an upstream TLP
  • the sending module is specifically configured to send the sending end delay measurement information to the MCP, where the sending end delay measurement information includes a sending end time stamp information, a service flow identifier, and a TLP identifier;
  • the DCP is a DCP that manages a downstream TLP
  • the sending module is specifically configured to send the receiving end delay measurement information to the MCP, where the receiving end delay measurement information includes the receiving end time stamp information, the service flow identifier, and the TLP identifier.
  • the foregoing DCP where the acquiring module includes:
  • the first acquiring unit is configured to obtain, by the at least one upstream TLP, the measurement information of the transmission delay measured by the service flow that is sent by the at least one upstream TLP, or obtain the measurement information of the received delay of the received service flow by the at least one downstream TLP. ;
  • a period identifier obtaining unit configured to: when the DCP of the management upstream TLP ends the measurement period, acquire a measurement period identifier, and send the measurement period identifier to the MCP, or the DCP in the management downstream TLP is At the beginning of the measurement period, the period identifier acquisition unit acquires a start time of the measurement period. If the difference between the start time and the time stamp information is less than or equal to a preset duration, the receiver delay measurement information belongs to The measurement period identifies the corresponding measurement information. If the difference between the start time and the time stamp information is greater than the preset duration, the measurement period identifier is incremented by one, and the time stamp information belongs to the next measurement period, and the management is acquired.
  • the DCP of the downstream TLP is identified in the measurement period of the measurement period;
  • the sending module includes:
  • the first sending unit sends the sending end delay measurement information to the MCP, or sends the receiving end delay measurement information to the MCP;
  • the second sending unit And sending, by the second sending unit, the DCP that manages the upstream TLP to the MCP by using the measurement period identifier acquired by the period identifier acquiring unit at the end of the measurement period, or acquiring the period identifier of the DCP of the management downstream TLP.
  • the measurement period identifier obtained by the unit is sent to the MCP.
  • the DCP where the preset duration is 2/3 of the measurement period.
  • the DCP where the method further includes:
  • a time synchronization module configured to perform time synchronization with the TLP by using an NTP or an IEEE 1588v2 clock, and use the NTP or the NTP or the IEEE 1588v2 clock before the obtaining module acquires the delay measurement information obtained by measuring the service flow by the at least one TLP
  • the DCP managing the upstream TLP is time synchronized with the DCP managing the downstream TLP by the IEEE 1588v2 clock.
  • the foregoing DCP where the acquiring module includes:
  • a second acquiring unit configured to acquire, by the at least one upstream TLP, the measurement information of the transmission delay measured by the service flow that is sent by the at least one upstream TLP, or obtain the measurement information of the received delay of the measurement of the sent service flow by the at least one downstream TLP.
  • a measurement packet obtaining unit configured to acquire, by the at least one downstream TLP, the measurement packet sent by the upstream TLP, and the arrival timestamp information of a measurement packet generated when the measurement packet arrives at the downstream TLP,
  • the measurement message includes: a sender time stamp information
  • the determining module is specifically configured to determine whether the arrival time stamp information and the receiving end time stamp information belong to a preset duration range, and if yes, determine the sender time stamp information and the receiving end time stamp information. Belong to the same data packet;
  • the sending module is specifically configured to send a result to the MCP.
  • the DCP where the DCP is a data collection point for managing an upstream TLP, and the acquiring module is configured to obtain the delay measurement information of the sending end, where the information about the delay information of the sending end further includes: The traffic flow feature information of the sender and the fragmentation reassembly identifier of the sender;
  • the sending module is specifically configured to send the sending end delay measurement information to the MCP;
  • the DCP is a data collection point for managing a downstream TLP
  • the acquiring module is specifically configured to acquire the delay measurement information of the receiving end, where the information about the delay information of the receiving end further includes: a service flow characteristic information of the receiving end and a fragment reassembly identifier of the receiving end;
  • the sending module is specifically configured to send the receiving end delay measurement information to the MCP, so that the MCP is configured according to the feature information of the sending end service flow, the fragment reassembly identifier of the sending end, the service flow characteristic information of the receiving end, and
  • the receiving end fragment reassembly identifier determines that the sending end time stamp information and the receiving end time stamp information are time stamp information corresponding to the same service flow.
  • an embodiment of the present invention provides a TLP, where
  • An identification module configured to identify a service flow according to the service flow feature information, and determine whether the service flow is a target service flow
  • a timestamp obtaining module configured to: add a delay measurement flag to the data packet of the service flow, and obtain data packet delay measurement information corresponding to the delay measurement flag;
  • a determining module configured to determine delay measurement information, where the delay measurement information includes: time stamp information, a service flow identifier, and a TLP identifier, so that the DCP performs the delay measurement after acquiring the delay measurement information
  • the information is sent to the MCP.
  • the timestamp obtaining module adds a delay measurement flag to the data packet of the service flow, and obtains packet delay measurement information corresponding to the delay measurement flag, including: an upstream TLP
  • the timestamp obtaining module is configured to: add a delay measurement flag to the data packet of the target service flow, and obtain a data packet sending end delay measurement information corresponding to the delay measurement flag, where the sending end delay measurement information
  • the time-stamp information, the service flow identifier, and the TLP identifier are sent, so that the DCP that manages the upstream TLP obtains the sender delay measurement information, and sends the sender delay measurement information to the MCP.
  • the timestamp obtaining module of the downstream TLP is configured to: when the identifying module identifies the data packet of the add delay measurement flag, the time stamp obtaining module acquires a data packet receiving corresponding to the delay measurement flag End delay measurement information; the receiving end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier, so that the DCP that manages the downstream TLP acquires the delay information of the receiving end, and then delays the receiving end. Measurement information is sent to the MCP.
  • the foregoing TLP where the method further includes:
  • the TLP is an upstream TLP
  • the time synchronization module is configured to: before the time stamp acquisition module of the upstream TLP adds a delay measurement flag to the data packet of the target service flow, using the NTP or IEEE 1588v2 clock and the DCP managing the upstream TLP Time synchronization
  • the TLP is a downstream TLP;
  • the time synchronization module is specifically configured to: before the identifying the module that adds the delay measurement flag, the NTP or IEEE 1588v2 clock and the DCP of the management downstream TLP before the identification module of the downstream TLP is identified Time synchronization
  • the TLP also includes:
  • a measurement period identifier obtaining module where the measurement period identifier acquisition module of the upstream TLP acquires a measurement period identifier corresponding to the delay measurement flag, so that the DCP that manages the upstream TLP acquires the measurement period identifier
  • the measurement period identifier is sent to the MCP; the measurement period identifier acquisition module of the downstream TLP acquires a measurement period identifier corresponding to the delay measurement flag and a start time of each measurement period, so as to manage the DCP of the downstream TLP.
  • a matching determination is performed, and the measurement period identifier is sent to the MCP.
  • the foregoing TLP where the method further includes:
  • the sending module is configured to send, by the sending module of the upstream TLP, a measurement message to the downstream TLP, where the measurement message includes: sending timestamp information;
  • the receiving module specifically, the receiving module of the downstream TLP receives the measurement packet, generates an arrival time stamp information of the measurement packet, and sends the measurement packet and the arrival time stamp information to the And managing the DCP of the downstream TLP, so that the DCP determines whether the arrival timestamp information and the receiving timestamp information belong to a preset duration range, and if yes, determining the sender timestamp information and the receiving end The timestamp information belongs to the same data packet, and the determination result is sent to the MCP.
  • the foregoing TLP where the timestamp obtaining module of the upstream TLP acquires the delay information of the sending end, further includes: a service flow characteristic information of the sending end and a fragment reassembly identifier of the sending end; the downstream TLP The obtaining the delay measurement information of the receiving end by the time stamp obtaining module further includes: a service flow characteristic information of the receiving end and a fragment reassembly identifier of the receiving end;
  • the DCP that manages the upstream TLP acquires the sender delay measurement information and sends the measurement information to the MCP, and the DCP that manages the downstream TLP acquires the receiver delay measurement information and sends the information to the MCP, so that Determining, by the MCP, that the sender timestamp information and the receiving end timestamp information are corresponding according to the sending end service flow characteristic information, the sending end fragment reassembly identifier, the receiving end service flow characteristic information, and the receiving end fragment reassembly identifier. Time stamp information for the same service flow.
  • the foregoing TLP where the timestamp obtaining module of the upstream TLP adds a delay measurement flag to the data packet of the target service flow, including:
  • the timestamp obtaining module retains a reserved bit of the TOS or a flag in the IP header of the data packet Add a force P delay measurement flag on the bit.
  • the identifying module is specifically configured to identify a service flow according to at least two binary information in the quintuple.
  • an embodiment of the present invention provides an MCP, where
  • a receiving module configured to receive the sending end delay measurement information of the DCP corresponding to the upstream TLP, and the receiving end delay measurement information of the DCP corresponding to the downstream TLP, where the sending end delay measurement information includes the time stamp information and the service flow
  • the sending end delay measurement information includes the time stamp information and the service flow
  • the identifier and the TLP identifier, the receiving end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier
  • a determining module configured to determine a single delay of the network according to the sending end delay measurement information and the receiving end delay measurement information.
  • the foregoing MCP where the method further includes:
  • Step; 'one ... the receiving module comprising:
  • the first receiving unit is configured to receive the sending end delay measurement information of the DCP corresponding to the upstream TLP and the receiving end delay measurement information of the DCP corresponding to the downstream TLP, where the sending end delay measurement information includes the time stamp information.
  • the service flow identifier and the TLP identifier where the receiving end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier;
  • a second receiving unit configured to receive a measurement period identifier sent by the DCP that manages the upstream TLP, and receive a measurement period identifier sent by the DCP that manages the downstream TLP;
  • the determining module further includes:
  • the first matching unit is configured to determine, according to the measurement period identifier that is sent by the DCP that manages the upstream TLP, and the measurement period identifier that is sent by the DCP that manages the downstream TLP, determining the delay measurement information of the sending end and the receiving end Whether the measurement information belongs to the same measurement period;
  • the receiving module is configured to receive the sending end delay measurement information sent by the DCP corresponding to the upstream TLP, and receive the determined and sent DCP corresponding to the downstream TLP.
  • the sending end delay measurement information belongs to the receiving end delay measurement information of the same data packet;
  • the determining module is specifically configured to: according to the sending end delay measurement information and the receiving end The measurement information is extended to determine the single delay of the network.
  • the receiving module is configured to receive the sending end delay measurement information sent by the DCP corresponding to the upstream TLP, where the sending end delay measurement information includes time stamp information and a service flow identifier.
  • the determining module includes:
  • the second matching unit is configured to determine, according to the feature information of the service flow of the sending end, the reassembly identifier of the sending end, the characteristic information of the receiving end service stream, and the reassembly identifier of the receiving end, and determining the time stamp information of the sending end and the receiving end Whether the time stamp information corresponds to time stamp information of the same service flow;
  • a network delay measurement system comprising: the DCP according to any one of the preceding claims, the TLP according to any one of the above, and the MCP according to any one of the above.
  • the network delay measurement method, device, and system of the embodiment of the present invention obtains the delay measurement information obtained by directly measuring the service flow by the at least one TLP, and sends the delay measurement information to the MCP, so that the MCP is configured according to the The relevant information in the delay measurement information determines the network delay condition, and realizes direct delay measurement of the service flow.
  • Embodiment 1 is a flowchart of Embodiment 1 of a network delay measurement method according to the present invention
  • FIG. 2 is a schematic flowchart of an implementation process of Embodiment 3 of a network delay measurement method according to the present invention
  • Embodiment 4 of a network delay measurement method according to the present invention
  • Embodiment 6 of a network delay measurement method according to the present invention
  • FIG. 6 is a flowchart of Embodiment 11 of a network delay measurement method according to the present invention
  • FIG. 7 is a schematic diagram of bidirectional delay measurement according to Embodiment 12 of the network delay measurement method of the present invention
  • FIG. 8 is a schematic structural diagram of Embodiment 1 of the DCP according to the present invention
  • FIG. 9 is a schematic structural diagram of a second embodiment of a DCP according to the present invention.
  • FIG. 10 is a schematic structural diagram of a third embodiment of a DCP according to the present invention.
  • FIG. 11 is a schematic structural view of a first embodiment of a TLP according to the present invention.
  • FIG. 12 is a schematic structural view of a second embodiment of a TLP according to the present invention.
  • FIG. 13 is a schematic structural view of a third embodiment of the TLP of the present invention.
  • Embodiment 1 of an MCP according to the present invention is a schematic structural diagram of Embodiment 1 of an MCP according to the present invention.
  • Embodiment 15 is a schematic structural diagram of Embodiment 2 of an MCP according to the present invention.
  • Embodiment 4 of an MCP according to the present invention is a schematic structural diagram of Embodiment 4 of an MCP according to the present invention.
  • Embodiment 17 is a schematic structural diagram of Embodiment 1 of a network delay measurement system according to the present invention.
  • FIG. 18 is a schematic diagram of Embodiment 2 of a network delay measurement system according to the present invention.
  • the technical solutions in the embodiments of the present invention will be clearly described below with reference to the accompanying drawings in the embodiments of the present invention. Examples are some embodiments of the invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
  • Embodiment 1 is a flowchart of Embodiment 1 of a network delay measurement method according to the present invention. As shown in FIG. 1, the method in this embodiment may include:
  • the delay measurement information includes: time stamp information, a service flow identifier, and a TLP identifier.
  • the Data Collecting Point obtains the delay measurement information obtained by measuring the service flow by at least one Target Logical Port (hereinafter referred to as TLP).
  • TLP corresponds to a network upstream sender entry or a network downstream receiver exit.
  • the DCPs corresponding to the TLPs are deployed on the upstream sending device and the downstream receiving device, and are used to read the delay measurement information of the TLPs on the devices where the DCPs are located. That is, when the TLP corresponds to the upstream portal of a certain network, the DCP corresponding to the TLP is deployed upstream of the corresponding network. The upstream port of the port is sent to the upstream device.
  • the DCP corresponding to the TLP is deployed on the downstream sending device corresponding to the downstream end of the network.
  • the service flow When a service flow enters the network, the service flow correspondingly generates a unique service flow identifier.
  • the service flow ID may be used as the service flow identifier. Since a service flow identifier uniquely corresponds to a service flow, when the service flow transmission scenario of the network occurs in a single point-to-multipoint or multi-point-to-multipoint, that is, the DCP on the upstream sending device and the downstream receiving device acquires multiple
  • the TLP measures the delay measurement information obtained by the service flow, it may determine, according to the service flow identifier, whether the relevant delay measurement information belongs to the same service flow.
  • Each of the delay measurement information includes the associated TLP identifier, so that when the time delay measurement information obtained by measuring the service flow of the plurality of TLPs is obtained, the measurement control point (MCP) can be based on each TLP.
  • the identification distinguishes delay measurement information from different TLPs.
  • S102 Send the delay measurement information to the measurement control point MCP, so that the MCP determines the network delay according to the timestamp information, the service flow identifier, and the TLP identifier.
  • the process of acquiring and transmitting the delay measurement information by the DCP is implemented by the management network of the network device.
  • a management port is also provided, and the management port is also adopted. That is, the above management network can be formed.
  • the transmission path of the delay measurement information may be transmitted out-of-band through the management network, or may be transmitted in-band through the same path as the transmission path of the target service flow.
  • the management network may use the virtual private network (Virtual Private Network (hereinafter referred to as VPN), Digital Communication Network (DCN) or public network with IP reachability.
  • VPN Virtual Private Network
  • DCN Digital Communication Network
  • the network delay measurement method provided in this embodiment obtains the delay measurement information of the at least one TLP measurement service flow by using the DCP, and implements the direct measurement of the service flow, and the DCP sends the delay measurement information to the MCP uniformly.
  • the MCP determines the network delay based on the relevant information in the delay measurement information.
  • the MCP is configured to determine the delay of the service flow according to the delay measurement information according to the transmission delay measurement information. Therefore, in the scenario where the network is single-point-to-single-point or single-point-to-multipoint, the traffic flow can be accurately measured for the delay of the service flow, and the real delay of the service flow is reflected.
  • the DCP obtains a delay measurement signal obtained by measuring at least one TLP for the service flow.
  • Interest including:
  • the DCP obtains the sender delay measurement information obtained by measuring at least one upstream TLP for the transmitted service flow.
  • the DCP managing the downstream TLP obtains the receiver delay measurement information obtained by measuring at least one downstream TLP for the received service flow.
  • the DCP managing the upstream TLP sends the sender delay measurement information to the MCP, and the sender delay measurement information includes the sender time stamp information, the service flow identifier, and the TLP identifier.
  • a DCP is deployed on each of the upstream sending ends, and the TLPs on the upstream sending ends are managed by the DCPs, and the upstream TLP identifies the sent service flows, and the data packets of the service flows are After the delay measurement flag is added, the time point of the delay measurement flag is the time-stamp information of the sender, and the upstream TLP generates the delay information of the sender.
  • the DCPs obtain the delay measurement information of the sender of the upstream TLP.
  • the sender delay measurement information includes: at least one TLP statistics of the sender time stamp information, the service flow identifier, and the TLP identifier of the upstream sender.
  • the DCP that manages the downstream TLP sends the receiver delay measurement information to the MCP.
  • the receiver delay measurement information includes the receiver time stamp information, the service flow identifier, and the TLP identifier.
  • the DCP is deployed on each of the downstream receiving ends, and the TLPs on the downstream receiving ends are managed by the DCPs, and the service flows are first identified by the downstream TLPs, if the service flows are The target traffic flow, when the downstream TLP identifies the data packet with the delay measurement flag, uses the time point as the receiving timestamp information, and generates the receiving end delay measurement information, and the DCP obtains the downstream TLP to measure the received service flow.
  • the received receiver delay measurement information is obtained.
  • the receiving end delay measurement information includes: the receiving end time stamp information, the service flow identifier, and the TLP label of the at least one TLP of the downstream receiving end.
  • the method for the network delay measurement method according to the second embodiment of the present invention further includes:
  • the DCP managing the upstream TLP obtains the measurement period identifier at the end of the measurement period and sends the measurement period identifier to the MCP.
  • the DCP managing the downstream TLP obtains the start time of the measurement period, if the start time and time stamp If the difference between the information is less than or equal to the preset duration, the measurement information of the receiver delay is the measurement information corresponding to the measurement period identifier; if the difference between the start time and the time stamp information is greater than the preset duration, the measurement period identifier is incremented by one, and the time stamp is added.
  • the information belongs to the next measurement cycle, and the measurement cycle identifier is sent to the MCP.
  • the preset duration is 2/3 of the measurement period.
  • the obtaining the measurement period identifier may be obtained by directly using the delay measurement information by the upstream TLP and the downstream TLP, or by measuring the DCP of the upstream TLP and the DCP managing the downstream TLP. After receiving the delay information of the receiving end, the measurement period identifier is obtained according to the time point of obtaining the measurement information of the delay of the transmitting end and the time point of obtaining the measurement information of the delay of the receiving end.
  • the DCP and the corresponding TLP can generate a corresponding measurement period identifier.
  • the formula for the DCP to obtain the period identification is as follows:
  • Measurement period identification global seconds / measurement period duration.
  • the upstream TLP and the DCP managing the upstream TLP, the downstream TLP, the DCP managing the downstream TLP, and each DCP are time synchronized by using the Network Time Protocol (NTP) or the IEEE 1588v2 clock.
  • the global seconds may be the time point at which the TLP generates the delay measurement information, or may be the time point at which the DCP reads the delay measurement information, and the measurement period is determined by dividing the global seconds by the result of the measurement period duration.
  • Each measurement cycle time is 2s.
  • a measurement period identifier is generated for each measurement period, for example, assume that the upstream TLP pair packet A Adding a delay measurement flag, the uplink delay measurement information is generated by the upstream TLP, and a corresponding measurement period identifier is generated, and the measurement period identifier is 10, in order to ensure that the MCP is based on the sender time stamp information corresponding to the data packet A and the receiving end. The time delay is calculated by the stamp information.
  • the DCP managing the upstream TLP obtains the sender timestamp information and the measurement period identifier generated by the upstream TLP at the end of the measurement period, and sends the sender timestamp information and the measurement period identifier (ie, 10).
  • the downstream TLP identifies the data packet A with the delay measurement flag in one measurement period, and generates the receiving end delay measurement information and A measurement period identifier is sent, and the receiver delay measurement information and the measurement period identifier are sent to the DCP managing the downstream TLP, and the DCP managing the downstream TLP is determined, if the difference between the start time and the time stamp information is less than or equal to the preset duration, Then, the measurement period identifier is 10, and the receiver delay measurement information and the measurement period identifier 10 are sent to the MCP.
  • the DCP adds the measurement period identifier to the first identifier. That is, the measurement period identifier is 11, and then the receiver delay measurement information and the new measurement period identifier 11 are sent to the MCP, and the MCP determines the receiver delay measurement information and the management upstream TLP according to the measurement period identifier 11.
  • the sender delay measurement information corresponding to the measurement period of the DCP is 11.
  • the DCP managing the upstream TLP is time synchronized with the upstream TLP by using the NTP or the IEEE 1588v2 clock, and the DCP managing the downstream TLP is used by the NTP or the IEEE 1588v2.
  • the clock is time synchronized with the downstream TLP.
  • the DCP that manages the upstream TLP and the DCP that manages the downstream TLP also implement time synchronization by using NTP or IEEE 1588v2 clocks.
  • NTP Network Time Protocol
  • IEEE 1588v2 clock is a high-precision clock that uses the IEEE 1588v2 protocol.
  • the time synchronization method according to the present invention is based on a common time reference (NTP or IEEE and the local time of the DCP managing the downstream TLP, and the local time between the DCP managing the upstream TLP and the DCP managing the downstream TLP.
  • the boundary points of various cycles are agreed by NTP or IEEE 1588v2 clock, that is, the upstream TLP and the DCP and the downstream TLP managing the upstream TLP and the DCP managing the downstream TLP are measured for each measurement cycle.
  • the network delay measurement method provided in this embodiment can optionally use the IEEE 1588v2 clock for time synchronization.
  • FIG. 2 is a schematic flowchart of the implementation of the network delay measurement method in the third embodiment of the present invention.
  • the third embodiment of the network delay measurement method of the present invention is described in detail below with reference to FIG.
  • R1 and R2 are network node devices.
  • the TLPs are deployed on R1 and R2, and the corresponding DCPs are deployed.
  • the MCPs are deployed on any network node device on the network.
  • the MCPs are deployed in functions. Strong on the node device. Referring to Figure 2, for R1 and R2, there can be The two opposite traffic flows are measured for delay.
  • T[N+1] represents the respective measurement period identifiers corresponding to two adjacent measurement period intervals.
  • the time axes of the two local times for R1 and R2 and the boundary points of each measurement period such as T[N], T[N+1] have been basically aligned by NTP or IEEE 1588v2 clocks, two of which The deviation of the local time axis is caused by the error of the network itself, the accuracy of the NTP or IEEE 1588v2 clock.
  • T[N] of the same measurement period identifier at both ends of R1 and R2, at the start and end of the measurement period (including the range of the pre-T/n period) on each TLP of the transmitting end and the receiving end of R1 and R2, A one-way delay measurement for the service flow data packet is initiated in the opposite direction, and only one data packet in the target service flow is added to each measurement cycle to add a delay measurement flag.
  • the transmitting end TLP adds a delay measurement flag to a data packet, and obtains a local transmission time stamp tl, t3, and a downstream receiving end TLP at R1 and R2, in the corresponding measurement period T[N], detecting
  • the downstream TLP can obtain the local receiving timestamps t2 and t4, respectively, and the TLP can report the delay measurement information including the timestamp information to the DCP managing the TLP, or at the end of each period.
  • the delay measurement information is read by the DCP managing each TLP.
  • All delay measurement information includes: timestamp information, traffic flow identification, and TLP identification. And carrying the same period identifier ⁇ [ ⁇ ], the service flow identifier and the TLP identifier embodies the receiving or sending direction information, the delay measurement information is read by the DCP and sent to the MCP, and the MCP can match and calculate according to T[N] deal with.
  • the equation also means that time-sequence coupling is not necessarily required between two one-way delay measurements, so it can be initiated and measured separately.
  • the time-stamp of the transmitting end of the N-th period is TX, and within 100 ms from the start time of each measurement period, if there is a service flow, the traffic flow is first.
  • the data packet adds a delay measurement flag, the upstream TLP records the time stamp Time_TX[N], and obtains the measurement period identifier N; otherwise, the weektime does not mark the delay measurement.
  • the receiving end receives the data packet of the time delay measurement flag, and the downstream TLP records the local time stamp Time_RX. If the downstream TLP obtains Time-RX at the end of this period, the DCP managing the downstream TLP is calculated as follows:
  • Time— RX-Time[N] (Time[N] is the start time of the current period, and is synchronized with the managed TLP by DCP with NTP or IEEE 1588v2 clock, so DCP can directly obtain the DCP local. Cycle start time in time);
  • the timestamp belongs to the next period (measurement period identifier +1: advance packet due to synchronization error), otherwise, the timestamp information belongs to the current period;
  • the obtained Time-RX is the receiving timestamp of this period N Time_RX[N];
  • the delay measurement period T is selected to be greater than 1 second.
  • the network delay measurement method obtains the measurement period identifier and the delay information of the transmission end at the end of the measurement period by managing the DCP of the upstream TLP, and sends the measurement period identifier to the MCP, and then passes the downstream TLP.
  • the DCP obtains the measurement period identifier and the receiver delay measurement information, and the DCP determines the obtained measurement period identifier, and then sends the determined measurement period identifier and the receiver delay measurement information to the MCP, so that the MCP is based on the upstream.
  • the measurement period identifier and the downstream measurement period indication are used to correlate the transmission delay measurement information and the receiver delay measurement information in the same period of the same service flow, and directly and accurately measure the service flow delay.
  • the method of the fourth embodiment of the network delay measurement method of the present invention further includes:
  • the DCP that manages the downstream TLP obtains the measurement packet sent by the upstream TLP and the arrival timestamp information of a measurement packet that is generated when the measurement packet arrives at the downstream TLP.
  • the measurement packet includes: .
  • the traffic flows in units of measurement cycles.
  • the line delay measurement, but the measurement period of the receiving end and the measurement period of the transmitting end are not time-synchronized by the time synchronization tool.
  • the upstream TLP adds a delay measurement flag to one data packet in each measurement period and generates a transmission delay.
  • the measurement information includes: the timestamp information, the service flow identifier, and the TLP identifier.
  • the upstream TLP sends the measurement packet including the sender timestamp information to the downstream TLP of the receiving end, and the DCP that manages the downstream TLP passes the The arrival time stamp information of the measurement data is compared with the measurement information of the delay of the receiving end, so that the measurement information of the delay of the transmitting end and the measurement information of the delay of the receiving end belong to the same measurement period.
  • the DCP of the downstream TLP is configured to perform matching identification, and determine whether the arrival time stamp information and the receiving time stamp information belong to the preset duration range. If yes, determine the sender time stamp information in the measurement packet and the receiving end generated by the downstream TLP.
  • the stamp information belongs to the same measurement period, that is, belongs to the same data packet (since only one packet is added per cycle to add a delay measurement flag), and the determination result is sent to the MCP.
  • the DCP managing the downstream TLP may send the sender timestamp information and the receiver timestamp information that belong to the same measurement period to the MCP, and the MCP performs the calculation.
  • the DCP may also directly determine the delay of the periodic data packet according to the timestamp information of the transmitting end and the timestamp information of the receiving end that belong to the same measurement period, and then send the calculated delay time to the MCP.
  • FIG. 3 is a schematic flowchart of the implementation process of the fourth embodiment of the network delay measurement method according to the present invention.
  • Ds when a data packet with a delay measurement flag added is transmitted to the receiving end through the network, there is a delay.
  • Dc when the measurement packet carrying the timestamp information of the sender sent by the upstream TLP arrives at the receiving end, there is also a delay Dc.
  • a delay difference is defined by the following formula, and the formula is as follows: Delay difference Z ⁇ Ds - Dc
  • the delay measurement flag is added to the service flow packet A at the upstream TLP of the transmitting end (TX), and the local transmitting end is obtained.
  • Poke information t1 and generate a measurement message carrying the sender timestamp information t1 and send it to the downstream TLP of the receiving end (RX).
  • the packet A may arrive first, and The measurement packet arrives first.
  • the downstream TLP of the receiving end obtains the time stamp information t2 of the service flow at the receiving end.
  • the measurement packet carrying the tl arrives at the receiving end, and receives The terminal obtains the arrival time stamp information tc of the measurement message.
  • the measurement message first reaches the downstream TLP the principle is the same.
  • the matching identification setting of the DCP managing the downstream TLP is
  • the DCP managing the downstream TLP first acquires the timestamp information t2 of the receiving end, and manages the downstream TLP at a certain time tc within the time range of adding or subtracting the delay difference based on the receiving timestamp information t2.
  • the sender timestamp information tl can match the receiver timestamp information t2, that is, the sender timestamp information and the receiver timestamp information belong to the same data packet of the same measurement period, or manage the downstream TLP.
  • the DCP first obtains the measurement message, and the DCP that manages the downstream TLP acquires the time-stamp information t2 of the data packet of the downstream TLP in the range of the time-delay time of the measurement message, based on the arrival time-stamp information tc of the measurement message.
  • the sender time stamp information t1 can match the receiver time stamp information t2, that is, the sender time stamp information and the receiver time stamp information belong to the same data packet of the same measurement period.
  • each delay measurement period interval (the interval between the upstream TLP and the service flow data packet adding the delay measurement flag is T)
  • the upstream TLP adds a delay measurement flag to only one service flow data packet, due to the actual network.
  • Ds and Dc have jitter and variable length, but limited, there is a delay difference / maximum value Z (MAX), as long as the measurement period interval ⁇ > 2 * ⁇ ( ⁇ ) + minimum safety interval time, you can determine each The t2 time stamp in the sample interval matches the tl protocol packet;
  • the network delay measurement method obtaineds the measurement packet sent by the upstream TLP received by the at least one downstream TLP by using the DCP of the downstream TLP, and performs matching identification by the DCP managing the downstream TLP, and determines the sender time stamp. Whether the information and the time-stamp information of the receiving end belong to the preset duration range. If yes, it is determined that the time-stamp information of the sending end and the time-stamp information of the receiving end belong to the same measurement period, and the determination result is sent to the MCP, so that the accurate and direct measurement service is realized. The delay of the flow.
  • the method for the network delay measurement method of the second embodiment of the present invention, the method for the fifth embodiment of the network delay measurement method of the present invention further includes:
  • the sender delay measurement information further includes: a service flow characteristic information of the sender and a fragment reassembly identifier of the sender
  • the receiver delay measurement information further includes: a service flow characteristic information of the receiver and a fragment reassembly identifier of the receiver, so that the MCP is based on
  • the sender service flow characteristic information, the sender fragment reassembly identifier, the receiver service stream feature information, and the receiver end fragment reassembly identifier, the sender time stamp information and the receiver end time stamp information are time stamp information corresponding to the same service flow.
  • the key is to determine that the time-stamp information of the sender and the time-stamp information of the receiver are the same service flow packet with the added delay measurement flag, and the packet is delayed at the transmitting end of the network. Obtained separately from the receiving end.
  • the service flow characteristic information and the fragmentation reassembly identifier contained in the data packet can uniquely identify the data packet.
  • the service flow characteristic information is information of a quintuple in the IP header and a service type TOS (Type of Service, hereinafter referred to as TOS) field, where the quintuple refers to the source IP address, the destination IP address, and the protocol type in the IP header.
  • TOS Type of Service
  • each sub-packet may be re-identified according to the fragment re-identification of each sub-packet.
  • the sub-packet is reorganized into the original data packet.
  • the upstream TLP adds a delay measurement flag to a data packet A of the service flow, and generates a sender delay measurement information including a sender time stamp information, a service flow identifier, a TLP identifier, The sender service flow feature information and the sender fragment reassembly identifier are obtained, and the sender delay measurement information is obtained by the DCP managing the upstream TLP, and sent to the MCP.
  • the generating terminal delay measurement information includes the receiving end time stamp information, the service flow identifier, the TLP identifier, the receiving end service flow characteristic information, and the receiving end fragment reassembly identifier. And receiving the delay measurement information of the receiver by the DCP managing the downstream TLP, and sending the information to the MCP.
  • the MCP can determine that the sender timestamp information and the receiver timestamp information are the same added delay measurement according to the sender service flow characteristic information, the sender fragment reassembly identifier and the receiver service stream feature information, and the receiver end fragment reassembly identifier.
  • the service flow data packet of the flag is obtained by the delay at the transmitting end and the receiving end of the network respectively, that is, the matching between the delay measurement information of the transmitting end and the measurement information of the delay of the receiving end is implemented. Therefore, the MCP performs delay measurement according to the matching sender delay measurement information and the receiver delay measurement information.
  • FIG. 4 is a schematic flowchart of an implementation process of a network delay measurement method according to Embodiment 5 of the present invention.
  • the fifth embodiment of the network delay measurement method of the present invention is described in detail below with reference to FIG.
  • the key is to determine that the timestamp information of the transmitting end and the timestamp information of the receiving end are obtained when the service flow data packet of the same added delay measurement flag passes through the network, and the quintuple of the data packet is used in this embodiment.
  • fragment reassembly identifier sharding reassembly ID
  • the quintuple of the IP packet can determine a traffic flow within a certain period of time (an ID loop).
  • the fragment reassembly ID (unfragmented) of the packet on the service stream determined by a quintuple is unique. Therefore, a data packet of a service flow within a measurement domain (which can contain multiple different service flows) can uniquely determine the data of the service flow at the sender and the receiver through the quintuple quintuple + fragment reassembly ID. Packet (for the same service flow, the fragmentation reassembly ID of each packet is different; different traffic flows, the quintuple of each service flow is different).
  • the DCP read delay measurement information of each TLP is managed on the transmitting end and the receiving end, and the service flow characteristic information (quintuple) and the fragment reassembly identifier (segment reassembly ID) are read, and the same service flow can be matched. Timestamp information for the packet.
  • the DCP that manages each TLP reads the timestamp information of the first packet received (the out-of-order and delay of the fragmented packets is usually smaller than the measurement period, and the host The side slice reassembly ID cycle will take longer than the measurement period).
  • the service flow data packet determined by the quintuple, at the transmitting end, the fragment reassembly ID is 100, and the receiving end is still 100 unchanged. Through this constant characteristic, the time obtained by both ends can be determined.
  • the stamp information is the result of the detection of the same packet.
  • the delay measurement information carries a local period identifier (no time synchronization is required), and the order of the delay information and the pairing of the two-way delay measurement may be further determined.
  • the network delay measurement method obtains the transmission delay measurement information obtained by the upstream TLP measurement by the DCP of the upstream TLP, and sends the information to the MCP.
  • the DCP that manages the downstream TLP obtains the receiver delay measurement information obtained by the downstream TLP measurement, and sends the information to the MCP, because the sender delay measurement information further includes the sender service flow characteristic information and the sender fragment reassembly identifier, and the receiver end time
  • the delay measurement information further includes the service flow characteristic information of the receiving end and the fragment reassembly identifier of the receiving end, so that the MCP according to the characteristic information of the sending end service flow, the fragment reassembly identifier of the transmitting end, the characteristic information of the receiving end service stream, and the fragment recombination identifier of the receiving end.
  • time-stamp information at the transmitting end and the time-stamp information at the receiving end are the same traffic flow data packet with the added delay measurement flag obtained by the delay at the transmitting end and the receiving end of the network respectively, thereby performing fast and accurate delay. measuring.
  • the foregoing embodiment describes the specific method S performed by the DCP in the network delay measurement method of the present invention.
  • the following describes the specific method S performed by the TLP in the network delay measurement method of the present invention. Bright.
  • FIG. 5 is a flowchart of Embodiment 6 of a network delay measurement method according to the present invention. As shown in FIG. 5, the method in this embodiment may include:
  • S200 Identify the service flow according to the service flow characteristic information, and determine whether the service flow is a target service flow.
  • the TLP is deployed on each of the upstream and downstream receiving ends.
  • the upstream TLP and the downstream TLP can be deployed on the user side or the network side of the sending end and the receiving end.
  • Each of the service flows has its own specific service flow characteristic information.
  • the service flow characteristic information the network delay measurement method is described in detail in Embodiment 5, and details are not described herein again. Therefore, when a service flow enters the network, the upstream sending port TLP first identifies the service flow according to the service flow characteristic information, and the identification process is based on matching the service flow feature information with the packet header information of the service flow. If the two match is successful, the upstream sending port TLP determines that the service flow is the target service flow.
  • each upstream sender may be determined according to the service flow characteristic information of the service flow. Whether the data packets on each downstream receiving end belong to the same service flow.
  • the TLP adds a delay measurement flag to the data packet of the target service flow, and generates delay measurement information, where the delay measurement information includes timestamp information, a service flow identifier, and a TLP identifier, where the timestamp information is added by the TLP.
  • the time point at which the time delay is measured For the service flow identifier and the TLP identifier, the network delay measurement method embodiment 1 has been described in detail, and is not mentioned here.
  • the DCP sends the delay measurement information to the MCP after acquiring the delay measurement information, so that the MCP determines the delay according to the delay measurement information.
  • the network delay measurement method identifies the service flow according to the service flow characteristic information by using the TLP, and determines whether the service flow is the target service flow. If yes, the TLP adds a delay measurement flag to the data packet of the service flow. Obtaining the packet delay measurement information corresponding to the delay measurement flag, and determining the delay measurement information by the TLP, so that the DCP sends the delay measurement information to the MCP after acquiring the delay measurement information, so that the MCP is based on the time. Deferred measurement information is delayed. Set. It realizes the delay measurement of the service flow data packet directly, and improves the accuracy and authenticity of the delay measurement.
  • the sixth embodiment of the network delay measurement method of the present invention optionally, the method for the seventh embodiment of the network delay measurement method of the present invention adds a delay measurement flag to the data packet of the service flow, and obtains a corresponding to the delay measurement flag.
  • Packet delay measurement information including:
  • the upstream TLP adds a delay measurement flag to the data packet of the target service flow, and obtains packet delay measurement information corresponding to the delay measurement flag, where the delay measurement information includes time stamp information, service flow identifier, and TLP.
  • the identifier is sent to the MCP after the DCP that manages the upstream TLP obtains the sender delay measurement information and sends the delay information to the MCP.
  • the packet receiving end delay measurement information corresponding to the delay measurement flag is obtained.
  • the receiver delay measurement information includes the timestamp information, the service flow identifier, and the TLP identifier, so that the DCP that manages the downstream TLP obtains the receiver delay measurement information and then sends the receiver delay measurement information to the MCP.
  • a delay measurement flag is added to the reserved bits of the TOS or the reserved bits of the Flags in the IP header of the packet.
  • the delay measurement flag can specify a range of 6 bits in the TOS and Flags fields of the packet IP header, 3 to 7 bits of the TOS, and 0 bits of the flags.
  • several bits (3rd to 7th bits) of the TOS are often not used in different specific networks, especially the 6th and 7th bits, which are rarely used, so these bits of the IP header can be borrowed for adding the identifier.
  • the 0th bit of Flags is the currently unique reserved bit in the IP header. In the normal IP header, this bit can be used to add the identifier of the data packet.
  • the identifying the service flow according to the service flow characteristic information may include: identifying the service flow according to at least two pieces of information in the quintuple.
  • the quintuple refers to the source IP address in the IP header or its IP address prefix, the destination IP address or its IP address prefix, the protocol type, the source protocol port number, and the destination protocol port number, except for the quintuple.
  • the information of the TOS field in the IP header may be added to specify the service flow feature information, and the above fields may all be specified, so that the measured service flow is finer; or may be partially specified, for example, in the specified quintuple At least one of the source IP address and the destination IP address; or the source IP address prefix or the destination IP address prefix; or the source IP address or its IP address prefix, the destination IP address or its IP address prefix, and the service type (Type of Service) , hereinafter referred to as TOS) letter Interest.
  • TOS service type
  • the method for the network delay measurement method of the seventh embodiment of the present invention further includes:
  • the NTP or IEEE 1588v2 clock is used to synchronize the time with the DCP managing the upstream TLP.
  • the downstream TLP identifies the data packet with the added delay measurement flag, Time synchronization with the DCP managing the downstream TLP with NTP or IEEE 1588v2 clock.
  • the upstream TLP adds a delay measurement flag to a data packet of the service flow in each measurement period interval in the measurement period, and generates a transmission delay measurement information and a measurement period identifier, and the downstream TLP is also measured by the measurement period. Identifying the data packet with the added delay measurement flag and generating the receiver delay measurement information and the measurement period identifier.
  • the key is to determine that the sender time stamp information is the same as the receiver time stamp information.
  • the data packet of the delay measurement flag is obtained by the upstream TLP and the downstream TLP after being transmitted through the network.
  • the time synchronization between the DCPs managing the downstream TLPs can be managed when the external time synchronization tool NTP or IEEE 1588v2 is deployed on the upstream TLP and the downstream TLP before the delay measurement is performed on the upstream TLP and the downstream TLP.
  • the DCP that manages the upstream TLP and the DCP that manages the downstream TLP also deploy an external time synchronization tool NTP or IEEE 1588v2 clock to ensure time synchronization between the TLP and the DCP, DCP, and DCP, thus ensuring each measurement.
  • the period, the sender measurement period identifier generated by the upstream TLP and the receiver measurement period identifier generated by the downstream TLP can be matched, thereby ensuring that the sender delay measurement information and the receiver delay measurement information with the same measurement period identifier can be matched, Make the MCP accurately determine the delay.
  • the upstream TLP adds a delay measurement flag to the data packet of the target service flow, and obtains a measurement period identifier corresponding to the delay measurement flag, so that the DCP that manages the upstream TLP obtains the measurement period identifier and sends the measurement period identifier information to the MCP. Because both the upstream TLP and the downstream TLP use the NTP or IEEE 1588v2 clock to synchronize time with the DCP managing the upstream TLP and the DCP managing the downstream TLP, it can also be read by the DCP managing the upstream TLP and the DCP managing the downstream TLP.
  • the DCP of the upstream TLP obtains the corresponding measurement period identifier according to the read delay measurement information of the sender, and manages the DCP of the downstream TLP according to the received receiver delay measurement.
  • the information acquires the corresponding measurement period identifier, and the two measurement period identifiers are consistent for the data packet to which the delay measurement flag is added for the same measurement period.
  • the upstream TLP adds a delay measurement flag to the first data packet of the target service flow, and records the current timestamp information tl And obtain the measurement period identifier T[N], as shown in Figure 2. If the upstream TLP does not recognize the target traffic flow, the measurement period does not add a delay measurement flag to the data packet.
  • the downstream TLP acquires the measurement period start time in each measurement period, and when the data packet with the added delay measurement flag is identified in each measurement period, the measurement period identifier corresponding to the delay measurement flag is obtained, so that the downstream TLP is managed. After the DCP acquires the start time and the measurement period identifier, the measurement period identification information is sent to the MCP.
  • the downstream TLP records the start time to of the measurement period at the beginning of the Nth measurement period, and if the data with the delay measurement flag is identified in the measurement period Packet, the downstream TLP records the current timestamp information t3, and generates the measurement period identifier M and the receiver delay measurement information, and at the end of the Nth measurement period, the receiver delay measurement information is acquired by the DCP managing the downstream TLP.
  • the measurement period identifier T[N] and the start time t0, the receiver delay measurement information includes: time stamp information t3, a service flow identifier, and a TLP identifier, and the DCP pair time stamp information t3 and the start time are managed by the downstream TLP.
  • T0 is calculated. If t3-tO ⁇ 2T/3, the timestamp information t3 belongs to the Nth measurement period. If t3-tO ⁇ 2T/3, the timestamp information t3 belongs to the (N+1)th measurement period.
  • the period flag T[N] is incremented by 1 to T[N+1], so that in the case that the time synchronization error causes the packet to be out of order, the MCP can still receive the management of the downstream TLP.
  • the end delay measurement information and the measurement period identifier are determined, according to the measurement period, it is determined that the time stamp information of the sender end and the time stamp information of the receiver end are time stamp information corresponding to the same data packet, thereby accurately and directly performing the delay measurement.
  • the TLP identifies the service flow data packet, and adds a delay measurement flag to the data packet in the measurement period interval in each period TLP, optionally, in each During the measurement period, the TLP adds a delay measurement flag to only one packet.
  • the network delay measurement method provided by the embodiment of the present invention uses the NTP or IEEE 1588v2 clock to manage the upstream before the delay measurement flag is added to the data packet of the target service flow by the upstream TLP.
  • the DCP of the TLP performs time synchronization.
  • the downstream TLP synchronizes the time of the downstream TLP with the downstream TLP by using the NTP or the IEEE 1588v2 clock to synchronize the time of the upstream TLP with the NTP or the IEEE 1588v2 clock.
  • the MCP determines that the sender time stamp information and the receiver time stamp information correspond to the same according to the same period measurement identifier.
  • the time stamp information of the data packet and in the case of packet out-of-order caused by the time synchronization error, the downstream TLP acquires the measurement cycle start time in each measurement cycle, and the added delay measurement is identified in each measurement cycle.
  • the measurement period identifier corresponding to the delay measurement flag is obtained, so that after the DCP that manages the downstream TLP acquires the start time and the measurement period identifier, the DCP determines the correct according to the start time and the time stamp information of the receiving end.
  • the measurement period identifier and send the correct measurement period identification information to the MCP to ensure that the MCP is accurate. Delay the case.
  • the method according to the seventh embodiment of the network delay measurement method of the present invention further includes:
  • the measurement packet is sent by the upstream TLP to the downstream TLP, and the measurement message includes: the timestamp information of the sender.
  • the receiving module of the downstream TLP receives the measurement packet, generates an arrival timestamp information of the measurement packet, and sends the measurement message and the arrival timestamp information to the DCP managing the downstream TLP, so that the DCP determines the arrival timestamp information and Whether the timestamp information of the receiving end belongs to the preset duration range, and if yes, it is determined that the timestamp information of the sending end and the timestamp information of the receiving end belong to the same measurement period, and the determination result is sent to the MCP.
  • the upstream TLP performs time delay measurement on the service flow in units of the measurement period T, but the measurement periods of the upstream TLP and the downstream TLP are not time synchronized by the time synchronization tool.
  • the upstream TLP generates a sender delay measurement information in units of measurement periods, where the delay measurement information includes the time stamp information, the service flow identifier, and the TLP identifier, and the DCP that manages the upstream TLP obtains the sender.
  • the upstream TLP When the delay measurement information is generated, the upstream TLP generates a measurement packet containing the sender timestamp information t1 and sends it to the downstream TLP of the receiving end.
  • the downstream TLP generates a receiver delay measurement information and receives it in one measurement period.
  • the measurement packet is obtained by the DCP managing the downstream TLP, and the measurement information and the measurement packet of the receiving end are compared, and the measurement message is compared with the measurement information of the delay of the receiving end, and the specific comparison method and technical solution are
  • the fourth embodiment of the network delay measurement method has been described in detail, and details are not described herein again.
  • the TLP identifies the service flow data packet, and adds a delay measurement flag to the data packet in the measurement period interval in each period TLP, optionally, in each During the measurement period, the TLP adds a delay measurement flag to only one packet.
  • the network delay measurement method provided by the embodiment of the present invention sends a measurement message to the downstream TLP by the upstream TLP, where the measurement message includes: the timestamp information of the sending end, and then the downlink TLP sends the received measurement message to the DCP. So that the DCP determines whether the timestamp information of the transmitting end and the timestamp information of the receiving end belong to the preset duration range. If yes, it is determined that the timestamp information of the sending end and the timestamp information of the receiving end belong to the same measurement period, and the determination result is sent to The MCP ensures that the measurement information of the delay of the transmitting end and the measurement information of the delay of the receiving end belong to the same measurement period, and implements a direct and accurate delay measurement for the service flow.
  • the method for the network delay measurement method of the present invention is based on the seventh embodiment of the network delay measurement method of the present invention, and further includes:
  • the sender delay measurement information further includes: a service flow characteristic information of the sender and a fragment reassembly identifier of the sender; the receiver delay measurement information further includes: a service flow characteristic information of the receiver and a fragment reassembly identifier of the receiver.
  • the DCP that manages the upstream TLP obtains the sender delay measurement information and sends it to the MCP, and the DCP that manages the downstream TLP acquires the receiver delay measurement information and sends the measurement information to the MCP, so that the MCP according to the feature information of the sender and the sender end.
  • the slice reassembly identifier, the receiving end service flow characteristic information, and the receiving end fragment reassembly identifier, and determining the sender time stamp information and the receiving end time stamp information are time stamp information corresponding to the same service flow.
  • the key is to determine that the sender time stamp information and the receiver time stamp information are the same time delay measurement flag packet transmitted by the network, and then the upstream TLP and the downstream TLP. Obtained separately.
  • the upstream TLP when the upstream TLP is identified by the upstream TLP, the upstream TLP adds a delay measurement flag 1 to a data packet A of the target service flow, and the sender delay measurement information generated by the upstream TLP includes the sender.
  • the packet information, the service flow identifier, the TLP identifier, the sender service flow characteristic information, and the sender fragment reassembly identifier are obtained, and the delay measurement information of the sender is obtained by the DCP managing the upstream TLP, and sent to the MCP. Moreover, since the present embodiment does not add a timestamp to the data packet based on the period, the TLP adds a time delay measurement flag to the data packet of one service flow in this embodiment. When the downstream TLP identifies the data packet A with the delay measurement flag, the receiver delay measurement information generated by the downstream TLP includes the receiving time stamp information.
  • the information, the service flow identifier, the TLP identifier, the service flow characteristic information of the receiver, and the fragmentation reassembly identifier of the receiver, and the delay measurement information of the receiver is obtained by the DCP managing the downstream TLP, and sent to the MCP.
  • the MCP can determine that the sender timestamp information and the receiver timestamp information are the same added delay measurement according to the sender service flow characteristic information, the sender fragment reassembly identifier and the receiver service stream feature information, and the receiver fragment reassembly identifier.
  • the service flow data packet of the flag is obtained by the delay at the transmitting end and the receiving end of the network respectively, that is, the matching between the delay measurement information of the transmitting end and the measurement information of the delay of the receiving end is implemented. Therefore, the MCP performs the delay measurement according to the successfully matched sender delay measurement information and the receiver delay measurement information, and the service flow feature information and the fragment reassembly identifier are described in detail in the fifth embodiment of the network delay measurement method. I won't go into details here.
  • the network delay measurement method provided by the embodiment of the present invention after the upstream TLP identifies and delays the measurement of the target service flow data packet, records the service flow characteristic information of the sender and the fragmentation reassembly identifier of the sender, and is recorded by the upstream TLP.
  • the transmitting end delay measurement information is generated, and the sending end delay measurement information provided by the embodiment of the present invention includes not only the sending end time stamp information, the service flow identifier, and the TLP identifier, but also the foregoing sending end service flow characteristic information and the sending end fragmentation. Reconstructing the identifier, so that the DCP that manages the upstream TLP obtains the sender delay measurement information and sends it to the MCP.
  • the downstream TLP also performs a similar operation, so that the DCP managing the downstream TLP obtains the receiver delay measurement information and sends the measurement information to the MCP.
  • the MCP implements matching between the sender delay measurement information and the receiver delay measurement information according to the traffic information characteristic information of the sender and the sender fragment reassembly identifier and the receiver service stream feature information and the receiver end fragment reassembly identifier. This ensures the accuracy of direct delay measurements.
  • FIG. 6 is a flowchart of Embodiment 11 of the network delay measurement method of the present invention. As shown in FIG. 6, the method in this embodiment may include:
  • S300 Receive delay measurement information of the transmitting end of the DCP corresponding to the upstream TLP, and receive delay measurement information of the DCP sent by the downlink TLP.
  • the sender delay measurement information includes the sender timestamp information, the service flow identifier, and the TLP identifier
  • the receiver delay measurement information includes the receiver timestamp information, the service flow identifier, and the TLP identifier.
  • the MCP delays the timestamp information of the sending end of the same data packet and the timestamp information of the receiving end according to the measurement information of the delay of the transmitting end and the measurement information of the delay of the receiving end.
  • the network delay measurement method provided by the embodiment of the present invention may deploy the MCP on any network element node in the entire network.
  • the MCP is deployed on a network element node with strong functions, and the MCP, Each DCP and each TLP is connected based on a management network.
  • the network delay measurement method receives the delay measurement information of the sending end of the DCP corresponding to the upstream TLP and the receiving end delay measurement information of the DCP corresponding to the downstream TLP through the MCP, and then passes the MCP according to the sending end.
  • the delay measurement information and the measurement information of the delay of the receiving end determine the single delay of the network, and realize the direct and accurate delay measurement for the service flow.
  • the method for the network delay measurement method according to the embodiment of the present invention further includes:
  • the MCP receives the measurement period identifier sent by the DCP that manages the upstream TLP, and the MCP receives the measurement period identifier sent by the DCP that manages the downstream TLP, and the measurement period identifier sent by the MCP according to the DCP that manages the upstream TLP and the measurement period identifier sent by the DCP that manages the downstream TLP. And determining whether the sender delay measurement information and the receiver delay measurement information belong to the same measurement period. If yes, the MCP determines the network single delay according to the sender delay measurement information and the receiver delay measurement information.
  • the target service flow measurement data summary table is maintained in the MCP.
  • Table 1 is a summary table of target service flow measurement data provided in this embodiment. The following describes how to determine the network delay in this embodiment according to Table 1:
  • Table 1 is the summary of the target business flow measurement data.
  • Traffic flow time stamp timestamp letter invalid invalid invalid invalid invalid invalid invalid invalid downstream information for the target traffic flow measurement data summary table 1 it should be noted that the concept of the left TLPs and the right TLPs in Table 1 is divided by the network as the boundary After defining one network side as the left side, the TLPs deployed on the left port are the left TLPs, and the corresponding TLPs are defined accordingly.
  • the MCP may use the source IP address or its IP address prefix in the quintuple on the two target service flows, and The destination IP address or its IP address prefix defines one of the target service flows as a forward traffic flow, and then defines another target traffic flow as a reverse traffic flow.
  • the scope of a network includes all devices and networks from the left port device to the right port device.
  • the target service flow A is accessed from each TLP on the left port device, and the TLP is removed from the TLP on the right port device; the target service flow B enters the network from each TLP on the right port device, Each TLP on the left port device leaves the network.
  • the left port device is the sender device, and the left TLP is the upstream sender.
  • the left port device is the receiver device and the left TLP is the downstream receiver.
  • each cycle MCP maintains a data entry of a forward traffic flow and one A data entry of a reverse traffic flow, thereby implementing a function of delay measurement for two reverse traffic flows simultaneously.
  • the DCP reads the delay measurement information of each TLP and sends it to the MCP.
  • the MCP can use the target service flow ID as the service flow identifier to find the corresponding target service flow measurement data summary table according to the service flow identifier.
  • the MCP then merges the data into the corresponding measurement data summary table data item according to the service flow identifier, the measurement period identifier, the TLP identifier, and the TLP identifier.
  • the measurement period identifier may uniquely determine that the sender delay measurement information and the receiver delay measurement information belong to the same data packet, that is, the data packet entry in Table 1, and one measurement period identifier corresponds to A packet of data.
  • the MCP For the MCP to receive the delay measurement information and maintain the process of Table 1, it should be noted that since the delay measurement is for the data packet, only one upstream TLP generated transmission end occurs during the transmission of one data packet. Poke information and receiver timestamp information generated by a downstream TLP, so after referring to Table 1, when the MCP receives a delay measurement information and updates it to the data item of one TLP in the left TLPs by the MCP, for the left of the data packet Other TLP data items of the side TLPs are set to be invalid by the MCP. For the right TLPs, the MCP performs a similar operation.
  • a traffic flow is assumed to be a forward traffic flow, and a data entry for each data packet of the forward traffic flow corresponds to a data arrival alignment flag in the target traffic flow measurement data summary table, in each measurement
  • the data of the measurement period of the measurement period is set to "not arrived" by the MCP, for example, in Table 1, the measurement period identifier
  • the MCP receives the sender delay measurement information obtained by the first TLP sent by the DCP managing the upstream TLP, and in the right TLPs The delay measurement information is not reached.
  • the MCP sets the corresponding data item arrival flag to "upstream to".
  • the MCP receives the downstream measurement TLP sent by the DCP that manages the downstream TLP
  • the MCP update table 1 The time-stamp information of the receiving end carried in the measurement information of the receiving end delay is filled in the corresponding TLP data entry, and the corresponding data item is set to the "all aligned".
  • the MCP When the MCP detects that a forward traffic flow identifier in a measurement period identifier or a data arrival parity flag of a reverse traffic flow is set to "all aligned", the MCP will be based on the corresponding sender time stamp. The delay calculation is performed on the information and the time-stamp information at the receiving end.
  • the MCP is based on the sender timestamp information corresponding to the forward traffic flow, and the receiving timestamp information.
  • the time-stamp information of the sender and the time-stamp information of the receiver corresponding to the reverse traffic flow are experimentally calculated, and the specific formula is as follows:
  • Delay (received timestamp information corresponding to the reverse traffic flow - timestamp information of the sender corresponding to the forward traffic flow) - (the sender's timestamp information corresponding to the reverse traffic flow - the receiver corresponding to the forward traffic flow) Time stamp information)
  • FIG. 7 is a schematic diagram of bidirectional delay measurement according to Embodiment 12 of the network delay measurement method of the present invention. Referring to FIG. 7, the calculation principle and method of the bidirectional delay are further described.
  • the traffic flow in the upper part of FIG. 7 is a forward traffic flow
  • the TLP on the left side of the network adds a delay measurement flag 1 to a data packet
  • the recording time is t1
  • the forward traffic flow pair should be sent.
  • the end time stamp information is tl
  • the TLP on the right side identifies the data packet with the delay measurement flag 1
  • the time t2 is recorded, that is, the time-stamp information of the receiving end of the forward traffic flow pair is t2
  • the left-side TLP and the right-side TLP respectively generate the transmission delay measurement information and the receiving end delay measurement information, and are managed by the left.
  • the DCP of the side TLP (for the upstream TLP of the forward traffic flow is the upstream TLP) and the DCP managing the right TLP read the transmission delay measurement information and the receiver delay measurement information, respectively, and send them to the MCP.
  • the MCP receives the sender timestamp information t3 and the receiver timestamp information t4.
  • the delay of the forward traffic flow data packet by the MCP is t2-tl
  • the delay of the reverse traffic flow data packet is t4-t3
  • the two-way delay is (t2-tl)+(t4-t3), that is, (t4- Tl ) -(t3-t2).
  • the network delay measurement method provided in this embodiment first performs time synchronization with each DCP by using an external time synchronization tool by the MCP, and then receives the delay measurement information sent by each DCP corresponding to the upstream TLP and the downstream TLP through the MCP, and the delay measurement is performed.
  • the information is the delay measurement information of the delay measurement information and the delay measurement information of the receiving end, and the measurement information of the delay of the TLP at the upstream end of the MCP record and maintenance target data flow summary table and the reception of the downstream receiving end TLP
  • the end delay measurement information is further calculated by the MCP based on the target traffic flow measurement data summary table in each measurement period identifier, and the delay measurement information of the transmission end delay measurement information and the delay information of the receiving end are calculated, thereby directly and accurately Determine the traffic delay of the network.
  • the method according to the thirteenth embodiment of the network delay measurement method of the present invention further includes:
  • the MCP receives the transmission delay measurement information sent by the DCP corresponding to the upstream TLP, and the MCP receives the received delay measurement information of the determined measurement and the transmission delay measurement information of the DCP corresponding to the downstream TLP, and the MCP is based on the sending end.
  • the delay measurement information and the receiver delay measurement information determine the network single delay.
  • the transmission delay measurement information and the receiver delay measurement information are guaranteed to be in the same measurement period, and the delay measurement information and the receiver delay measurement are ensured at the transmitting end.
  • the information belongs to the same data packet.
  • the uplink TLP, the downstream TLP, and the DCP managing the downstream TLP are used to measure the delay information of the sender and the receiver.
  • the measurement information is matched and the DCP matches the sender delay measurement information and the receiver delay measurement information to the MCP.
  • the MCP also maintains the table 1.
  • the MCP matches the transmission delay measurement information according to the matching.
  • Receiver delay measurement information refer to Table 1, update the sender delay measurement information and the receiver delay measurement information to the data table entry of the corresponding data packet of Table 1, need to say
  • the DCP can send the matched delay measurement information and the receiver delay measurement information to the MCP in the MCP maintained by the MCP in this embodiment.
  • the MCP sends the delay measurement information to the receiver delay.
  • the measurement information is updated to the data entry of the corresponding data packet, and the delay condition is determined by the MCP.
  • the DCP can also send the determined delay to the MCP according to the consistent delay of the sender delay measurement information and the receiver delay measurement information, and the MCP directly receives the delay.
  • the data entry corresponding to the data packet is determined by matching the consistent sender delay measurement information and the receiver delay measurement information, and then calculating the delay by the MCP, and the specific calculation method and formula are in the network delay measurement method embodiment of the present invention. The detailed description has been made in the 12th, and will not be repeated here.
  • the network delay measurement method receives the delay measurement information of the transmission end sent by the DCP corresponding to the upstream TLP through the MCP, and the determined measurement information of the delay and the transmission delay sent by the DCP corresponding to the downstream TLP belongs to the same measurement period.
  • the receiver delay measurement information, the MCP based on the transmitter delay measurement information and the receiver delay measurement information, ensures that the MCP accurately measures the network traffic flow delay without the time synchronization tool.
  • the method for the network delay measurement method according to the fourth embodiment of the present invention further includes:
  • the MCP receives the sender delay measurement information sent by the DCP corresponding to the upstream TLP, and the sender delay measurement information includes time stamp information, a service flow identifier, a TLP identifier, a sender service flow characteristic information, and a sender fragment reassembly identifier.
  • the MCP receives the receiver delay measurement information sent by the DCP corresponding to the downstream TLP, and the delay information of the receiving end includes time stamp information, a service flow identifier, a TLP identifier, a service flow characteristic information of the receiver, and a fragment reassembly identifier of the receiver.
  • the MCP determines whether the timestamp information of the sender and the timestamp information of the receiver correspond to the same service flow according to the feature information of the service flow of the sender, the fragmentation reassembly identifier of the sender, the service stream feature information of the receiver, and the fragmentation reassembly identifier of the receiver. Stamp information. If yes, the MCP determines the network single delay based on the sender delay measurement information and the receiver delay measurement information.
  • the method of the fifth embodiment of the network delay measurement method and the method for the network delay measurement method in the tenth embodiment of the present invention may be based on the feature information of the service flow of the sender, the reassembly identifier of the sender, and the service flow of the receiver.
  • the information and the fragmentation reassembly identifier of the receiving end determine whether the timestamp information of the transmitting end and the timestamp information of the receiving end belong to the timestamp information of a data packet.
  • the MCP will The sender timestamp information and the receiver timestamp information are updated to the data entry of the corresponding data packet in Table 1, and if not, the MCP updates the sender timestamp information and the receiver timestamp information to their respective data respectively. In the data entry of the package.
  • the MCP is based on the characteristic information of the service flow of the sender, that is, the quintuple: source IP address, destination The IP address, the protocol type, the source protocol port number, and the destination protocol port number determine which service flow the delay measurement information of the sender belongs to.
  • the MCP After determining that the delay measurement information of the sender belongs to a target service flow, the MCP finds and The target traffic flow measurement data summary table corresponding to the service flow, and because the sender fragment reassembly identifier is unique to a data packet, the MCP determines, according to the sender fragment reassembly identifier, that the sender delay measurement information belongs to the target.
  • a specific data packet of the service flow for example, belongs to the Nth data packet, and the MCP updates the sender time stamp information carried by the sender delay measurement information to the data entry of the corresponding Nth data packet.
  • the MCP also uses the above operation to identify the receiver delay measurement information, and updates the receiver delay measurement information to the corresponding entry of the corresponding data packet, and then the corresponding MCP performs corresponding
  • the delay calculation, the specific calculation method and the formula are described in detail in the twelfth embodiment of the network delay measurement method of the present invention, and details are not described herein again.
  • the network delay measurement method provided in this embodiment receives the delay measurement information of the sender sent by the DCP corresponding to the upstream TLP through the MCP, and the delay measurement information of the sender includes the timestamp information, the service flow identifier, the TLP identifier, and the sender service flow.
  • the feature information and the fragment reassembly identifier of the sender end the MCP receives the receiver delay measurement information sent by the DCP corresponding to the downstream TLP, and the delay measurement information of the receiver includes the timestamp information, the service flow identifier, the TLP identifier, and the service flow characteristic information of the receiver.
  • the receiving end fragment reassembly identifier, and the MCP identifies and matches according to the sending end service flow characteristic information, the sending end fragment reassembly identifier, the receiving end service flow characteristic information, and the receiving end fragment recombination identifier, thereby ensuring the sending end delay
  • the measurement information and the measurement information of the delay of the receiving end belong to the same data packet, and the delay calculation of the matched transmission delay measurement information and the reception delay measurement information is performed by the MCP, thereby realizing direct and accurate measurement of the service flow delay.
  • FIG. 8 is a schematic structural diagram of Embodiment 1 of a DCP according to the present invention. As shown in FIG. 8, the DCP includes: an obtaining module 10 and a sending module 12.
  • the obtaining module 10 is configured to obtain time delay measurement information obtained by measuring at least one TLP, where the time delay measurement information includes: time stamp information, a service flow identifier, and a TLP identifier.
  • the sending module 12 is configured to send the delay measurement information to the measurement control point MCP, so that the MCP determines the network delay according to the timestamp information, the service flow identifier, and the TLP identifier.
  • the DCP is deployed on the sending end device and the receiving end device.
  • the working principle and the technical solution are described in detail in Embodiment 1 of the network delay measuring method of the present invention, and details are not described herein again.
  • the DCP of this embodiment may be used to implement the technical solution of the method embodiment shown in FIG. 1.
  • the principle and the technical effect are similar, and details are not described herein again.
  • the obtaining module 10 is configured to obtain the transmission delay measurement information obtained by measuring the service flow sent by the at least one upstream TLP, and the sending module 12 is specific to the DCP that manages the upstream TLP.
  • the sending end delay measurement information is sent to the MCP, and the sending end delay measurement information includes the sending end time stamp information, the service flow identifier, and the TLP identifier.
  • the DCP is a DCP that manages the downstream TLP
  • the acquiring module 10 is configured to obtain the receiving end delay measurement information obtained by measuring the received service flow by the at least one downstream TLP
  • the sending module 12 is specifically configured to delay the receiving end.
  • the measurement information is sent to the MCP, and the receiving end delay measurement information includes the receiving end time stamp information, the service flow identifier, and the TLP identifier.
  • FIG. 9 is a schematic structural diagram of a second embodiment of a DCP according to the present invention.
  • the acquisition module 20 of the DCP embodiment 2 of the present invention includes: a first acquisition unit 200, and a periodic identifier acquisition. Unit 202.
  • the first obtaining unit 200 is configured to obtain, by the at least one upstream TLP, the transmit end delay measurement information obtained by measuring the sent service flow, or obtain the receive end delay measurement by the at least one downstream TLP to measure the sent service flow. information.
  • the period identifier obtaining unit 202 is configured to: when the DCP managing the upstream TLP ends the measurement period, acquire the measurement period identifier, and send the measurement period identifier to the MCP, or, when the DCP managing the downstream TLP starts at the measurement period, the period identifier
  • the obtaining unit 202 can obtain the boundary time point of each measurement period according to the NTP or the IEEE 1588v2 clock, that is, obtain the start time of the measurement period, and if the difference between the start time and the time stamp information is less than or equal to the preset duration, the receiving end
  • the delay measurement information belongs to the measurement information corresponding to the measurement period identifier.
  • the measurement period identifier is incremented by one, and the time stamp information belongs to the next measurement period, and the management downstream TLP is obtained.
  • the DCP is identified during the measurement period of the measurement period.
  • the preset duration is 2/3 of the measurement period.
  • the corresponding DCP can directly read the sender delay measurement information and the receiver delay measurement information generated by the upstream TLP and the downstream TLP, and the period identifier acquisition unit 202 of the DCP is configured according to the delay information of the sender and the receiver.
  • the delay measurement information obtains two measurement period identifiers respectively, and for one data packet to which the delay measurement flag is added in one cycle, the DCP for managing the upstream TLP and the two measurement period identifiers for managing the DCP of the downstream TLP are Consistent.
  • the DCP embodiment 2 sending module 22 of the present invention includes: a first sending unit 220 and a second sending unit 222.
  • the first sending unit 220 sends the sending end delay measurement information to the MCP, or sends the receiving end delay measurement information to the MCP.
  • the second sending unit 222 sends the measurement period identifier acquired by the period identifier acquiring unit 202 to the MCP at the end of the measurement period, or the measurement period acquired by the period identifier acquiring unit 202 of the DCP managing the downstream TLP.
  • the identity is sent to the MCP.
  • the DCP of the DCP embodiment of the present invention further includes: a time synchronization module 24.
  • the time synchronization module 24 is configured to perform time synchronization with the TLP by using the NTP or IEEE 1588v2 clock, and use NTP or IEEE 1588v2 clock management before the acquisition module 20 acquires the delay measurement information obtained by measuring the traffic flow by the at least one TLP.
  • the DCP of the upstream TLP is time synchronized with the DCP managing the downstream TLP.
  • the third embodiment of the network delay measurement method of the present invention has been described in detail, and details are not described herein again.
  • the DCP of the present embodiment can be used to implement the technical solution of the third embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 10 is a schematic structural diagram of Embodiment 3 of the DCP of the present invention.
  • the method includes: an obtaining module 30, a determining module 32, and a sending module 34.
  • the obtaining module 30 includes: a second obtaining unit 300 and a measuring message acquiring unit 302.
  • the second obtaining unit 300 is configured to obtain, by the at least one upstream TLP, the sender delay measurement information obtained by measuring the sent service flow, or obtain the receiver delay measurement obtained by measuring, by the at least one downstream TLP, the sent service flow.
  • the sender delay measurement information includes: a sender time stamp information, a service flow identifier, and
  • the TLP identifier, the receiver delay measurement information includes: the receiving end time stamp information, the service flow identifier, and the TLP identifier.
  • the measurement packet obtaining unit 302 is configured to obtain, by the at least one downstream TLP, the measurement packet sent by the upstream TLP and the arrival timestamp information of a measurement packet generated when the measurement packet reaches the downstream TLP, where the measurement message includes: End time stamp information.
  • the determining module 32 is specifically configured to determine whether the arrival time stamp information and the receiving time stamp information belong to the preset duration range, and if yes, determine that the sender time stamp information and the receiving end time stamp information belong to the same data packet.
  • the specific determination principle and method are described in detail in Embodiment 4 of the network delay measurement method of the present invention, and are not mentioned here.
  • the sending module 34 is specifically configured to send the result to the MCP.
  • the determining module 32 may send the sending end time stamp information and the receiving end time stamp information belonging to the same data packet to the sending module 34, and send the same to the MCP through the sending module 34, and perform delay calculation by the MCP.
  • the determining module 32 may directly determine the delay of the periodic data packet according to the sender time stamp information and the receiving time stamp information belonging to the same measurement period, and then send the calculated delay time to the MCP.
  • the DCP of the present embodiment can be used to implement the technical solution of the fourth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • the DCP of the DCP embodiment of the present invention includes: an obtaining module 10 and a sending module 12.
  • the obtaining module 10 when the DCP is a data collection point for managing the upstream TLP, the acquiring module 10 is specifically configured to obtain the delay information of the sending end, and the information about the delay of the sending end includes: the feature information of the sending end service stream and the sending end Fragment reorganization logo.
  • the acquiring module 10 is specifically configured to obtain the receiver delay measurement information, and the receiving end delay measurement information further includes: the receiving end service flow characteristic information and the receiving end fragment recombination identifier.
  • the sending module 12 is configured to: when the DCP is a data collection point of the upstream TLP, the sending module 12 is configured to send the sending end delay measurement information to the MCP, so that the MCP reassembles according to the feature information of the sending end service stream and the sending end.
  • the identifier, the service flow characteristic information of the receiving end, and the fragment reassembly identifier of the receiving end determine that the time stamp information of the sending end and the time stamp information of the receiving end are time stamp information corresponding to the same service flow.
  • the DCP of the present embodiment can be used to implement the technical solution of the fifth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 11 is a schematic structural diagram of Embodiment 1 of the TLP of the present invention. As shown in FIG. 11, the TLP includes: an identification module 40, a time stamp acquisition module 42, and a determination module 44.
  • the identifying module 40 is configured to identify the service flow according to the service flow feature information, and determine whether the service flow is the target service flow.
  • the method for determining the service flow according to the service flow characteristic information is described in detail in the sixth embodiment of the network delay measurement method of the present invention, and details are not described herein again.
  • the timestamp obtaining module 42 is configured to: if yes, add a delay measurement flag to the data packet of the service flow, and obtain the data packet delay measurement information corresponding to the delay measurement flag.
  • the time stamp obtaining module 42 adds a delay measurement flag to the data packet of the target service flow, and generates delay measurement information, where the time delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier, where the time stamp Information is the point in time at which the TLP adds a time delay measurement flag.
  • the network delay measurement method embodiment 1 has been described in detail, and will not be described here.
  • the determining module 44 is configured to determine delay measurement information, where the delay measurement information includes: time stamp information, a service flow identifier, and a TLP identifier, so that the DCP sends the delay measurement information to the MCP after acquiring the delay measurement information.
  • the TLP of the present embodiment can be used to implement the technical solution of the sixth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • the time stamp acquisition module 42 adds a delay measurement flag to the data packet of the service flow, and obtains the packet delay measurement information corresponding to the delay measurement flag, including:
  • the timestamp obtaining module 42 of the upstream TLP is specifically configured to add a delay measurement flag to the data packet of the target service flow, where the timestamp obtaining module obtains the packet sending end delay measurement information corresponding to the delay measurement flag, and sends the
  • the end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier, so that the DCP that manages the upstream TLP obtains the sender delay measurement information and then sends the sender delay measurement information to the MCP.
  • the time stamp obtaining module 42 of the downstream TLP is specifically configured to: when the identifying module identifies the data packet with the added delay measurement flag, the time stamp obtaining module acquires the data packet receiving end delay measurement information corresponding to the delay measurement flag .
  • the receiving end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier, so that the DCP that manages the downstream TLP obtains the receiver delay measurement information and then sends the receiver delay measurement information to the MCP.
  • FIG. 12 is a schematic structural diagram of Embodiment 2 of the TLP according to the present invention. As shown in FIG. 12, the method further includes: a time synchronization module 41 and a measurement period identifier acquisition module 43.
  • the time synchronization module 41 when the TLP is an upstream TLP, specifically for the timestamp acquisition module 42 of the upstream TLP, before adding the delay measurement flag to the data packet of the target service flow, using the NTP or IEEE 1588v2 clock and managing the upstream TLP DCP performs time synchronization.
  • the time synchronization module 41 is specifically configured to use the NTP or IEEE 1588v2 clock to manage the DCP of the downstream TLP before the identification module 40 of the downstream TLP identifies the data packet with the added delay measurement flag. Synchronize.
  • the measurement period identifier obtaining module 43 is configured to obtain the measurement period identifier corresponding to the delay measurement flag, so that the DCP that manages the upstream TLP acquires the measurement period identifier and sends the measurement period identifier information to the MCP;
  • the TLP is a downstream TLP, and is specifically configured to obtain a measurement period identifier corresponding to the delay measurement flag and a start time of each measurement period, so that the DCP that manages the downstream TLP acquires the start time and the measurement period identifier, and then performs matching determination, and then The receiver delay measurement information is sent to the MCP.
  • the DCP can directly read the delay measurement information generated by the TLP, and the DCP obtains the corresponding measurement period identifier by using the delay measurement information.
  • the upstream TLP or the downstream TLP may not use the measurement period identifier acquisition module. 43.
  • the TLP identifies the service flow data packet based on each period, and adds a delay measurement flag. Optionally, only one data packet is operated in each measurement period. .
  • the TLP of the present embodiment can be used to implement the technical solution of the eighth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 13 is a schematic structural diagram of Embodiment 3 of the TLP of the present invention. As shown in FIG. 13, the method further includes: a transmitting module 46 and a receiving module 48.
  • the sending module 46 is specifically configured to send, by the sending module of the upstream TLP, the measurement packet to the downstream TLP.
  • the measurement message includes: a sender time stamp information.
  • the time stamp acquisition module 42 adds a delay measurement flag to the data packet, and acquires data corresponding to the delay measurement flag. Packet delay measurement information, at this time, the upstream TLP starts the sending module 46, and the transmitting module 46 sends a measurement message to the downstream TLP.
  • the receiving module 48 is configured to receive the measurement packet by the receiving module of the downstream TLP, generate an arrival time stamp information of the measurement packet, and send the measurement packet and the arrival time stamp information to the DCP managing the downstream TLP, so as to enable the DCP. Determining whether the arrival time stamp information and the receiving time stamp information belong to the preset duration range, and if yes, determining that the sender time stamp information and the receiving end time stamp information belong to the same data packet, and send the determination result to the MCP.
  • the identification module 40 of the downstream TLP identifies the data packet with the delay measurement flag and generates corresponding receiving time stamp information, and the receiving module 48 receives the measurement message, records the time point when the measurement message is received, and generates the arrival time.
  • the measurement information is sent to the determining module 44, so that the determining module 48 sends the measurement message, the receiver delay measurement information, and the arrival time stamp information to the DCP managing the downstream TLP.
  • the DCP of the downstream TLP is managed to perform corresponding operations.
  • the TLP identifies the service flow data packet, and adds a delay measurement flag to the data packet in the measurement period interval in each period TLP, optionally, in each During the measurement period, the TLP adds a delay measurement flag to only one packet.
  • the TLP of the present embodiment can be used to implement the technical solution of the ninth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • the TLP embodiment of the present invention further includes:
  • the obtaining, by the timestamp obtaining module 42 of the upstream TLP, the information about the delay of the transmitting end of the transmitting end includes: the information about the service flow of the sending end and the fragmentation and reassembly of the transmitting end; the time delay obtaining module 42 of the downstream TLP acquiring the information of the receiving end delay measurement further includes: The service flow characteristic information of the receiving end and the fragment reassembly identifier of the receiving end are configured to enable the DCP that manages the upstream TLP to obtain the measurement information of the delay of the transmitting end and send it to the MCP, and the DCP that manages the downstream TLP acquires the measurement information of the delay of the receiving end and sends the information to the MCP.
  • the MCP determines that the timestamp information of the sending end and the timestamp information of the receiving end correspond to the same service flow according to the feature information of the service flow of the sending end, the fragmentation reassembly identifier of the sending end, the characteristic information of the receiving end service stream, and the fragmentation reassembly identifier of the receiving end.
  • Time stamp information The TLP of the present embodiment can be used to implement the technical solution of the tenth embodiment of the network delay measurement method of the present invention. The implementation principle and technical effects are similar, and details are not described herein again.
  • the time stamp obtaining module of the upstream TLP adds a delay measurement flag to the data packet of the target service flow, including: the IP header of the time stamp obtaining module in the data packet
  • a delay measurement flag is added to the reserved bits of the TOS or the reserved bits of the Flags, so that the delay flag of the TOS or the reserved bits of the Flags can be used to add a delay measurement flag to the data packet, and the normal transmission of the data packet can be ensured.
  • the identification module 60 is specifically configured to identify the service flow according to at least two pieces of information in the quintuple to ensure effective identification of the service flow.
  • the reserved bits of the TOS or the reserved bits of the Flags, and the quintuple are described in detail in the seventh embodiment of the network delay measurement method of the present invention, and details are not described herein again.
  • FIG. 14 is a schematic structural diagram of Embodiment 1 of an MCP according to the present invention.
  • the MCP includes: a receiving module 70 and a determining module 72.
  • the receiving module 70 is configured to receive the sending end delay measurement information of the DCP corresponding to the upstream TLP and the receiving end delay measurement information of the DCP corresponding to the downstream TLP, where the sending end delay measurement information includes the time stamp information and the service flow identifier. And the TLP identifier, the receiving end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier;
  • the determining module 72 is configured to determine a single delay of the network according to the delay information of the sending end and the measurement information of the delay of the receiving end.
  • the MCP is deployed on any network element node device of the network, and is optionally deployed on a network element node device with strong functions.
  • a target traffic flow measurement data summary table is maintained by the determination module 72, as described in Table 1 above. The calculation principle and formula of the determination module 72 are described in detail in the embodiment 12 of the network delay measurement method of the present invention, and details are not described herein again.
  • the MCP of this embodiment may be used to implement the technical solution of the network delay measurement method in the eleventh embodiment of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 15 is a schematic structural diagram of Embodiment 2 of the MCP according to the present invention, and includes a receiving module 70 and a determining module 72.
  • the receiving module 70 includes: a first receiving unit 700 and a second receiving unit 702. .
  • the determining module 72 includes a first matching unit 720 and a first determining unit 722.
  • the first receiving unit 700 is configured to receive, according to the sending end delay measurement information of the DCP corresponding to the upstream TLP, and the receiving end delay measurement information of the DCP corresponding to the downstream TLP, and send the
  • the end delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier
  • the receiver delay measurement information includes time stamp information, a service flow identifier, and a TLP identifier.
  • the second receiving unit 702 is specifically configured to receive a measurement period identifier that is sent by the DCP that manages the upstream TLP, and receive a measurement period identifier that is sent by the DCP that manages the downstream TLP.
  • the first matching unit 720 is specifically configured to determine, according to the measurement period identifier sent by the DCP that manages the upstream TLP, and the measurement period identifier that is sent by the DCP of the downstream TLP, whether the measurement information of the delay information of the transmitting end and the measurement information of the delay of the receiving end belong to the same measurement. cycle.
  • the first matching unit 720 After receiving the measurement period identifier received by the second receiving unit 702, the first matching unit 720, according to the measurement period identifier sent by the DCP managing the upstream TLP, and the measurement period identifier sent by the DCP managing the downstream TLP, will belong to the same measurement.
  • the periodic transmit delay measurement information and the receive delay measurement information are updated in the corresponding entries of the target service flow measurement data summary table maintained by the MCP.
  • the first determining unit 722 is specifically configured to: if yes, determine a network single delay according to the sending end delay measurement information and the receiving end delay measurement information.
  • the first determining unit 722 detects that one of the forward traffic flow identifiers of one measurement cycle identifier in the table 1 or the data arrival data of the data traffic corresponding to the reverse traffic flow is set to "all aligned"
  • the first determining unit 722 performs delay calculation according to the corresponding sender time stamp information and the receiving end time stamp information.
  • the specific calculation principle and formula are described in detail in the embodiment of the network delay measurement method of the present invention, and are not described herein again.
  • the MCP of this embodiment may be used to implement the technical solution of the network delay measurement method in the twelfth embodiment of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • the MCP of the third embodiment of the MCP of the present invention includes: a receiving module 70 and a determining module 72.
  • the receiving module 70 is configured to receive, by the receiving end, the delay of the transmitting end of the DCP corresponding to the upstream TLP, and measure the delay information of the receiving end of the same data packet.
  • the determining module 72 is specifically configured to determine a single delay of the network according to the delay information of the sending end and the measurement information of the delay of the receiving end.
  • the receiving module 70 of the MCP receives and updates the transmission delay measurement information and the receiver delay measurement information. Go to the corresponding data packet entry of the data packet, and then pass the determination module. 72 Determine the delay situation.
  • the DCP can also send the determined delay to the MCP by using the DCP to determine the delay according to the matched delay measurement information and the receiver delay measurement information.
  • the determination module 72 is not required to be activated.
  • the MCP of this embodiment may be used to implement the technical solution of the thirteenth embodiment of the network delay measurement method of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 16 is a schematic structural diagram of Embodiment 4 of the MCP of the present invention. As shown in FIG. 16, the method includes: a receiving module 70 and a determining module 72.
  • the receiving module 70 is specifically configured to receive the sending end delay measurement information sent by the DCP corresponding to the upstream TLP, where the sending end delay measurement information includes the time stamp information, the service flow identifier, the TLP identifier, the sending end service flow characteristic information, and the sending end
  • the determining module 72 includes: a second matching unit 721 and a second determining unit 723.
  • the second matching unit 721 is specifically configured to determine the timestamp information of the sending end and the timestamp information of the receiving end according to the feature information of the service flow of the sending end, the fragment reassembly identifier of the sending end, the characteristic information of the receiving end service stream, and the reassembly identifier of the receiving end fragment. Whether it corresponds to the time stamp information of the same data packet.
  • the specific working principle and method are described in detail in Embodiment 5, Embodiment 10 and Embodiment 14 of the network delay measurement method of the present invention, and details are not described herein again.
  • the second determining unit 723 is specifically configured to, if yes, determine a single delay of the network according to the measurement information of the delay information of the transmitting end and the measurement information of the receiving end delay.
  • the specific working principle and method are described in detail in Embodiment 5, Embodiment 10 and Embodiment 14 of the network delay measurement method of the present invention, and details are not described herein again.
  • the MCP of this embodiment may be used to implement the technical solution of the network delay measurement method in the embodiment of the present invention.
  • the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 17 is a schematic structural diagram of Embodiment 1 of a network delay measurement system according to the present invention.
  • the system in this embodiment includes: DCP, TLP, and MCP.
  • TLP is performed on any two sides of the network.
  • the deployment of the DCP and the corresponding DCP are not limited in the embodiment of the present invention.
  • the DCP can use the structure of FIG. 8 to FIG. 10, and can perform the technical solutions of the network delay measurement method of the first embodiment to the fifth embodiment of the present invention; the TLP can use the structure of FIG. 11 to FIG.
  • the technical solution of the sixth embodiment to the tenth embodiment of the network delay measurement method of the present invention can be implemented;
  • the MCP can use the structure of FIG. 14 to FIG. 16 correspondingly, and can implement the delay measurement method provided by the present invention.
  • a technical solution of the fourteenth embodiment which realizes the original The technical and technical effects are similar and will not be described here.
  • FIG. 18 is a schematic diagram of Embodiment 2 of a network delay measurement system according to the present invention. Referring to FIG. 17 and FIG. 18, the network delay measurement method, apparatus, and system provided by the present invention are generally described below.
  • a TLP is deployed on the network side of the upstream sending end and the downstream receiving end.
  • the TLP may be deployed on the user side, and the DCP is deployed on the upstream sending end and the downstream receiving end respectively.
  • the base station side gateway CSG CSG
  • the radio network controller site gateway RSG
  • RSG2 is a downstream receiving device.
  • the MCP is deployed on any network element node in the entire network.
  • MCP is deployed on RSG1.
  • the MCP is deployed on a relatively powerful network element node, and the transmission path of the delay measurement information is separated from the transmission path of the target service flow (shown by a solid line in FIG. 18). , to ensure the independence of optional delay measurement information reading and transmission.
  • the management network (as shown in Figure 18, the path of the management network layer is indicated by a dotted line) can be used to measure VPN, DCN or public network with IP reach.
  • each TLP, each DCP, and the MCP cooperate to perform direct delay measurement on the service flow.
  • the specific method and technical solution are in the network delay measurement method of the present invention.
  • Embodiment 1 to Embodiment 14 The DCP embodiment 1 to 4, the TLP embodiment 1 to the embodiment 4 of the present invention, and the MCP embodiment 1 to the fourth embodiment of the present invention have been described in detail, and are not described herein again.
  • the acquisition and transmission of the delay measurement information between each TLP, each DCP, and the MCP in the embodiment of the present invention is implemented by the management network to implement the external transmission, which is effective.
  • the delay measurement information is caused by the difference between the measurement reference of the Layer 2 VPN network and the Layer 3 VPN network.
  • the delay measurement information provided in the embodiment of the present invention includes the timestamp information and the service flow.
  • the identifier and the TLP identifier may be used by the TLP on the new receiving device to perform delay measurement on the data packet of the target service flow when the RSG1 and the RSG2 are switched.
  • the target service flow starts from the left user and is identified by the TLP on the CSG, enabling network delay measurement, and then the target service flows through the network to the TLP of the RSG1, and the TLP performs the corresponding receiving end. Delay measurement.
  • RSG1 fails, the target traffic flow is switched to RSG2, at which point the TLP on RSG2 can identify the target traffic flow and continue the corresponding delay measurement.
  • the aforementioned program can be stored in a computer readable storage medium.
  • the S includes the above-described method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Health & Medical Sciences (AREA)
  • Cardiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

一种网络时延测量方法、装置和系统，包括：获取至少一个目标逻辑端口TLP对业务流进行测量得到的时延测量信息，时延测量信息包括：时戳信息、业务流标识以及TLP标识，将时延测量信息发送给测量控制点MCP，以使MCP根据时戳信息、业务流标识以及TLP标识，确定网络时延情况。本发明实施例还提供一种网络时延测量装置和系统。本发明实施例实现了在网络为单点对单点、或者单点对多点等场景下直接对业务流实现准确的时延测量，反应业务流真实的时延情况。

Description

网络时延测量方法、 装置和系统 技术领域 本发明涉及通信技术， 尤其涉及一种网络时延测量方法、 装置和系统。 背景技术 随着网络信息技术的不断进步， 网络的 IP ( Internet Protocol, 网络协议 ) 化已经成为趋势。 在这种趋势之下， 如何对基于 IP协议的业务进行时延性能 质量评价已经成为越发突出的问题。
现有技术对于网络业务流时延的测量， 主要是在测量一端插入专门的时 延测量报文， 在时延测量报文中携带在发送端和接收端的时戳， 再根据时延 测量报文中的收发时戳， 来计算网络业务流的时延结果。
但是， 由于现有技术是间接地测量时延测量报文， 并不能够真实准确的 反应网络业务流本身的时延性能。 发明内容 本发明实施例提供一种网络时延测量方法、 装置和系统， 以实现对于网 络业务流时延的测量。
一方面， 本发明实施例提供一种网络时延测量方法， 其中， 包括： 获取至少一个目标逻辑端口 TLP对业务流进行测量得到的时延测量信 息， 所述时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识；
将所述时延测量信息发送给测量控制点 MCP , 以使所述 MCP根据所述时 戳信息、 业务流标识以及 TLP标识， 确定网络时延情况。
可选地， 上述网络时延测量方法， 可以包括： 所述获取至少一个 TLP对 业务流进行测量得到的时延测量信息， 包括：
管理上游 TLP的数据收集点 DCP获取至少一个上游 TLP对发送的业务流 进行测量得到的发送端时延测量信息；
管理下游 TLP的 DCP获取至少一个下游 TLP对接收的业务流进行测量得 到的接收端时延测量信息； 所述将所述时延测量信息发送给 MCP, 包括：
所述管理上游 TLP的 DCP将所述发送端时延测量信息发送给 MCP, 所述 发送端时延测量信息包括发送端时戳信息、 业务流标识以及 TLP标识；
所述管理下游 TLP的 DCP将所述接收端时延测量信息发送给 MCP , 所述 接收端时延测量信息包括接收端时戳信息、 业务流标识以及 TLP标识。
可选地， 上述网络时延测量方法， 可以包括：
所述管理上游 TLP的 DCP在测量周期结束时， 获取测量周期标识， 并将 所述测量周期标识发送给所述 MCP;
所述管理下游 TLP的 DCP获取该测量周期的起始时间， 若所述起始时间 与所述时戳信息之差小于等于预设时长， 则所述接收端时延测量信息属于所 述测量周期标识对应的测量信息； 若所述起始时间与所述时戳信息之差大于 预设时长， 则将测量周期标识加 1 , 所述时戳信息属于下一个测量周期， 将所 述测量周期标识发送给所述 MCP。
可选地， 上述网络时延测量方法， 其中， 所述预设时长为所述测量周期 的 2/3时长。
可选地， 上述网络时延测量方法， 可以包括：
所述管理上游 TLP的 DCP釆用 NTP或 IEEE 1588v2时钟与所述上游 TLP进 行时间同步， 所述管理下游 TLP的 DCP釆用 NTP或 IEEE 1588v2时钟与所述下
DCP也通过所述釆用 NTP或 IEEE 1588v2时钟来实现时间同步。
可选地， 上述网络时延测量方法， 可以包括：
所述管理下游 TLP的 DCP获取至少一个下游 TLP接收的由所述上游 TLP 发送的测量报文以及所述测量报文到达所述下游 TLP时产生的一个测量报文 的到达时戳信息， 所述测量报文包括： 发送端时戳信息；
所述管理下游 TLP的 DCP确定所述到达时戳信息和所述接收端时戳信息 是否同属于预设时长范围， 若是， 则确定所述发送端时戳信息和所述接收端 时戳信息属于同一数据包， 并将确定结果发送给所述 MCP。
可选地， 上述网络时延测量方法， 其中， 所述发送端时延测量信息还包 括： 发送端业务流特征信息以及发送端分片重组标识， 所述接收端时延测量 信息还包括：接收端业务流特征信息以及接收端分片重组标识，以使所述 MCP 根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息 以及接收端分片重组标识， 确定所述发送端时戳信息和所述接收端时戳信息 是对应于同一业务流的时戳信息。
另一方面， 本发明实施例提供一种网络时延测量方法， 其中， 包括： 根据业务流特征信息对业务流进行识别， 确定所述业务流是否是目标业 务流；
若是， 则对所述业务流的数据包添加时延测量标志， 获取与该时延测量 标志对应的数据包时延测量信息；
确定时延测量信息， 所述时延测量信息包括： 时戳信息、 业务流标识以 及 TLP标识， 以使所述 DCP在获取所述时延测量信息后将所述时延测量信息 发送给 MCP。
可选地， 上述网络时延测量方法， 其中， 对所述业务流的数据包添加时 延测量标志， 获取与该时延测量标志对应的数据包时延测量信息， 包括： 上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与该时延测 量标志对应的数据包发送端时延测量信息， 所述发送端时延测量信息包括时 戳信息、 业务流标识以及 TLP标识， 以使管理上游 TLP的 DCP获取所述发送端 时延测量信息后将所述发送端时延测量信息发送给 MCP;
下游 TLP在识别到所述添加时延测量标志的数据包时 , 获取与该时延测 量标志对应的数据包接收端时延测量信息。 所述接收端时延测量信息包括时 戳信息、 业务流标识以及 TLP标识， 以使管理下游 TLP的 DCP获取所述接收端 时延测量信息后将所述接收端时延测量信息发送给 MCP。
可选地， 上述网络时延测量方法， 其中， 还包括：
所述上游 TLP在目标业务流的数据包上添加时延测量标志之前，釆用 NTP 在识别到所述添加时延测量标志的数据包之前， 釆用 NTP或 IEEE 1588v2时钟 与所述管理下游 TLP的 DCP进行时间同步；
对所述业务流的数据包添加时延测量标志， 获取与该时延测量标志对应 的数据包时延测量信息， 还包括：
所述上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与所述 时延测量标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取所述测量 周期标识后将所述测量周期标识信息发送给 MCP;
所述下游 TLP在每个测量周期获取该测量周期起始时间， 在每个测量周 期内识别到所述添加时延测量标志的数据包时， 获取与所述时延测量标志对 应的测量周期标识， 以使管理下游 TLP的 DCP获取所述起始时间和所述测量 周期标识后将所述测量周期标识信息发送给 MCP。
可选地， 上述网络时延测量方法， 其中， 还包括：
由所述上游 TLP向下游 TLP发送测量报文， 所述测量报文包括： 发送端时 戳信息；
所述下游 TLP的接收模块接收到所述测量报文， 产生一个测量报文的到 达时戳信息， 并将所述测量报文以及所述到达时戳信息发送给所述管理下游 TLP的 DCP,以使所述 DCP确定所述达时戳信息和所述接收端时戳信息是否 同属于预设时长范围， 若是， 则确定所述发送端时戳信息和所述接收端时戳 信息属于同一数据包， 并将确定结果发送给所述 MCP。
可选地， 上述网络时延测量方法， 其中， 所述发送端时延测量信息还包 括： 发送端业务流特征信息以及发送端分片重组标识； 所述接收端时延测量 信息还包括： 接收端业务流特征信息以及接收端分片重组标识；
以使所述管理上游 TLP的 DCP获取所述发送端时延测量信息并发送给所 述 MCP , 所述管理下游 TLP的 DCP获取所述接收端时延测量信息并发送给所 述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及接收端分片重组标识， 确定所述发送端时戳信息 和所述接收端时戳信息是对应于同一数据包的时戳信息。
可选地， 上述网络时延测量方法， 其中， 所述上游 TLP在目标业务流的 数据包上添加时延测量标志， 包括：
在所述数据包的 IP头中 TOS的保留位或者 Flags的保留位上添加时延测量 标志。
可选地， 上述网络时延测量方法， 其中， 所述根据业务流特征信息对业 务流进行识别， 包括：
根据五元组中的至少两元信息， 对所述业务流进行识别。
再一方面， 本发明实施例提供一种网络时延测量方法， 其中， 包括： 接收上游 TLP对应的 DCP发送的发送端时延测量信息以及与下游 TLP对 应的 DCP发送的接收端时延测量信息， 所述发送端时延测量信息包括时戳信 息、 业务流标识以及 TLP标识， 所述接收端时延测量信息包括时戳信息、 业 务流标识以及 TLP标识；
根据所述发送端时延测量信息和所述接收端时延测量信息， 确定网络单 次时延情况。
可选地， 上述网络时延测量方法， 其中， 还包括：
MCP接收所述管理上游 TLP的 DCP发送的测量周期标识， 所述 MCP接收 所述管理下游 TLP的 DCP发送的测量周期标识， 由所述 MCP根据所述管理上 游 TLP的 DCP发送的测量周期标识和所述管理下游 TLP的 DCP发送的测量周 期标识， 判断所述发送端时延测量信息和所述接收端时延测量信息是否属于 同一测量周期， 若是， 则所述 MCP根据所述发送端时延测量信息和所述接收 端时延测量信息， 确定网络单次时延情况。
可选地， 上述网络时延测量方法， 其中， 还包括：
息属于同一数据包的所述接收端时延测量信息， 所述 MCP根据所述发送端时 延测量信息和所述接收端时延测量信息， 确定网络单次时延情况。
可选地， 上述网络时延测量方法， 其中， 还包括：
MCP接收所述上游 TLP对应的 DCP发送的发送端时延测量信息， 所述发 送端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 发送端业务流特 征信息以及发送端分片重组标识；
所述 MCP接收所述下游 TLP对应的 DCP发送的接收端时延测量信息， 所 述接收端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 接收端业务 流特征信息以及接收端分片重组标识；
所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业 务流特征信息以及接收端分片重组标识， 判断所述发送端时戳信息和所述接 收端时戳信息是否对应于同一数据包的时戳信息； 若是， 则所述 MCP根据所 述发送端时延测量信息和所述接收端时延测量信息，确定网络单次时延情况。
一方面， 本发明实施例提供一种 DCP, 其中， 包括：
获取模块， 用于获取至少一个 TLP对业务流进行测量得到的时延测量信 息， 所述时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识 发送模块， 用于将所述时延测量信息发送给测量控制点 MCP, 以使所述 MCP根据所述时戳信息、 业务流标识以及 TLP标识， 确定网络时延情况。
可选地， 上述 DCP, 其中， 所述 DCP为管理上游 TLP的 DCP;
所述获取模块， 具体用于获取至少一个上游 TLP对发送的业务流进行测 量得到的发送端时延测量信息；
或者，
所述 DCP为管理下游 TLP的 DCP；
所述获取模块， 具体用于获取至少一个下游 TLP对接收的业务流进行测 量得到的接收端时延测量信息；
所述 DCP将时延测量信息发送给 MCP, 包括：
所述 DCP为管理上游 TLP的 DCP；
所述发送模块， 具体用于将所述发送端时延测量信息发送给 MCP, 所述 发送端时延测量信息包括发送端时戳信息、 业务流标识以及 TLP标识；
或者，
所述 DCP为管理下游 TLP的 DCP；
所述发送模块， 具体用于将所述接收端时延测量信息发送给 MCP, 所述 接收端时延测量信息包括接收端时戳信息、 业务流标识以及 TLP标识。
可选地， 上述 DCP, 其中， 所述获取模块包括：
第一获取单元， 用于获取至少一个上游 TLP对发送的业务流进行测量得 到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务流进 行测量得到的接收端时延测量信息；
周期标识获取单元， 用于在所述管理上游 TLP的 DCP在测量周期结束时， 获取测量周期标识， 并将所述测量周期标识发送给所述 MCP, 或者， 在所述 管理下游 TLP的 DCP在测量周期开始时， 所述周期标识获取单元获取该测量 周期的起始时间， 若所述起始时间与所述时戳信息之差小于等于预设时长， 则所述接收端时延测量信息属于所述测量周期标识对应的测量信息； 若所述 起始时间与所述时戳信息之差大于预设时长， 则将测量周期标识加 1 , 所述时 戳信息属于下一个测量周期， 获取管理下游 TLP的 DCP在该测量周期的测量 周期标识； 所述发送模块包括：
第一发送单元， 将所述发送端时延测量信息发送给 MCP, 或者， 将所述 接收端时延测量信息发送给所述 MCP；
第二发送单元， 将所述管理上游 TLP的 DCP在测量周期结束时将周期标 识获取单元获取的所述测量周期标识发送给所述 MCP, 或者， 将所述管理下 游 TLP的 DCP的周期标识获取单元获取的所述测量周期标识发送给 MCP。
可选地， 上述 DCP, 其中， 所述预设时长为所述测量周期的 2/3时长。 可选地， 上述 DCP, 其中， 还包括：
时间同步模块， 用于在所述获取模块获取至少一个 TLP对业务流进行测 量得到的时延测量信息之前， 釆用 NTP或 IEEE 1588v2时钟与所述 TLP进行时 间同步， 以及釆用所述 NTP或 IEEE 1588v2时钟所述管理上游 TLP的 DCP与所 述管理下游 TLP的 DCP进行时间同步。
可选地， 上述 DCP, 其中， 所述获取模块包括：
第二获取单元， 用于获取至少一个上游 TLP对发送的业务流进行测量得 到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务流进 行测量得到的接收端时延测量信息；
测量报文获取单元， 用于获取至少一个下游 TLP接收的由所述上游 TLP 发送的测量报文以及所述测量报文到达所述下游 TLP时产生的一个测量报文 的到达时戳信息， 所述测量报文包括： 发送端时戳信息；
所述确定模块， 具体用于确定所述到达时戳信息和所述接收端时戳信息 是否同属于预设时长范围， 若是， 则确定所述发送端时戳信息和所述接收端 时戳信息属于同一数据包；
所述发送模块， 具体用于确定结果发送给所述 MCP。
可选地， 上述 DCP, 其中， 所述 DCP为管理上游 TLP的数据收集点； 所述获取模块， 具体用于获取所述发送端时延测量信息， 所述发送端时 延测量信息还包括： 发送端业务流特征信息以及发送端分片重组标识；
所述发送模块， 具体用于将所述发送端时延测量信息给所述 MCP;
所述 DCP为管理下游 TLP的数据收集点；
所述获取模块， 具体用于获取所述接收端时延测量信息， 所述接收端时 延测量信息还包括： 接收端业务流特征信息以及接收端分片重组标识； 所述发送模块， 具体用于将所述接收端时延测量信息发送给所述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业 务流特征信息以及接收端分片重组标识， 确定所述发送端时戳信息和所述接 收端时戳信息是对应于同一业务流的时戳信息。
另一方面， 本发明实施例提供一种 TLP, 其中， 包括：
识别模块， 用于根据业务流特征信息对业务流进行识别， 确定所述业务 流是否是目标业务流；
时戳获取模块， 用于若是， 则对所述业务流的数据包添加时延测量标志， 获取与该时延测量标志对应的数据包时延测量信息；
确定模块， 用于确定时延测量信息， 所述时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识， 以使所述 DCP在获取所述时延测量信息后将所述 时延测量信息发送给 MCP。
可选地， 上述 TLP, 其中， 所述时戳获取模块对所述业务流的数据包添 加时延测量标志， 获取与该时延测量标志对应的数据包时延测量信息， 包括： 上游 TLP的所述时戳获取模块， 具体用于在目标业务流的数据包上添加 时延测量标志， 获取与该时延测量标志对应的数据包发送端时延测量信息， 所述发送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 以使管 理上游 TLP的 DCP获取所述发送端时延测量信息后将所述发送端时延测量信 息发送给 MCP;
下游 TLP的所述时戳获取模块， 具体用于在所述识别模块识别到所述添 加时延测量标志的数据包时， 所述时戳获取模块获取与该时延测量标志对应 的数据包接收端时延测量信息； 所述接收端时延测量信息包括时戳信息、 业 务流标识以及 TLP标识 ,以使管理下游 TLP的 DCP获取所述接收端时延测量信 息后将所述接收端时延测量信息发送给 MCP。
可选地， 上述 TLP, 其中， 还包括：
所述 TLP为上游 TLP;
时间同步模块， 具体用于在所述上游 TLP的所述时戳获取模块在目标业 务流的数据包上添加时延测量标志之前， 釆用 NTP或 IEEE 1588v2时钟与所述 管理上游 TLP的 DCP进行时间同步；
所述 TLP为下游 TLP; 所述时间同步模块， 具体用于在所述下游 TLP的所述识别模块在识别到 所述添加时延测量标志的数据包之前， 釆用 NTP或 IEEE 1588v2时钟与所述管 理下游 TLP的 DCP进行时间同步；
所述 TLP, 还包括：
测量周期标识获取模块， 具体用于所述上游 TLP的所述测量周期标识获 取模块获取与所述时延测量标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取所述测量周期标识后将所述测量周期标识发送给 MCP; 所述下游 TLP的所述测量周期标识获取模块获取与所述时延测量标志对应的测量周期 标识以及每个测量周期起始时间， 以使管理下游 TLP的 DCP获取所述起始时 间和所述测量周期标识后进行匹配确定，再将所述测量周期标识发送给 MCP。
可选地， 上述 TLP, 其中， 还包括：
发送模块， 具体用于由所述上游 TLP的发送模块向下游 TLP发送测量报 文， 所述测量报文包括： 发送端时戳信息；
接收模块， 具体用于所述下游 TLP的接收模块接收到所述测量报文， 产 生一个测量报文的到达时戳信息， 并将所述测量报文以及所述到达时戳信息 发送给所述管理下游 TLP的 DCP,以使所述 DCP确定所述到达时戳信息和所述 接收端时戳信息是否同属于预设时长范围， 若是， 则确定所述发送端时戳信 息和所述接收端时戳信息属于同一数据包， 并将确定结果发送给所述 MCP。
可选地， 上述 TLP, 其中， 所述上游 TLP的所述时戳获取模块获取所述发 送端时延测量信息还包括：发送端业务流特征信息以及发送端分片重组标识； 所述下游 TLP的所述时戳获取模块获取所述接收端时延测量信息还包括： 接 收端业务流特征信息以及接收端分片重组标识；
以使所述管理上游 TLP的 DCP获取所述发送端时延测量信息并发送给所 述 MCP , 所述管理下游 TLP的 DCP获取所述接收端时延测量信息并发送给所 述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及接收端分片重组标识， 确定所述发送端时戳信息 和所述接收端时戳信息是对应于同一业务流的时戳信息。
可选地， 上述 TLP, 其中， 所述上游 TLP的所述时戳获取模块在目标业务 流的数据包上添加时延测量标志， 包括：
所述时戳获取模块在所述数据包的 IP头中 TOS的保留位或者 Flags的保留 位上添力 P时延测量标志。
可选地， 上述 TLP, 其中， 所述识别模块， 具体用于根据五元组中的至 少两元信息， 对业务流进行识别。
再一方面， 本发明实施例提供一种 MCP, 其中， 包括：
接收模块， 用于接收上游 TLP对应的 DCP发送的发送端时延测量信息以 及与下游 TLP对应的 DCP发送的接收端时延测量信息， 所述发送端时延测量 信息包括时戳信息、 业务流标识以及 TLP标识， 所述接收端时延测量信息包 括时戳信息、 业务流标识以及 TLP标识；
确定模块， 用于根据所述发送端时延测量信息和所述接收端时延测量信 息， 确定网络单次时延情况。
可选地， 上述 MCP, 其中， 还包括：
步； ' 一 … 所述接收模块， 包括：
第一接收单元， 具体用于接收上游 TLP对应的 DCP发送的发送端时延测 量信息以及与下游 TLP对应的 DCP发送的接收端时延测量信息， 所述发送端 时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 所述接收端时延测 量信息包括时戳信息、 业务流标识以及 TLP标识；
第二接收单元， 具体用于接收所述管理上游 TLP的 DCP发送的测量周期 标识， 以及接收所述管理下游 TLP的 DCP发送的测量周期标识；
所述确定模块， 还包括：
第一匹配单元， 具体用于根据所述管理上游 TLP的 DCP发送的测量周期 标识和所述管理下游 TLP的 DCP发送的测量周期标识， 判断所述发送端时延 测量信息和所述接收端时延测量信息是否属于同一测量周期；
确定单元， 具体用于， 若是， 则根据所述发送端时延测量信息和所述接 收端时延测量信息， 确定网络单次时延情况。
可选地， 上述 MCP, 其中， 所述接收模块， 具体用于接收所述上游 TLP 对应的 DCP发送的所述发送端时延测量信息， 以及接收所述下游 TLP对应的 DCP发送的已确定与所述发送端时延测量信息属于同一数据包的所述接收端 时延测量信息；
所述确定模块， 具体用于根据所述发送端时延测量信息和所述接收端时 延测量信息， 确定网络单次时延情况。
可选地， 上述 MCP, 其中， 所述接收模块， 具体用于接收所述上游 TLP 对应的 DCP发送的发送端时延测量信息， 所述发送端时延测量信息包括时戳 信息、 业务流标识、 TLP标识、 发送端业务流特征信息以及发送端分片重组 标识， 以及接收所述下游 TLP对应的 DCP发送的接收端时延测量信息， 所述 接收端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 接收端业务流 特征信息以及接收端分片重组标识；
所述确定模块包括：
第二匹配单元， 具体用于根据发送端业务流特征信息、 发送端分片重组 标识、 接收端业务流特征信息以及接收端分片重组标识， 判断所述发送端时 戳信息和所述接收端时戳信息是否对应于同一业务流的时戳信息；
确定单元， 具体用于， 若是， 则根据所述发送端时延测量信息和所述接 收端时延测量信息， 确定网络单次时延情况。
本发明实施例一种网络时延测量系统， 其中， 包括： 上述任一项所述的 DCP、 上述任一项所述的 TLP以及上述任一项所述的 MCP。
本发明实施例网络时延测量方法、 装置和系统， 通过 DCP获取至少一个 TLP对业务流进行直接测量得到的时延测量信息， 将该时延测量信息统一发 送给 MCP , 使 MCP根据所述该时延测量信息中的相关信息确定网络时延情 况， 实现了对业务流的直接时延测量。 附图说明 为了更清楚地说明本发明实施例或现有技术中的技术方案， 下面将对实 施例或现有技术描述中所需要使用的附图作简单地介绍， 显而易见地， 下面 描述中的附图仅仅是本发明的一些实施例， 对于本领域普通技术人员来讲， 在不付出创造性劳动性的前提下， 还可以根据这些附图获得其他的附图。
图 1为本发明网络时延测量方法实施例一的流程图；
图 2为本发明网络时延测量方法实施例三的实现流程示意图；
图 3为本发明网络时延测量方法实施例四的实现流程示意图；
图 4为本发明网络时延测量方法实施例五的实现流程示意图；
图 5为本发明网络时延测量方法实施例六的流程图；
图 6为本发明网络时延测量方法实施例十一的流程图； 图 7为本发明网络时延测量方法实施例十二的双向时延测量示意图； 图 8为本发明 DCP实施例一的结构示意图；
图 9为本发明 DCP实施例二的结构示意图；
图 10为本发明 DCP实施例三的结构示意图；
图 11为本发明 TLP实施例一的结构示意图；
图 12为本发明 TLP实施例二的结构示意图；
图 13为本发明 TLP实施例三的结构示意图；
图 14为本发明 MCP实施例一的结构示意图；
图 15为本发明 MCP实施例二的结构示意图；
图 16为本发明 MCP实施例四的结构示意图；
图 17为本发明网络时延测量系统实施例一的结构示意图；
图 18为本发明网络时延测量系统实施例二的示意图。 具体实施方式 为使本发明实施例的目的、 技术方案和优点更加清楚， 下面将结合本发 明实施例中的附图， 对本发明实施例中的技术方案进行清楚地描述， 显然， 所描述的实施例是本发明一部分实施例， 而不是全部的实施例。 基于本发明 中的实施例， 本领域普通技术人员在没有作出创造性劳动前提下所获得的所 有其他实施例， 都属于本发明保护的范围。
图 1为本发明网络时延测量方法实施例一的流程图， 如图 1所示， 本实 施例的方法可以包括：
S100, 获取至少一个目标逻辑端口 TLP对业务流进行测量得到的时延测 量信息。
具体的， 时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识。 数据釆集点 （ Data Collecting Point, 以下简称 DCP )获取至少一个目标 逻辑端口 （ Target Logical Port, 以下简称 TLP )对业务流进行测量得到的时 延测量信息。 该 TLP对应一个网络上游发送端入口或者一个网络下游接收端 出口。 对应该 TLP的 DCP分别部署在上游发送设备、 下游接收设备上， 用 于读取各 DCP所在设备上 TLP的时延测量信息。 也就是说， 当 TLP对应某 个网络上游发送端入口时， 对应该 TLP的 DCP部署在对应该网络上游发送 端入口的上游发送设备上； 当 TLP对应某个网络下游接收端入口时， 对应该 TLP的 DCP部署在对应该网络下游接收端入口的下游发送设备上。
在一个业务流进入网络时， 该业务流对应产生唯一的业务流标识， 可选 的， 可以将业务流 ID作为该业务流标识。 由于一个业务流标识与一个业务流 唯一对应， 所以当网络的业务流传送场景出现单点对多点， 或者多点对多点 时， 即， 上游发送设备、 下游接收设备上的 DCP获取多个 TLP对业务流进 行测量得到的时延测量信息时， 可以根据该业务流标识判断相关的时延测量 信息是否属于同一个业务流。
每一个时延测量信息中都含有所属的 TLP标识，这样在获取多个 TLP对 业务流进行测量得到的时延测量信息时， 测量控制点 （ Measurement Control Point, 以下简称 MCP )可以根据每一个 TLP标识区分来自不同 TLP的时延 测量信息。
S102, 将时延测量信息发送给测量控制点 MCP, 以使 MCP根据时戳信 息、 业务流标识以及 TLP标识， 确定网络时延情况。
具体的， DCP获取和发送时延测量信息的过程^^于网络设备的管理网 络来实现的， 对于网络各节点设备， 除了进行业务流发送的业务端口， 还设 有管理端口， 通过这些管理端口即可以组成上述管理网络。 这样时延测量信 息的发送路径可以通过管理网络进行带外传送， 也可以通过业务端口与目标 业务流的发送路径同路径进行带内传送， 可选的， 管理网络可以釆用虚拟专 用网络（Virtual Private Network , 以下简称 VPN)、 数字通信网络 ( Data Communication Network, 以下简称 DCN )或者具备 IP可达的公网。
本实施例提供的网络时延测量方法， 通过 DCP获取至少一个 TLP测量 业务流得到时延测量信息， 实现了对业务流的直接测量， 并由 DCP将该时延 测量信息统一发送给 MCP, 使 MCP根据该时延测量信息中的相关信息确定 网络时延情况。 并且在当有多个 TLP分别对业务流进行测量得到各自的数据 包测量信息时，根据发送时延测量信息以使 MCP根据时延测量信息统一确定 业务流的时延情况。 从而在网络为单点对单点、 或者单点对多点等场景下都 能直接对业务流实现准确的时延测量， 反应业务流真实的时延情况。
基于本发明网络时延测量方法实施例一， 本发明网络时延测量方法实施 例二的方法中， DCP获取至少一个 TLP对业务流进行测量得到的时延测量信 息， 包括：
管理上游 TLP的数据收集点 DCP获取至少一个上游 TLP对发送的业务流 进行测量得到的发送端时延测量信息。
管理下游 TLP的 DCP获取至少一个下游 TLP对接收的业务流进行测量得 到的接收端时延测量信息。
将时延测量信息发送给 MCP, 包括：
管理上游 TLP的 DCP将发送端时延测量信息发送给 MCP, 发送端时延测 量信息包括发送端时戳信息、 业务流标识以及 TLP标识。
具体的， 本实施例通过在上游的每个发送端部署 DCP, 通过该些 DCP 来管理上游各发送端上的 TLP, 当上游 TLP对发送的业务流进行识别， 并对 该业务流的数据包添加时延测量标志后， 添加时延测量标志的时间点即为发 送端时戳信息，上游 TLP产生发送端时延测量信息，该些 DCP获取上游 TLP 的发送端时延测量信息。 该发送端时延测量信息包括， 上游的发送端至少一 个 TLP统计的发送端时戳信息、 业务流标识以及 TLP标识。
管理下游 TLP的 DCP将接收端时延测量信息发送给 MCP, 接收端时延 测量信息包括接收端时戳信息、 业务流标识以及 TLP标识。
具体的， 与上文类似， 本实施例在下游的每个接收端部署 DCP, 通过该 些 DCP来管理下游各接收端上的 TLP、 由下游 TLP首先对业务流进行识别， 若该业务流为目标业务流，下游 TLP在识别到具有时延测量标志的数据包时， 以该时间点为接收端时戳信息， 并产生接收端时延测量信息， 由 DCP获取下 游 TLP对接收的业务流测量得到的接收端时延测量信息。 该接收端时延测量 信息包括， 下游的接收端至少一个 TLP统计的接收端时戳信息、 业务流标识 以及 TLP标。 需要说明的是， 上游、 下游的概念是针对一个业务流的在网络 中的传送方向而定的， 对于不同的业务流， 同一个 TLP既可以是上游的， 同 时也可以是下游的。
一方面， 基于本发明网络时延测量方法实施例二， 本发明网络时延测量 方法实施例三的方法， 还包括：
管理上游 TLP的 DCP在测量周期结束时， 获取测量周期标识， 并将测量 周期标识发送给 MCP。
管理下游 TLP的 DCP获取该测量周期的起始时间， 若起始时间与时戳信 息之差小于等于预设时长， 则接收端时延测量信息属于测量周期标识对应的 测量信息；若起始时间与时戳信息之差大于预设时长，则将测量周期标识加 1 , 时戳信息属于下一个测量周期， 将测量周期标识发送给 MCP。 可选的， 预设 时长为测量周期的 2/3时长。
具体的，对于获取测量周期标识， 可以通过上游 TLP和下游 TLP直接通过 时延测量信息获得；也可以由管理上游 TLP的 DCP和管理下游 TLP的 DCP在读 取到发送端时延那测量信息和接收端时延测量信息后， 根据获取发送端时延 那测量信息的时间点和获取接收端时延测量信息的时间点获得测量周期标 识。
每一个测量周期， DCP与对应的 TLP均可以产生一个对应的测量周期标 识， DCP得到周期标识的公式如下：
测量周期标识 =全局秒数 /测量周期时长。
需要说明的是， 上游 TLP与管理上游 TLP的 DCP、 下游 TLP与管理下 游 TLP 的 DCP 以及各 DCP 之间通过釆用网络时间协议 （Network Time Protocol， 以下简称 NTP)或 IEEE 1588v2时钟进行了时间同步 , 全局秒数可 以是 TLP产生时延测量信息的时间点， 也可以是 DCP读取该时延测量信息 的时间点， 测量周期标识为全局秒数除以测量周期时长的结果取整。 例如， 每个测量周期时长 Is, 当上游 TLP对一个测量周期内的数据包添加时延测量 标志的时间点为 10s, 则该测量周的测量周期标识位根据上述公式计算得为 10s/ls=10; 每个测量周期时间为 2s, 当上游 TLP对一个测量周期内的数据包 添加时延测量标志的时间点为 7s, 7/2=3.5, 则测量周期标识为 3。
由于上游 TLP以测量周期为单位， 在业务流中的每一个测量区间内选取 一个数据包添加时延测量标志，所以每一个测量周期产生一个测量周期标识， 例如， 假设在上游 TLP对数据包 A添加时延测量标志， 由上游 TLP产生发 送端时延测量信息并且产生对应的测量周期标识， 该测量周期标识为 10, 为 了保证 MCP根据与数据包 A对应的发送端时戳信息和接收端时戳信息进行 时延计算， 管理上游 TLP的 DCP在测量周期结束时， 获取上游 TLP产生的 发送端时戳信息和测量周期标识， 并将发送端时戳信息和测量周期标识 (即， 10 )发送给 MCP, 数据包 A经网络传送到达接收端后， 下游 TLP在一个测 量周期内识别到带有时延测量标志的数据包 A后产生接收端时延测量信息和 一个测量周期标识， 并且将该接收端时延测量信息和测量周期标识发送给管 理下游 TLP的 DCP , 管理下游 TLP的 DCP经判断， 若起始时间与时戳信息 之差小于等于预设时长， 则确定该测量周期标识为 10, 并将该接收端时延测 量信息和测量周期标识 10发送给 MCP, 若起始时间与时戳信息之差大于预 设时长， 则 DCP将测量周期标识加 1 , 即该测量周期标识为 11 , 然后将将该 接收端时延测量信息和新的测量周期标识 11发送给 MCP, MCP根据测量周 期标识 11 , 将该接收端时延测量信息与管理上游 TLP的 DCP发送的测量周 期标识为 11的发送端时延测量信息对应。
可选地，为了保证上游 TLP和下游 TLP基于相同的时间产生周期测量标 识， 管理上游 TLP的 DCP釆用 NTP或 IEEE 1588v2时钟与上游 TLP进行时 间同步， 管理下游 TLP的 DCP釆用 NTP或 IEEE 1588v2时钟与下游 TLP进 行时间同步。并且管理上游 TLP的 DCP与管理下游 TLP的 DCP也通过釆用 NTP或 IEEE 1588v2时钟来实现时间同步。
具体的 , 网络时间协议（ Network Time Protocol, 以下简称 NTP )和 IEEE 1588v2时钟都是外部同步工具，其中， NTP是一个普遍釆用的网络同步工具， NTP的同步偏差为 lms〜50ms, 能够保证本实施例提供的网络时延测量方法 的同步要求。 IEEE 1588v2时钟是一种釆用 IEEE 1588v2协议的高精度时钟。 本发明涉及的时间同步方法， 是基于一个共同的时间基准（NTP或 IEEE 以及管理下游 TLP的 DCP的本地时间 ,还有管理上游 TLP的 DCP和管理下 游 TLP的 DCP之间的本地时间。 可选的 , 就是通过 NTP或 IEEE 1588v2时 钟约定各种周期的边界点 （即每个周期的开始时间点）， 即将上游 TLP以及 管理上游 TLP的 DCP和下游 TLP以及管理下游 TLP的 DCP每个测量周期 的起始时间点对齐。 对于已经部署了 IEEE 1588v2时钟的网络， 本实施例提 供的网络时延测量方法可选的釆用 IEEE 1588v2时钟来进行时间同步。
图 2为本发明网络时延测量方法实施例三的实现流程示意图， 下面结合 图 2, 对本发明网络时延测量方法实施例三进行详细说明。
如图 2所示， Rl、 R2为网络节点设备， 在 Rl、 R2上部署 TLP, 以及相 对应的 DCP, 并且在网络任意一个网络节点设备上部署 MCP, 可选的， 将 MCP部署在功能较强的节点设备上。 参照图 2, 对于 R1和 R2, 可以同时有 两个方向相反的业务流进行时延测量。
考虑到 Rl、 R2,都有自己的本地时间，对应时间轴分别为 R1本地时间、 R2本地时间， 所以 Rl、 R2间通过釆用外部时间同步工具等方式， 实现了时 间及周期同步， T[N]、 T[N+1]表示两个相邻测量周期区间对应的各自的测量 周期标识。
由图 2可以看到， 对于 R1以及 R2的两个本地时间的时间轴以及各测量 周期边界点如 T[N]、 T[N+1], 已通过 NTP或 IEEE 1588v2时钟基本对齐， 其 中两个本地时间轴的偏差是网络本身的误差、 NTP或 IEEE 1588v2时钟的精 度导致的。
在 Rl、 R2两端相同测量周期标识的一个测量周期 T[N]内， 在 R1 以及 R2的发送端、接收端的各 TLP上，测量周期开始时（包括前 T/n周期范围内）， 同时相向发起一个对业务流数据包的单向时延测量， 每个测量周期只选取一 个目标业务流内的数据包添加时延测量标志。
在 R1以及 R2的上游发送端 TLP对一个数据包添加时延测量标志，并得 到本地发送时戳 tl、 t3 , 在 Rl以及 R2的下游接收端 TLP , 在相应测量周期 T[N]内， 检测到一个具有时延测量标志的数据包， 下游 TLP可分别得到本地 接收时间戳 t2、 t4, TLP可向管理该 TLP的 DCP上报一个包含时戳信息的时 延测量信息， 或在每周期结束时由管理各 TLP的 DCP读取时延测量信息。
所有时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识。 并且携 带相同周期标识 Τ[Ν] ,业务流标识以及 TLP标识体现了接收或发送方向信息， 该时延测量信息由 DCP读取并发送给 MCP, MCP根据 T[N]即可匹配和进行 计算处理。
对于双向时延， 可以视为两个单向时延之和， 如下面公式：
双向时延=(^241)+( 143)=( 141)-(1342)
该公式也表示两个单向时延测量之间并不必然要求时间顺序耦合 , 因此 可以单独发起和测量。
如果网络部署了精确时间同步， 单向时延 ld(Rl→R2)=t2-tl , ld(R2→ Rl)=t4-t3 , 对于测量周期 T的选择， 假设， 测量周期为 Τ, 业务流传输时延 与乱序延迟之和为 D, 发送端和接收端测量周期同步误差为△, 则测量周期 Τ满足以下两个条件： 一、 (2*A-D) < T/3; 二、 （2* \+D) < 2*T/3。 对于测量周期标识和接收端时戳信息周期归属的规则， 假设， 第 N周期 发送端时戳为 TX, 则每个测量周期开始时刻起 100ms 内， 如果有业务流， 那么对该业务流第一个数据包添加时延测量标志， 上游 TLP 记录当时时戳 Time— TX[N], 并获得测量周期标识 N; 否则， 本周期不标记时延测量 4艮文。
对于接收端所属测量周期的判定， 假设第 N周期内， 接收端收到有时延 测量标志的数据包， 下游 TLP记录本地时戳 Time— RX。 若在本周期结束时下 游 TLP获得 Time— RX , 则有管理下游 TLP的 DCP进行如下计算：
计算 Time— RX-Time[N](Time[N]为当前周期的起始时间， 由与 DCP釆用 NTP或 IEEE 1588v2时钟与被管理的 TLP进行了时间同步， 所以 DCP可以 直接获取该 DCP本地时间上的周期起始时间）；
若大于 2T/3 ( T为周期时长）， 该时戳属于下一周期（测量周期标识 +1 : 由于同步误差导致的超前数据包）， 否则， 该时戳信息属于当前周期；
若在本周期 2T/3处读取， 得到的 Time— RX就是本周期 N的接收时间戳 Time— RX[N];
如果/ <1001118, 传输时延 +乱序延迟 D小于 200ms, 则选取时延测量周 期 T大于 1秒即可。
本发明实施例提供的网络时延测量方法， 通过管理上游 TLP的 DCP在 测量周期结束时， 获取测量周期标识以及发送端时延测量信息， 并将测量周 期标识发送给 MCP, 再通过下游 TLP的 DCP获取测量周期标识以及接收端 时延测量信息， 由该 DCP对获取的测量周期标识进行判断， 再将判断后的测 量周期标识以及接收端时延测量信息发送给 MCP, 以使 MCP根据上游的测 量周期标识和下游的测量周期标示， 将属于同一业务流同一周期内的发送端 时延测量信息和接收端时延测量信息对应起来， 直接准确地测量业务流时延 情况。
另一方面， 基于本发明网络时延测量方法实施例二， 本发明网络时延测 量方法实施例四的方法， 还包括：
管理下游 TLP的 DCP获取至少一个下游 TLP接收的由上游 TLP发送的测 量报文以及测量报文到达下游 TLP时产生的一个测量报文的到达时戳信息， 测量报文包括： 发送端时戳信息。
具体的， 对于网络接收端和发送端设备均以测量周期为单位对业务流进 行时延测量， 但是接收端的测量周期与发送端的测量周期并没有通过时间同 步工具进行时间同步的网络， 上游 TLP在每一个测量周期对一个数据包添加 时延测量标志并产生一个发送端时延测量信息， 该发送端时延测量信息包括 时戳信息、 业务流标识以及 TLP标识， 上游 TLP将包含有发送端时戳信息的测 量报文发送给接收端下游 TLP ,管理下游 TLP的 DCP通过该测量 4艮文的到达时 戳信息与接收端时延测量信息进行比对， 保证发送端时延测量信息与接收端 时延测量信息属于同一测量周期。
管理下游 TLP的 DCP进行匹配识别， 确定到达时戳信息和接收端时戳 信息是否同属于预设时长范围， 若是， 则确定测量报文中的发送端时戳信息 和下游 TLP产生的接收端时戳信息属于同一测量周期， 也即属于同一个数据 包 (由于每周期仅选取一个数据包添加时延测量标志）， 并将确定结果发送给 MCP。
可选的， 管理下游 TLP的 DCP可以将属于同一测量周期的发送端时戳 信息和接收端时戳信息发送给 MCP , 由 MCP来进行计算。 也可以由该 DCP 根据属于同一测量周期的发送端时戳信息和接收端时戳信息直接确定该周期 数据包的时延情况， 再将计算好的时延情况发送给 MCP。
具体的， 图 3为本发明网络时延测量方法实施例四的实现流程示意图， 如图 3所示， 对于添加了时延测量标志的数据包在经过网络传送到达接收端 时会有一个时延 Ds,对于上游 TLP发送的携带发送端时戳信息的测量报文到 达接收端时也会有一个时延 Dc,通过下面的公式定义一个时延差，公式如下： 时延差 Z^Ds - Dc|。
参照图 3 , 假设发送端的上游 TLP与接收端的下游 TLP的测量周期都为 T , 在发送端（TX )的上游 TLP, tl时刻对业务流数据包 A添加时延测量标志， 得 到本地发送端时戳信息 tl ,并生成携带发送端时戳信息 tl的测量报文发送给接 收端 （RX ) 的下游 TLP, 经过时延 Ds, 由于可能会出现乱序， 所以可能是数 据包 A先到达， 也可能是测量报文先到达， 当数据包 A先到达接收端的下游 TLP, 接收端的下游 TLP得到业务流在接收端时戳信息 t2, 经过时延 Dc, 携带 tl的测量报文到达接收端， 接收端得到测量报文的到达时戳信息 tc。 当测量报 文先到达下游 TLP时， 原理相同。
管理下游 TLP的 DCP的匹配识别设置 |tc-t2|<时延差△,而在一个测量时延 间隔周期 T中， 只在测量周期开始时对一个业务流数据包添加时延测量标志， 且满足 Τ»时延差△, 管理下游 TLP的 DCP因此有下面匹配识别操作：
单次测量中， 管理下游 TLP的 DCP首先获取到接收端时戳信息 t2, 以接收 端时戳信息 t2为基准在加或减时延差 的时间范围内，在某一时刻 tc管理下游 TLP的 DCP获取到测量报文， 则发送端时戳信息 tl可与接收端时戳信息 t2匹 配， 即， 发送端时戳信息和接收端时戳信息属于同一测量周期的同一数据包， 或者管理下游 TLP的 DCP首先获取到测量报文， 以该测量报文的到达时戳信 息 tc为基准在加或减时延差 时间范围内管理下游 TLP的 DCP获取到数据包 的接收端时戳信息 t2, 则发送端时戳信息 tl可与接收端时戳信息 t2匹配， 即， 发送端时戳信息和接收端时戳信息属于同一测量周期的同一数据包。
周期性测量中， 每个时延测量周期间隔 （上游 TLP对业务流数据包添加 时延测量标志的时间间隔为 T ) , 上游 TLP只对一个业务流数据包添加时延测 量标志， 由于实际网络中 Ds、 Dc都有抖动、 变长， 但是有限的， 存在时延差 /的最大值 Z (MAX),只要测量周期间隔 Τ>2* Δ(ΜΑΧ)+最小安全间隔时间， 即可以确定每釆样间隔内的 t2时戳与 tl协议报文匹配；
假设网络中 Δ(ΜΑΧ) =500ms,考虑 100ms安全处理时间，则1 2*500+100 = LIS 可以实现周期测量。
本发明实施例提供的网络时延测量方法， 通过管理下游 TLP的 DCP获 取至少一个下游 TLP接收的由上游 TLP发送的测量报文， 并且由管理下游 TLP的 DCP进行匹配识别，确定发送端时戳信息和接收端时戳信息是否同属 于预设时长范围， 若是， 则确定发送端时戳信息和接收端时戳信息属于同一 测量周期， 并将确定结果发送给 MCP, 实现了准确直接地测量业务流的时延 情况。
再一方面， 基于本发明网络时延测量方法实施例二， 本发明网络时延测 量方法实施例五的方法， 还包括：
发送端时延测量信息还包括： 发送端业务流特征信息以及发送端分片重 组标识， 接收端时延测量信息还包括： 接收端业务流特征信息以及接收端分 片重组标识， 以使 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及接收端分片重组标识， 发送端时戳信息和接收端 时戳信息是对应于同一业务流的时戳信息。 具体的， 在业务流时延的测量过程中， 关键是确定发送端时戳信息和接 收端时戳信息是同一个被添加时延测量标志的业务流数据包经过一段时延在 网络的发送端和接收端分别获得的。 对于一个业务流的数据包， 该数据包所 包含的业务流特征信息以及分片重组标识就可以唯一地识别该数据包。 业务 流特征信息为 IP头中的五元组以及服务类型 TOS ( ( Type of Service , 以下简 称 TOS ) )字段的信息， 其中五元组指 IP头中的源 IP地址、 目的 IP地址、 协议类型、 源协议端口号、 目的协议端口号。 在数据包的传送过程中， 经常 会将一个过大的数据包分成多个字数据包再进行传输的情况。 对于一个被分 片的数据包， 所属每一片子数据包的分片重组标识都是一样的， 在接收端接 收到各分片子数据包后， 可以根据各子数据包的分片重组标识将各子数据包 重组为原数据包。
因此当一个业务流被上游 TLP识别后，上游 TLP对该业务流的一个数据 包 A添加时延测量标志， 并生成发送端时延测量信息包括发送端时戳信息、 业务流标识、 TLP标识、 发送端业务流特征信息以及发送端分片重组标识， 并由管理上游 TLP的 DCP获取该发送端时延测量信息， 并发送给 MCP。 当 下游 TLP识别到具有时延测量标志的数据包 A时，生成接收端时延测量信息 包括接收端时戳信息、 业务流标识、 TLP标识、 接收端业务流特征信息以及 接收端分片重组标识， 并由管理下游 TLP的 DCP获取该接收端时延测量信 息， 并发送给 MCP。 MCP根据发送端业务流特征信息以及发送端分片重组 标识和接收端业务流特征信息以及接收端分片重组标识可以确定发送端时戳 信息和接收端时戳信息是同一个被添加时延测量标志的业务流数据包经过一 段时延在网络的发送端和接收端分别获得的， 即实现了发送端时延测量信息 和接收端时延测量信息的匹配。从而由 MCP依据匹配成功的发送端时延测量 信息和接收端时延测量信息进行时延测量。
图 4为本发明网络时延测量方法实施例五的实现流程示意图， 下面结合 图 4, 对本发明网络时延测量方法实施例五进行详细说明。
在时延测量中， 关键是确定发送端的时戳信息和接收端的时戳信息是同 一个被添加时延测量标志的业务流数据包经过网络时获得的， 本实施例用数 据包的五元组（源 IP地址、 目的 IP地址、 协议类型、 源协议端口号、 目的协 议端口号） ， 以及分片重组标识（分片重组 ID ) 的方式来匹配确定。 本发明 网络时延测量方法实施例五的匹配原理如下：
在 IP网络中， 对于同一个 VPN (在一个 VPN中， 数据包具有不重叠的 地址空间）内，通过 IP才艮文的五元组可以确定一个业务流，在一定时间内（一 个 ID循环） ， 一个五元组确定的业务流上数据包的分片重组 ID (未分片） 是独一无二的。 因此一个测量域（可以包含多个不同的业务流） 内业务流的 数据包， 在发送端和接收端， 通过 ^艮文五元组 +分片重组 ID, 即可唯一确定 该业务流的数据包（对于同一个业务流， 各数据包的分片重组 ID不同； 不同 的业务流， 各业务流的五元组不同） 。
在发送端和接收端上管理各 TLP的 DCP读取时延测量信息， 并读取业 务流特征信息（五元组） 以及分片重组标识（分片重组 ID ) , 即可匹配同一 个业务流数据包的时戳信息。
对于在网络中被分片的数据包， 管理各 TLP的 DCP读取接收到的第一 个数据包的时戳信息即可 (分片数据包的乱序及延迟通常会小于测量周期， 而主机侧分片重组 ID循环的时间则会长于测量周期）。
对于错误！ 未找到引用源。 所示测量实现流程， 五元组确定的业务流数 据包， 在发送端， 分片重组 ID为 100, 到接收端依然是 100不变， 通过这个不 变的特性即可确定两端获得的时戳信息是对同一数据包的检测结果。
另外， 可选的， 在接收端， 时延测量信息如携带本地周期标识（不需要 进行时间同步）， 可以进一步确定时延信息的顺序和双向时延测量的配对。
本发明实施例提供的网络时延测量方法， 通过管理上游 TLP的 DCP获 取上游 TLP测量得到的发送端时延测量信息， 发送给 MCP。 管理下游 TLP 的 DCP获取下游 TLP测量得到的接收端时延测量信息， 发送给 MCP, 因为 该发送端时延测量信息还包括发送端业务流特征信息以及发送端分片重组标 识， 该接收端时延测量信息还包括接收端业务流特征信息以及接收端分片重 组标识，以使 MCP根据发送端业务流特征信息以及发送端分片重组标识和接 收端业务流特征信息以及接收端分片重组标识可以确定发送端时戳信息和接 收端时戳信息是同一个被添加时延测量标志的业务流数据包经过一段时延在 网络的发送端和接收端分别获得的， 从而进行时延的快速准确测量。
上述实施例描述了本发明网络时延测量方法中 DCP 所执行的具体方法 S, 下面对本发明网络时延测量方法中 TLP所执行的具体方法 S进行详细说 明。
图 5为本发明网络时延测量方法实施例六的流程图， 如图 5所示， 本实 施例的方法可以包括：
S200, 根据业务流特征信息对业务流进行识别， 确定业务流是否是目标 业务流。
具体的， 首先在每一个上游发送端和下游接收端上部署 TLP, 可选的， 上游 TLP和下游 TLP 可以同时被部署在发送端和接收端的用户侧或者网络 侧。 因为每一个业务流都有其特定的业务流特征信息， 对于业务流特征信息， 上文网络时延测量方法实施例五中已进行了详细说明， 此处不再赘述。 所以 当一个业务流进入网络后， 上游发送端口 TLP首先根据业务流特征信息对业 务流进行识别， 该识别过程是根据预先设置的业务流特征信息与该业务流的 报文头信息进行匹配识别， 如果两者匹配成功， 上游发送端口 TLP确定该业 务流为目标业务流。 当出现网络的业务流传送场景为单点对多点， 或者多点 对多点的场景时， 无论业务流的具体路径如何， 都可以依据该业务流的业务 流特征信息确定各上游发送端以及各下游接收端上的数据包是否属于同一个 业务流。
S202, 若是， 则对业务流的数据包添加时延测量标志， 获取与该时延测 量标志对应的数据包时延测量信息。
具体的， TLP对目标业务流的数据包添加一个时延测量标志， 并且生成 时延测量信息， 该时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 其中， 时戳信息就是 TLP添加时延测量标志的时间点。 对于业务流标识以及 TLP标识， 网络时延测量方法实施例一已进行了详细说明 , 此处不再赞述。
S204, 确定时延测量信息。
具体的， TLP在生成时延测量信息后， 以使 DCP在获取该时延测量信息 后将时延测量信息发送给 MCP,以使 MCP根据时延测量信息确定时延情况。
本发明实施例提供的网络时延测量方法， 通过 TLP根据业务流特征信息 对业务流进行识别， 确定业务流是否是目标业务流， 若是， 再由 TLP对业务 流的数据包添加时延测量标志， 获取与该时延测量标志对应的数据包时延测 量信息， 并由 TLP确定时延测量信息， 以使 DCP在获取时延测量信息后将 时延测量信息发送给 MCP, 以使 MCP根据时延测量信息进行时延情况的确 定。 实现了直接对业务流数据包的时延测量， 提高了时延测量的准确性和真 实性。
基于本发明网络时延测量方法实施例六， 可选的， 本发明网络时延测量 方法实施例七的方法， 对业务流的数据包添加时延测量标志， 获取与该时延 测量标志对应的数据包时延测量信息， 包括：
上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与该时延测 量标志对应的数据包发送端时延测量信息， 发送端时延测量信息包括时戳信 息、 业务流标识以及 TLP标识， 以使管理上游 TLP的 DCP获取发送端时延测量 信息后将发送端时延测量信息发送给 MCP。
下游 TLP在识别到添加时延测量标志的数据包时， 获取与该时延测量标 志对应的数据包接收端端时延测量信息。接收端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 以使管理下游 TLP的 DCP获取接收端时延测量 信息后将接收端时延测量信息发送给 MCP。
可选的，在数据包的 IP头中 TOS的保留位或者 Flags的保留位上添加时 延测量标志。
具体的， 时延测量标志可以指定的范围是数据包 IP头中 TOS、 Flags两 个域中的共 6个位， TOS第 3 ~ 7位， flags第 0位。 具体的， 在不同的具体 网络中 TOS后几位（第 3 ~ 7位） 经常不用， 特别是第 6、 7位， 很少使用， 因此 IP头的这几位可借用用于添加标识。 在 IPv4的 IP头中， Flags的第 0 位是 IP头中当前唯一的一个保留位， 在通常 IP头中， 该位可以用于对数据 包进行标识的添加。
在具体实现时， 根据业务流特征信息对业务流进行识别， 可以包括： 才艮据五元组中的至少两元信息， 对业务流进行识别。
具体的， 五元组指 IP头中的源 IP地址或其 IP地址前缀、 目的 IP地址或 其 IP地址前缀、 协议类型、 源协议端口号、 目的协议端口号， 除了五元组之 外， 可选的， 也可以添加 IP头中 TOS字段的信息来对业务流特征信息进行 指定， 上述的字段可以全部指定， 这样测量的业务流比较精细； 也可以部分 指定， 例如， 指定五元组中的至少源 IP地址、 目的 IP地址这两元信息； 或 者， 源 IP地址前缀、 目的 IP地址前缀； 或者源 IP地址或其 IP地址前缀、 目 的 IP地址或其 IP地址前缀、 服务类型 （Type of Service, 以下简称 TOS )信 息。
一方面， 基于本发明网络时延测量方法实施例七， 本发明网络时延测量 方法实施例八的方法， 还包括：
上游 TLP在目标业务流的数据包上添加时延测量标志之前， 釆用 NTP或 IEEE 1588v2时钟与管理上游 TLP的 DCP进行时间同步， 下游 TLP在识别到添 加时延测量标志的数据包之前， 釆用 NTP或 IEEE 1588v2时钟与管理下游 TLP 的 DCP进行时间同步。
具体的， 时间同步的方法和原理， 本发明网络时延测量方法实施例三的 方法已进行了详细说明， 此处不再赘述。 参见图 2, 上游 TLP以测量周期为单 位在每个测量周期区间对业务流的一个数据包添加时延测量标志并生成发送 端时延测量信息以及测量周期标识， 下游 TLP也以测量周期为单位对添加时 延测量标志的数据包进行识别并生成接收端时延测量信息以及测量周期标 识， 在时延的测量中， 关键在于确定发送端时戳信息与接收端时戳信息是同 一个被添加时延测量标志的数据包经过网络传输后由上游 TLP和下游 TLP分 别获得的。对于本实施例 ,可以通过在上游 TLP和下游 TLP进行时延测量之前 , 在上游 TLP和下游 TLP上分别部署外部时间同步工具 NTP或 IEEE 1588v2时 管理下游 TLP的 DCP之间的时间同步。 可选的， 管理上游 TLP的 DCP和管理下 游 TLP的 DCP也均部署外部时间同步工具 NTP或 IEEE 1588v2时钟， 以保证 TLP与 DCP、 DCP与 DCP之间的时间同步， 这样就保证了每个测量周期， 上游 TLP产生的发送端测量周期标识和下游 TLP产生的接收端测量周期标识能够 匹配， 进而保证了具有相同测量周期标识的发送端时延测量信息和接收端时 延测量信息能够匹配， 以使 MCP准确地确定时延情况。
对业务流的数据包添加时延测量标志， 获取与该时延测量标志对应的数 据包时延测量信息， 还包括：
上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与时延测量 标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取测量周期标识后将 测量周期标识信息发送给 MCP。 因为上游 TLP和下游 TLP均釆用了 NTP或 IEEE 1588v2时钟与管理上游 TLP的 DCP和管理下游 TLP的 DCP进行时间同 步， 所以， 也可以由管理上游 TLP的 DCP和管理下游 TLP的 DCP在读取发送端 时延测量信息和接收端时延测量信息后， 上游 TLP的 DCP根据读取到的发送 端时延测量信息获取对应的测量周期标识， 管理下游 TLP的 DCP根据读取到 的接收端时延测量信息获取对应的测量周期标识， 并且两个测量周期标识对 于属于同一个测量周期， 被添加时延测量标志的该数据包来说是一致的。
具体的， 每个测量周期开始时刻起 100ms内， 如果上游 TLP识别到目标 业务流， 那么上游 TLP对该目标业务流的第一个数据包添加时延测量标志， 并记录当时的时戳信息 tl , 并获得测量周期标识 T[N], 如图 2。 若上游 TLP 没有识别到目标业务流， 则该测量周期不对数据包添加时延测量标志。
下游 TLP在每个测量周期获取该测量周期起始时间， 在每个测量周期内 识别到添加时延测量标志的数据包时， 获取与时延测量标志对应的测量周期 标识， 以使管理下游 TLP的 DCP获取起始时间和测量周期标识后将测量周 期标识信息发送给 MCP。
具体的， 参照图 2, 假设一个测量周期为 T, 下游 TLP在第 N个测量周 期起始时记录该测量周期的起始时间 to , 若在该测量周期内识别到了具有时 延测量标志的数据包， 则下游 TLP记录当时的时戳信息 t3 , 并生成测量周期 标识 M以及接收端时延测量信息， 并且在第 N各测量周期结束时， 由管理下 游 TLP的 DCP获取接收端时延测量信息、 测量周期标识 T[N]以及起始时间 t0, 该接收端时延测量信息包括： 时戳信息 t3、 业务流标识以及 TLP标识， 由管理下游 TLP的 DCP对时戳信息 t3与起始时间 t0进行计算，若 t3- tO<2T/3 , 则该时戳信息 t3属于该第 N测量周期， 若 t3- tO≥2T/3 , 则该时戳信息 t3属 于第 N+1测量周期， 此时由管理下游 TLP的 DCP对周期标志 T[N]加 1为 T[N+1],这样在具有时间同步误差导致数据包乱序的情况下， MCP依然能够, 在接收到管理下游 TLP的 DCP发送的接收端时延测量信息以及测量周期标 识后， 根据测量周期， 确定发送端时戳信息和接收端时戳信息是对应于同一 数据包的时戳信息， 进而准确直接地进行时延测量。
并且， 对于本实施例提供的网络时延测量方法， TLP对业务流数据包进 行识别，在每一个周期 TLP对该测量周期区间内的数据包添加时延测量标志， 可选的， 在每一个测量周期内， TLP只对一个数据包添加时延测量标志。
本发明实施例提供的网络时延测量方法， 通过上游 TLP在目标业务流的 数据包上添加时延测量标志之前， 釆用 NTP或 IEEE 1588v2时钟与管理上游 TLP的 DCP进行时间同步， 下游 TLP在识别到添加时延测量标志的数据包 之前，釆用 NTP或 IEEE 1588v2时钟与管理下游 TLP的 DCP进行时间同步， 保证了上游 TLP与下游 TLP时间同步， 从而可以保证在相同的周期内上游 TLP 产生的测量周期标识与下游 TLP 产生的测量周期标识一致， 因此以使 MCP依据相同的周期测量标识确定发送端时戳信息和接收端时戳信息是对 应于同一数据包的时戳信息， 并且在发生时间同步误差导致的数据包乱序情 况下， 由下游 TLP在每个测量周期获取该测量周期起始时间， 在每个测量周 期内识别到添加时延测量标志的数据包时， 获取与时延测量标志对应的测量 周期标识， 以使管理下游 TLP的 DCP获取起始时间和测量周期标识后， 由 DCP根据起始时间和接收端时戳信息， 确定正确的测量周期标识， 并将正确 的测量周期标识信息发送给 MCP, 保证 MCP准确的确定时延情况。
另一方面， 基于本发明网络时延测量方法实施例七， 本发明网络时延测 量方法实施例九的方法， 还包括：
由上游 TLP向下游 TLP发送测量报文， 测量报文包括： 发送端时戳信息。 下游 TLP的接收模块接收到测量报文， 产生一个测量报文的到达时戳信 息， 并将测量报文、 以及到达时戳信息发送给管理下游 TLP 的 DCP, 以使 DCP确定到达时戳信息和接收端时戳信息是否同属于预设时长范围 , 若是， 则确定发送端时戳信息和接收端时戳信息属于同一测量周期， 并将确定结果 发送给 MCP。
具体的， 参照图 3 , 对于上游 TLP以测量周期 T为单位对业务流进行时 延测量，但是上游 TLP和下游 TLP的测量周期并没有通过时间同步工具进行 时间同步。 为了保证发送端时延测量信息与接收端时延测量信息属于同一测 量周期。 本实施例当上游 TLP以测量周期为单位产生一个发送端时延测量信 息， 该发送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 并由 管理上游 TLP的 DCP获取该发送端时延测量信息时， 由上游 TLP生成一个 包含有发送端时戳信息 tl的测量报文， 并发送给接收端下游 TLP, 下游 TLP 在一个测量周期内生成一个接收端时延测量信息并接收到该测量报文， 由管 理下游 TLP的 DCP获取该接收端时延测量信息和测量报文， 并通过该测量 报文与接收端时延测量信息进行比对， 具体的比对方法和技术方案在网络时 延测量方法实施例四已经进行了详细说明， 此处不再赘述。 并且， 对于本实施例提供的网络时延测量方法， TLP对业务流数据包进 行识别，在每一个周期 TLP对该测量周期区间内的数据包添加时延测量标志， 可选的， 在每一个测量周期内， TLP只对一个数据包添加时延测量标志。
本发明实施例提供的网络时延测量方法，通过由上游 TLP向下游 TLP发 送测量报文， 该测量报文包括： 发送端时戳信息， 再由下游 TLP将接收到的 测量报文发送给 DCP,以使 DCP确定发送端时戳信息和接收端时戳信息是否 同属于预设时长范围， 若是， 则确定发送端时戳信息和接收端时戳信息属于 同一测量周期， 并将确定结果发送给 MCP, 保证了发送端时延测量信息与接 收端时延测量信息属于同一测量周期， 实现了对于业务流直接准确的时延测 量。
再一方面， 基于本发明网络时延测量方法实施例七， 本发明网络时延测 量方法实施例十的方法， 还包括：
发送端时延测量信息还包括： 发送端业务流特征信息以及发送端分片重 组标识； 接收端时延测量信息还包括： 接收端业务流特征信息以及接收端分 片重组标识。
以使管理上游 TLP的 DCP获取发送端时延测量信息并发送给 MCP, 管 理下游 TLP的 DCP获取接收端时延测量信息并发送给 MCP,以使 MCP根据 发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及 接收端分片重组标识， 确定发送端时戳信息和接收端时戳信息是对应于同一 业务流的时戳信息。
具体的， 参照图 4, 在网络时延测量中， 关键在于确定发送端时戳信息 与接收端时戳信息是同一个被添加时延测量标志的数据包经过网络传输后由 上游 TLP和下游 TLP分别获得的。对于本实施例， 因此当一个业务流被上游 TLP识别后，上游 TLP对该目标业务流的一个数据包 A添加时延测量标志 1 , 由上游 TLP生成的发送端时延测量信息包括发送端时戳信息、 业务流标识、 TLP标识、 发送端业务流特征信息以及发送端分片重组标识， 并由管理上游 TLP的 DCP获取该发送端时延测量信息， 并发送给 MCP。 并且， 由于本实 施例并不是基于周期对数据包添加时戳， 所以本实施例中 TLP对于一个业务 流的数据包添加时延测量标志可以艮密集。 当下游 TLP识别到具有时延测量 标志的数据包 A时，由下游 TLP生成接收端时延测量信息包括接收端时戳信 息、 业务流标识、 TLP标识、 接收端业务流特征信息以及接收端分片重组标 识， 并由管理下游 TLP的 DCP获取该接收端时延测量信息， 并发送给 MCP。 MCP 根据发送端业务流特征信息以及发送端分片重组标识和接收端业务流 特征信息以及接收端分片重组标识可以确定发送端时戳信息和接收端时戳信 息是同一个被添加时延测量标志的业务流数据包经过一段时延在网络的发送 端和接收端分别获得的， 即实现了发送端时延测量信息和接收端时延测量信 息的匹配。从而由 MCP依据匹配成功的发送端时延测量信息和接收端时延测 量信息进行时延测量， 网络时延测量方法实施例五中对业务流特征信息以及 分片重组标识进行了详细说明， 此处不再赘述。
本发明实施例提供的网络时延测量方法， 通过上游 TLP识别并对目标业 务流数据包进行时延测量后， 将发送端业务流特征信息以及发送端分片重组 标识记录下来， 并由上游 TLP生成发送端时延测量信息， 在本发明实施例提 供的发送端时延测量信息不仅包括发送端时戳信息、 业务流标识以及 TLP标 识， 还包括上述发送端业务流特征信息以及发送端分片重组标识， 以使管理 上游 TLP的 DCP获取发送端时延测量信息并发送给 MCP, 下游 TLP也做类 似操作，以使管理下游 TLP的 DCP获取接收端时延测量信息并发送给 MCP, 以使 MCP根据发送端业务流特征信息以及发送端分片重组标识和接收端业 务流特征信息以及接收端分片重组标识实现了发送端时延测量信息和接收端 时延测量信息的匹配。 进而保证了直接延时测量的准确性。
下面对本发明网络时延测量方法中 MCP所执行的方法 S进行详细说明。 图 6为本发明网络时延测量方法实施例十一的流程图， 如图 6所示， 本 实施例的方法可以包括：
S300, 接收上游 TLP对应的 DCP发送的发送端时延测量信息以及与下 游 TLP对应的 DCP发送的接收端时延测量信息。
具体的，发送端时延测量信息包括发送端时戳信息、业务流标识以及 TLP 标识，接收端时延测量信息包括接收端时戳信息、业务流标识以及 TLP标识。
S302, 根据发送端时延测量信息和接收端时延测量信息， 确定网络单次 时延情况。
具体的， MCP根据发送端时延测量信息和接收端时延测量信息， 将属于 同一个业务流的同一个数据包的发送端时戳信息与接收端时戳信息进行时延 计算。 另外，本发明实施例提供的网络时延测量方法可以将 MCP部署在整个 网络中任意一个网元节点上，可选的，将 MCP部署在一个功能较强的网元节 点上 , 并且 , MCP、 各 DCP以及各 TLP基于管理网络进行连接。
本发明实施例提供的网络时延测量方法， 通过 MCP接收上游 TLP对应 的 DCP发送的发送端时延测量信息以及与下游 TLP对应的 DCP发送的接收 端时延测量信息，再通过 MCP根据发送端时延测量信息和接收端时延测量信 息， 确定网络单次时延情况， 实现了对于业务流直接准确的时延测量。
一方面， 基于本发明网络时延测量方法实施例十一， 本发明网络时延测 量方法实施例十二的方法， 还包括：
MCP接收管理上游 TLP的 DCP发送的测量周期标识， MCP接收管理下 游 TLP的 DCP发送的测量周期标识 , 由 MCP根据管理上游 TLP的 DCP发 送的测量周期标识和管理下游 TLP的 DCP发送的测量周期标识， 判断发送 端时延测量信息和接收端时延测量信息是否属于同一测量周期， 若是， 则 MCP根据发送端时延测量信息和接收端时延测量信息， 确定网络单次时延情 况。
具体的， 当对目标业务流进行时延测量时， 可选的， 在 MCP中维护该目 标业务流测量数据汇总表。 表 1为本实施例提供的目标业务流测量数据汇总 表， 下面根据表 1 , 对本实施例如何确定网络时延情况进行说明：
表 1为目标业务流测量数据汇总表
Figure imgf000032_0001
反向业 接收端
务流标 时戳 时戳信 无效 无效 无效 无效 无效 下游 识 到 息 对于目标业务流测量数据汇总表 1 , 需要说明的是， 表 1中左侧 TLPs和右 侧 TLPs的概念是以网络为界限进行划分的，在将一个网络一侧定义为左侧后， 左侧端口部署的 TLPs为左侧 TLPs, 相对应的定义右侧 TLPs。 在本发明实施例 提供的网络时延测量过程中，可能同时出现两个方向相反的目标业务流， MCP 可以依据两个目标业务流上五元组中的源 IP地址或其 IP地址前缀，以及目的 IP 地址或其 IP地址前缀， 将其中一个目标业务流定义为正向业务流， 进而将另 一个目标业务流定义为反向业务流。 例如， 一个网络的范围包括由左侧端口 设备至右侧端口设备在内的所有设备和网络。 支设目标业务流 A从左侧端口 设备上的各 TLP进入该网络，从右侧端口设备上的各 TLP离开该网络； 目标业 务流 B从右侧端口设备上的各 TLP进入该网络， 从左侧端口设备上的各 TLP离 开该网络。 对于目标业务流 A来说， 左侧端口设备为发送端设备， 左侧 TLP 为上游发送端， 对目标业务流 B来说， 左侧端口设备为接收端设备， 左侧 TLP 为下游接收端。 因此基于一个 TLP同时对两个方向不同的目标业务流进行时 延测量时， 在 MCP的目标业务流测量数据汇总表 1上， 每一个周期 MCP都维 护一个正向业务流的数据表项和一个反向业务流的数据表项， 从而实现同时 对两个反向业务流进行时延测量的功能。
DCP读取各 TLPs的时延测量信息并发送给 MCP, 首先 MCP根据该业务流 标识， 可选的， 可以以目标业务流 ID作为该业务流标识， 找到对应的目标业 务流测量数据汇总表， MCP再根据业务流标识、 测量周期标识和 TLP标识， TLP标识， 将数据凑入对应相应的测量数据汇总表数据项中。 在本实施例中， 对于表 1 ,测量周期标识可以唯一确定发送端时延测量信息和接收端时延测量 信息属于同一数据包， 即， 表 1中的数据包表项， 一个测量周期标识对应一个 数据包。
对于 MCP接收时延测量信息并维护表 1的过程， 需要说明的是， 由于时延 测量针对的是数据包， 所以在一个数据包的传输过程中， 只会出现一个上游 TLP产生的发送端时戳信息和一个下游 TLP产生的接收端时戳信息，因此参照 表 1当 MCP收到一个时延测量信息并由 MCP更新到左侧 TLPs中一个 TLP的数 据项内后，对于该数据包的左侧 TLPs的其他 TLP数据项就被 MCP设置为无效， 对于右侧 TLPs, MCP也进行类似操作。
参照表 1 ,假设一个业务流为正向业务流， 对于该正向业务流的每个数据 包的数据表项都会在目标业务流测量数据汇总表中对应一个数据到齐标志， 在每一个测量周期标识对应的一个数据包的左侧 TLPs和右侧 TLPs的时延测 量信息没有到时， 由 MCP将该测量周期的数据到齐标志设置为 "未到" ， 例 如表 1中， 测量周期标识为 N-1的正向业务流对应的数据项中， 左侧 TLPs中， MCP收到管理上游 TLP的 DCP发送的第一个 TLP测量得到的发送端时延测量 信息， 并且右侧 TLPs中的时延测量信息未到， 此时 MCP将对应的数据项到齐 标志设置为 "上游到" ， 当 MCP收到管理下游 TLP的 DCP发送的下游 TLP的时 接收延测量信息后， MCP更新表 1 ,将接收端时延测量信息中携带的接收端时 戳信息填入对应的 TLP数据表项中， 将对应的数据项到齐标志设置为 "全部 到齐" 。
当 MCP检测到一个测量周期标识中一个正向业务流标识或者一个反向 业务流对应的数据项的数据到齐标志设置为 "全部到齐"后， 此时 MCP将根 据对应的发送端时戳信息和接收端时戳信息进行时延计算， 具体公式如下： 时延 =接收端时戳信息-发送端时戳信息
即， 数据包在上游 TLP 添加时延测量标志的时间点与该数据包被下游 TLP识别的时间点之差。
参照表 1 , 对于一个测量周期标识， 可以同时进行正向业务流和反向业 务流的双向时延测量， 此时 MCP根据正向业务流对应的发送端时戳信息，和 接收端时戳信息以及反向业务流对应的发送端时戳信息和接收端时戳信息进 行实验计算， 具体公式如下：
时延 = (反向业务流对应的接收端时戳信息-正向业务流对应的发送端时 戳信息)- （反向业务流对应的发送端时戳信息-正向业务流对应的接收端时戳 信息）
图 7为本发明网络时延测量方法实施例十二的双向时延测量示意图， 参 照图 7 , 对上述双向时延的计算原理和方法进行进一步说明。
如图 7所示， 殳图 7中上方的业务流为正向业务流， 网络左侧 TLP对 一个数据包添加时延测量标志 1 , 并记录时刻为 tl , 所以正向业务流对应当 的发送端时戳信息为 tl , 当右侧的 TLP识别该具有时延测量标志 1的数据包 时， 记录该时刻 t2, 即正向业务流对应当的接收端时戳信息为 t2, 左侧 TLP 和右侧 TLP分别生成发送端时延测量信息和接收端时延测量信息， 并由管理 左侧 TLP的 DCP (对于该正向业务流左侧 TLP为上游 TLP )和管理右侧 TLP 的 DCP 分别读取发送端时延测量信息和接收端时延测量信息， 并发送给 MCP。 同理， 对于反向业务流， MCP接收发送端时戳信息 t3和接收端时戳 信息 t4。 MCP计算正向业务流数据包的时延为 t2-tl ,反向业务流数据包的时 延 t4-t3 , 则双向时延为 (t2-tl)+(t4-t3), 即 （t4-tl ) -(t3-t2)。
本实施例提供的网络时延测量方法，首先通过 MCP釆用外部时间同步工 具与各 DCP进行时间同步，再通过 MCP接收上游 TLP和下游 TLP对应的各 DCP发送的时延测量信息， 时延测量信息为时延测量信息发送端时延测量信 息和接收端时延测量信息，通过 MCP记录和维护目标业务流测量数据汇总表 上游发送端的 TLP的发送端时延测量信息和下游接收端 TLP的接收端时延测 量信息，再由 MCP根据该目标业务流测量数据汇总表以每个测量周期标识为 单位， 对发送端时延测量信息与接收端时延测量信息进行时延计算， 从而直 接准确地确定网络的业务流时延情况。
另一方面， 基于本发明网络时延测量方法实施例十一， 本发明网络时延 测量方法实施例十三的方法， 还包括：
MCP接收上游 TLP对应的 DCP发送的发送端时延测量信息 , MCP接收 下游 TLP对应的 DCP发送的已确定与发送端时延测量信息属于同一测量周 期的接收端时延测量信息， MCP根据发送端时延测量信息和接收端时延测量 信息， 确定网络单次时延情况。
具体的， 对于没有釆用外部时间同步工具的网络来说， 为了保证发送端 时延测量信息与接收端时延测量信息属于同一测量周期， 既保证发送端时延 测量信息与接收端时延测量信息属于同一数据包。 根据本发明网络时延测量 方法实施例四和本发明网络时延测量方法实施例九提供的方法可知， 由上游 TLP、 下游 TLP以及管理下游 TLP的 DCP对发送端时延测量信息与接收端 时延测量信息进行对匹配， DCP将匹配一致的发送端时延测量信息与接收端 时延测量信息发送给 MCP, 本实施例 MCP也维护表 1 , MCP根据匹配一致 的发送端时延测量信息与接收端时延测量信息， 参照表 1 , 将发送端时延测 量信息与接收端时延测量信息更新到表 1对应数据包的数据表项中， 需要说 明的是， 本实施例 MCP维护的表 1 中， DCP可以将匹配一致的发送端时延 测量信息与接收端时延测量信息发送给 MCP, MCP将发送端时延测量信息 与接收端时延测量信息更新到相应的数据包的数据表项中，再通过 MCP进行 时延情况的确定。 DCP也可以根据匹配一致的发送端时延测量信息与接收端 时延测量信息确定了时延情况后， 由 DCP将已确定的时延情况发送给 MCP, 此时 MCP直接接收时延情况。
数据包对应的数据表项通过匹配一致的发送端时延测量信息与接收端时 延测量信息来确定，再由 MCP进行时延计算， 具体计算方法和公式在本发明 网络时延测量方法实施例十二中已经进行了详细说明， 此处不再赘述。
本实施例提供的网络时延测量方法 ,通过 MCP接收上游 TLP对应的 DCP 发送的发送端时延测量信息， MCP接收下游 TLP对应的 DCP发送的已确定 与发送端时延测量信息属于同一测量周期的接收端时延测量信息， MCP根据 发送端时延测量信息和接收端时延测量信息， 保证了在没有时间同步工具的 情况下， MCP准确的直接的测量网络业务流时延情况。
再一方面， 基于本发明网络时延测量方法实施例十一， 本发明网络时延 测量方法实施例十四的方法， 还包括：
MCP接收上游 TLP对应的 DCP发送的发送端时延测量信息 , 发送端时延 测量信息包括时戳信息、 业务流标识、 TLP标识、 发送端业务流特征信息以 及发送端分片重组标识。
MCP接收下游 TLP对应的 DCP发送的接收端时延测量信息 , 接收端时延 测量信息包括时戳信息、 业务流标识、 TLP标识、 接收端业务流特征信息以 及接收端分片重组标识。
MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流 特征信息以及接收端分片重组标识， 判断发送端时戳信息和接收端时戳信息 是否对应于同一业务流的时戳信息。若是， 则 MCP根据发送端时延测量信息 和接收端时延测量信息， 确定网络单次时延情况。
具体的， 配合网络时延测量方法实施例五和网络时延测量方法实施例十 的方法，本发明实施例 MCP可以根据发送端业务流特征信息、发送端分片重 组标识、 接收端业务流特征信息以及接收端分片重组标识， 判断发送端时戳 信息和接收端时戳信息是否属于一个数据包的时戳信息。若是， 则 MCP将该 发送端时戳信息和接收端时戳信息更新到表 1 中对应数据包的数据表项中， 若不是，则 MCP将该发送端时戳信息和接收端时戳信息分别更新到各自对应 的数据包的数据表项中。 参照表 1 , 对上述 MCP的操作过程进行详细说明。 当 MCP接收到和接收端时延测量信息后， 因为对于一个业务流来说， 其业务 流特征信息是独一无二的，所以 MCP根据发送端业务流特征信息，即五元组： 源 IP地址、 目的 IP地址、 协议类型、 源协议端口号、 目的协议端口号， 确 定该发送端时延测量信息属于哪一个业务流， 在确定该发送端时延测量信息 属于一个目标业务流后， MCP找到与该业务流对应的目标业务流测量数据汇 总表， 又因为发送端分片重组标识对于一个数据包也是独一无二的， 所以 MCP根据该发送端分片重组标识， 确定该发送端时延测量信息属于该目标业 务流的具体的一个数据包， 例如属于第 N个数据包， 于是 MCP将发送端时 延测量信息携带的发送端时戳信息更新到对应的第 N各数据包的数据表项 中。 对于接收端时延测量信息， MCP也釆用上述操作进行接收端时延测量信 息识别， 并将接收端时延测量信息识别更新到对应数据包的对应表项中， 之 后再由 MCP进行相应的时延计算，具体的计算方法和公式在本发明网络时延 测量方法实施例十二中已经进行了详细说明， 此处不再赘述。
本实施例提供的网络时延测量方法 ,通过 MCP接收上游 TLP对应的 DCP 发送的发送端时延测量信息， 发送端时延测量信息包括时戳信息、 业务流标 识、 TLP标识、 发送端业务流特征信息以及发送端分片重组标识， MCP接收 下游 TLP对应的 DCP发送的接收端时延测量信息， 接收端时延测量信息包 括时戳信息、 业务流标识、 TLP标识、 接收端业务流特征信息以及接收端分 片重组标识， 并通过 MCP根据发送端业务流特征信息、发送端分片重组标识 和接收端业务流特征信息、 接收端分片重组标识进行识别匹配， 从而保证了 发送端时延测量信息与接收端时延测量信息属于同一数据包， 再通过 MCP 对匹配的发送端时延测量信息和接收端时延测量信息进行时延计算， 实现了 对于业务流时延的直接准确的测量。
图 8为本发明 DCP实施例一的结构示意图， 如图 8所示， DCP包括： 获 取模块 10、 发送模块 12。
获取模块 10, 用于获取至少一个 TLP对业务流进行测量得到的时延测量 信息， 时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识。 发送模块 12, 用于将时延测量信息发送给测量控制点 MCP, 以使 MCP 根据时戳信息、 业务流标识以及 TLP标识， 确定网络时延情况。
具体的， 将 DCP部署在发送端设备和接收端设备上， 其工作原理和技 术方案在本发明网络时延测量方法实施例一已进行详细说明，此处不再赘述。
本实施例的 DCP, 可以用于执行图 1所示方法实施例的技术方案， 其实 现原理和技术效果类似， 此处不再赘述。
对于本实施例提供的 DCP, 对于 DCP为管理上游 TLP的 DCP, 获取模 块 10,具体用于获取至少一个上游 TLP对发送的业务流进行测量得到的发送 端时延测量信息，发送模块 12,具体用于将发送端时延测量信息发送给 MCP, 发送端时延测量信息包括发送端时戳信息、 业务流标识以及 TLP标识。 其中 具体的原理和方法在本发明网络时延测量方法实施例二已进行了详细说明， 此处不再赘述。
或者， DCP为管理下游 TLP的 DCP, 获取模块 10, 具体用于获取至少 一个下游 TLP对接收的业务流进行测量得到的接收端时延测量信息， 发送模 块 12, 具体用于将接收端时延测量信息发送给 MCP,接收端时延测量信息包 括接收端时戳信息、 业务流标识以及 TLP标识。 其中具体的原理和方法在本 发明网络时延测量方法实施例二已进行了详细说明， 此处不再赘述。
一方面， 在图 8的基础上， 图 9为本发明 DCP实施例二的结构示意图， 如图 9所示本发明 DCP实施例二的获取模块 20, 包括： 第一获取单元 200、 周期标识获取单元 202。
第一获取单元 200, 用于获取至少一个上游 TLP对发送的业务流进行测量 得到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务流 进行测量得到的接收端时延测量信息。
周期标识获取单元 202, 用于在管理上游 TLP的 DCP在测量周期结束时， 获取测量周期标识， 并将测量周期标识发送给 MCP, 或者， 在管理下游 TLP 的 DCP在测量周期开始时， 周期标识获取单元 202根据 NTP或 IEEE 1588v2时 钟就可以得到每个测量周期的边界时间点， 即获取该测量周期的起始时间， 若起始时间与时戳信息之差小于等于预设时长， 则收端时延测量信息属于测 量周期标识对应的测量信息； 若起始时间与时戳信息之差大于预设时长， 则 将测量周期标识加 1 , 时戳信息属于下一个测量周期， 获取管理下游 TLP的 DCP在该测量周期的测量周期标识。 可选的， 上述预设时长为测量周期的 2/3 时长。 具体的， 其具体的工作原理和方法在本发明网络时延测量方法实施例 三以及图 2中已进行了详细说明， 此处不再赘述。
可选的，对应的 DCP可以直接读取上游 TLP和下游 TLP产生的发送端时延 测量信息和接收端时延测量信息， 由 DCP的周期标识获取单元 202根据发送端 时延测量信息和接收端时延测量信息分别获得两个测量周期标识， 并且， 对 于一个周期内被添加时延测量标志的一个数据包来说， 管理上游 TLP的 DCP 以及管理下游 TLP的 DCP获取的两个测量周期标识是一致的。
如图 9所示， 本发明 DCP实施例二发送模块 22, 包括： 第一发送单元 220、 第二发送单元 222。
第一发送单元 220, 将发送端时延测量信息发送给 MCP, 或者， 将接收端 时延测量信息发送给 MCP。
第二发送单元 222, 将管理上游 TLP的 DCP在测量周期结束时将周期标识 获取单元 202获取的测量周期标识发送给 MCP, 或者， 将管理下游 TLP的 DCP 的周期标识获取单元 202获取的测量周期标识发送给 MCP。
可选的， 本发明 DCP实施例二的 DCP , 还包括： 时间同步模块 24。 时间同步模块 24, 用于在获取模块 20获取至少一个 TLP对业务流进行 测量得到的时延测量信息之前，釆用 NTP或 IEEE 1588v2时钟与 TLP进行时 间同步， 以及釆用 NTP或 IEEE 1588v2时钟管理上游 TLP的 DCP与管理下 游 TLP的 DCP进行时间同步。 对于 NTP或 IEEE 1588v2时钟以及时间同步 的方和原理， 在本发明网络时延测量方法实施例三已进行了详细说明， 此处 不再赘述。
本实施例的 DCP, 可以用于执行本发明网络时延测量方法实施例三的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
另一方面， 在图 8的基础上， 图 10为本发明 DCP实施例三的结构示意 图， 如图 8所示， 包括： 获取模块 30、 确定模块 32、 发送模块 34。 其中获 取模块 30, 包括： 第二获取单元 300、 测量报文获取单元 302。
第二获取单元 300, 用于获取至少一个上游 TLP对发送的业务流进行测量 得到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务流 进行测量得到的接收端时延测量信息。 具体的， 发送端时延测量信息包括： 发送端时戳信息、 业务流标识以及
TLP标识， 接收端时延测量信息包括： 接收端时戳信息、 业务流标识以及 TLP 标识。
测量报文获取单元 302, 用于获取至少一个下游 TLP接收的由上游 TLP发 送的测量报文以及测量报文到达下游 TLP时产生的一个测量报文的到达时戳 信息， 测量报文包括： 发送端时戳信息。
确定模块 32 , 具体用于确定到达时戳信息和接收端时戳信息是否同属于 预设时长范围， 若是， 则确定发送端时戳信息和接收端时戳信息属于同一数 据包。 具体的确定原理和方法， 在本发明网络时延测量方法实施例四进行了 详细说明， 此处不再赞述。
发送模块 34, 具体用于确定结果发送给 MCP。
需要说明的是，确定模块 32可以将属于同一数据包的发送端时戳信息和 接收端时戳信息发送给发送模块 34, 通过发送模块 34发送给 MCP, 由 MCP 来进行时延计算。也可以由该确定模块 32根据属于同一测量周期的发送端时 戳信息和接收端时戳信息直接确定该周期数据包的时延情况， 再将计算好的 时延情况发送给 MCP。
本实施例的 DCP, 可以用于执行本发明网络时延测量方法实施例四的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
再一方面， 参照图 8, 本发明 DCP实施例四的 DCP包括： 获取模块 10、 发送模块 12。
获取模块 10, 当 DCP为管理上游 TLP的数据收集点时，该获取模块 10, 具体用于获取发送端时延测量信息， 发送端时延测量信息还包括： 发送端业 务流特征信息以及发送端分片重组标识。 当 DCP为管理下游 TLP的数据收 集点时， 该获取模块 10具体用于获取接收端时延测量信息， 接收端时延测量 信息还包括： 接收端业务流特征信息以及接收端分片重组标识。
发送模块 12, 当 DCP为管理上游 TLP的数据收集点时，该发送模块 12, 具体用于将发送端时延测量信息给 MCP, 以使 MCP根据发送端业务流特征 信息、 发送端分片重组标识、 接收端业务流特征信息以及接收端分片重组标 识，确定发送端时戳信息和接收端时戳信息是对应于同一业务流的时戳信息。
具体的， 对于业务流特征信息、 分片重组标识， 以及 DCP基于业务流特 征信息、 分片重组标识进行的相应操作， 本发明网络时延测量方法实施例五 已进行了详细说明， 此处不再赘述。
本实施例的 DCP, 可以用于执行本发明网络时延测量方法实施例五的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
图 11为本发明 TLP实施例一的结构示意图， 如图 11所示， TLP包括： 识别模块 40、 时戳获取模块 42、 确定模块 44。
识别模块 40, 用于根据业务流特征信息对业务流进行识别， 确定业务流 是否是目标业务流。
具体的， 对于如何根据业务流特征信息对业务流进行识别， 在本发明网 络时延测量方法实施例六已进行了详细说明， 此处不再赘述。
时戳获取模块 42, 用于若是， 则对业务流的数据包添加时延测量标志， 获取与该时延测量标志对应的数据包时延测量信息。
具体的，时戳获取模块 42对目标业务流的数据包添加一个时延测量标志， 并且生成时延测量信息，该时延测量信息包括时戳信息、业务流标识以及 TLP 标识， 其中， 时戳信息就是 TLP添加时延测量标志的时间点。 对于业务流标 识以及 TLP标识， 网络时延测量方法实施例一已进行了详细说明， 此处不再 赘述。
确定模块 44, 用于确定时延测量信息， 时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识， 以使 DCP在获取时延测量信息后将时延测量信 息发送给 MCP。
本实施例的 TLP, 可以用于执行本发明网络时延测量方法实施例六的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
对于本实施例提供的 TLP, 可选的， 对于时戳获取模块 42对业务流的数 据包添加时延测量标志，获取与该时延测量标志对应的数据包时延测量信息， 包括：
上游 TLP的时戳获取模块 42, 具体用于在目标业务流的数据包上添加时 延测量标志， 该时戳获取模块获取与该时延测量标志对应的数据包发送端时 延测量信息， 发送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 以使管理上游 TLP的 DCP获取发送端时延测量信息后将发送端时延测量信息 发送给 MCP。 下游 TLP的时戳获取模块 42,具体用于在识别模块识别到添加时延测量 标志的数据包时， 该时戳获取模块获取与该时延测量标志对应的数据包接收 端端时延测量信息。接收端时延测量信息包括时戳信息、业务流标识以及 TLP 标识， 以使管理下游 TLP的 DCP获取接收端时延测量信息后将接收端时延 测量信息发送给 MCP。
一方面，在图 11的基础上， 图 12为本发明 TLP实施例二的结构示意图， 如图 12所示， 还包括： 时间同步模块 41、 测量周期标识获取模块 43。
时间同步模块 41 , 当 TLP为上游 TLP, 具体用于在上游 TLP的时戳获 取模块 42在目标业务流的数据包上添加时延测量标志之前，釆用 NTP或 IEEE 1588v2时钟与管理上游 TLP的 DCP进行时间同步。 当 TLP为下游 TLP , 该 时间同步模块 41 , 具体用于在下游 TLP的识别模块 40在识别到添加时延测 量标志的数据包之前，釆用 NTP或 IEEE 1588v2时钟与管理下游 TLP的 DCP 进行时间同步。 时间同步的方法和原理， 本发明网络时延测量方法实施例三 的方法已进行了详细说明， 此处不再赘述。
测量周期标识获取模块 43 , 当 TLP为上游 TLP, 具体用于获取与时延测 量标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取测量周期标识 后将测量周期标识信息发送给 MCP; 当 TLP为下游 TLP,具体用于获取与时 延测量标志对应的测量周期标识以及每个测量周期起始时间， 以使管理下游 TLP的 DCP获取起始时间和测量周期标识后进行匹配确定，再将接收端时延 测量信息发送给 MCP。
需要说明的是， DCP可以直接读取 TLP产生的时延测量信息， 由 DCP 通过时延测量信息得到相应的测量周期标识， 在这个方案中， 上游 TLP或者 下游 TLP可以不使用测量周期标识获取模块 43。 并且，对于本实施例提供的 TLP, TLP 在基于每一个周期对业务流数据包进行识别， 并添加时延测量标 志， 可选的， 只在每一个测量周期内， 对一个数据包进行相应操作。
本实施例的 TLP, 可以用于执行本发明网络时延测量方法实施例八的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
另一方面， 在图 11的基础上， 图 13为本发明 TLP实施例三的结构示意 图， 如图 13所示， 还包括： 发送模块 46、 接收模块 48。
发送模块 46, 具体用于由上游 TLP的发送模块向下游 TLP发送测量报文， 该测量报文包括： 发送端时戳信息。
具体的， 当上游 TLP的识别模块 40对目标业务流的数据包添加时延测量 标志后， 由时戳获取模块 42对该数据包添加时延测量标志， 获取与该时延测 量标志对应的数据包时延测量信息， 此时上游 TLP启动发送模块 46, 并由发 送模块 46向下游 TLP发送测量报文。
接收模块 48, 具体用于下游 TLP的接收模块接收到测量报文， 产生一个 测量报文的到达时戳信息， 并将测量报文以及到达时戳信息发送给管理下游 TLP的 DCP,以使 DCP确定到达时戳信息和接收端时戳信息是否同属于预设 时长范围， 若是， 则确定发送端时戳信息和接收端时戳信息属于同一数据包， 并将确定结果发送给 MCP。
具体的， 下游 TLP的识别模块 40识别具有时延测量标志的数据包并产 生相应的接收端时戳信息， 接收模块 48接收到测量报文， 记录接收到测量报 文的时间点， 生成到达时戳信息， 将该测量报文以及到达时戳信息发送给确 定模块 44, 以使确定模块 48将该测量报文、 接收端时延测量信息以及到达 时戳信息发送给管理下游 TLP的 DCP, 使管理下游 TLP的 DCP进行相应操 作。
并且， 对于本实施例提供的网络时延测量方法， TLP对业务流数据包进 行识别，在每一个周期 TLP对该测量周期区间内的数据包添加时延测量标志， 可选的， 在每一个测量周期内， TLP只对一个数据包添加时延测量标志。
本实施例的 TLP, 可以用于执行本发明网络时延测量方法实施例九的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
再一方面 , 参照图 11 , 本发明 TLP实施例四 , 还包括：
上游 TLP的时戳获取模块 42获取发送端时延测量信息还包括： 发送端业 务流特征信息以及发送端分片重组标识； 下游 TLP的时戳获取模块 42获取接 收端时延测量信息还包括：接收端业务流特征信息以及接收端分片重组标识； 以使管理上游 TLP的 DCP获取发送端时延测量信息并发送给 MCP, 管 理下游 TLP的 DCP获取接收端时延测量信息并发送给 MCP,以使 MCP根据 发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及 接收端分片重组标识， 确定发送端时戳信息和接收端时戳信息是对应于同一 业务流的时戳信息。 本实施例的 TLP, 可以用于执行本发明网络时延测量方法实施例十的技 术方案， 其实现原理和技术效果类似， 此处不再赘述。
可选的， 对于本发明 TLP实施例一〜实施例四， 对于上游 TLP的时戳获 取模块在目标业务流的数据包上添加时延测量标志， 包括： 时戳获取模块在 数据包的 IP头中 TOS的保留位或者 Flags的保留位上添加时延测量标志，这 样可以利用 TOS的保留位或者 Flags的保留位对数据包添加时延测量标志， 而且还能保证数据包的正常传输。
并且， 对于识别模块 60, 具体用于根据五元组中的至少两元信息， 对业 务流进行识别， 以保证对于业务流的有效识别。
具体的， 对于 TOS的保留位或者 Flags的保留位， 以及五元组， 在本发 明网络时延测量方法实施例七中进行了详细的说明， 此处不再赘述。
图 14为本发明 MCP实施例一的结构示意图， 如图 14所示， MCP包括： 接收模块 70、 确定模块 72。
接收模块 70, 用于接收上游 TLP对应的 DCP发送的发送端时延测量信息 以及与下游 TLP对应的 DCP发送的接收端时延测量信息， 发送端时延测量信 息包括时戳信息、 业务流标识以及 TLP标识， 接收端时延测量信息包括时戳 信息、 业务流标识以及 TLP标识；
确定模块 72, 用于根据发送端时延测量信息和接收端时延测量信息， 确 定网络单次时延情况。
具体的， MCP部署在网络的任一网元节点设备上， 可选的， 部署在一个 功能较强的网元节点设备上。由确定模块 72维护一个目标业务流测量数据汇 总表， 参照上文表 1。 确定模块 72依据的计算原理和公式， 在本发明网络时 延测量方法实施例十二已进行了详细说明， 此处不再赘述。
本实施例的 MCP, 可以用于执行本发明网络时延测量方法实施例十一的 技术方案， 其实现原理和技术效果类似， 此处不再赘述。
一方面，在图 14的基础上，图 15为本发明 MCP实施例二的结构示意图， 包括：接收模块 70、确定模块 72,其中接收模块 70包括： 第一接收单元 700、 第二接收单元 702。确定模块 72包括：第一匹配单元 720、第一确定单元 722。
第一接收单元 700,具体用于接收上游 TLP对应的 DCP发送的发送端时 延测量信息以及与下游 TLP对应的 DCP发送的接收端时延测量信息， 发送 端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 接收端时延测量 信息包括时戳信息、 业务流标识以及 TLP标识。
第二接收单元 702,具体用于接收管理上游 TLP的 DCP发送的测量周期 标识， 以及接收管理下游 TLP的 DCP发送的测量周期标识。
第一匹配单元 720, 具体用于根据管理上游 TLP的 DCP发送的测量周期标 识和管理下游 TLP的 DCP发送的测量周期标识， 判断发送端时延测量信息和 接收端时延测量信息是否属于同一测量周期。
具体的， 第一匹配单元 720接收到由第二接收单元 702接收的测量周期标 识后，根据管理上游 TLP的 DCP发送的测量周期标识和管理下游 TLP的 DCP发 送的测量周期标识， 将属于同一测量周期的发送端时延测量信息和接收端时 延测量信息更新到 MCP维护的目标业务流测量数据汇总表的对应表项中。
第一确定单元 722, 具体用于， 若是， 则根据发送端时延测量信息和接收 端时延测量信息， 确定网络单次时延情况。
具体的当第一确定单元 722检测到表 1中一个测量周期标识中一个正向业 务流标识或者一个反向业务流对应的数据项的数据到齐标志设置为 "全部到 齐"后， 此时第一确定单元 722将根据对应的发送端时戳信息和接收端时戳信 息进行时延计算。 具体的计算原理和公式在本发明网络时延测量方法实施例 十二已进行了详细说明， 此处不再赘述。
本实施例的 MCP, 可以用于执行本发明网络时延测量方法实施例十二的 技术方案， 其实现原理和技术效果类似， 此处不再赘述。
另一方面， 参见图 12, 本发明 MCP实施例三的 MCP 包括： 接收模块 70、 确定模块 72。
接收模块 70, 具体用于接收上游 TLP对应的 DCP发送的发送端时延测量 于同一数据包的接收端时延测量信息。
确定模块 72,具体用于根据发送端时延测量信息和接收端时延测量信息， 确定网络单次时延情况。
具体的， DCP可以将匹配一致的发送端时延测量信息与接收端时延测量 信息发送给 MCP时， 由 MCP的接收模块 70接收并将发送端时延测量信息 与接收端时延测量信息更新到相应的数据包的数据表项中， 再通过确定模块 72进行时延情况的确定。 DCP也可以根据匹配一致的发送端时延测量信息与 接收端时延测量信息确定了时延情况后， 由 DCP将已确定的时延情况发送给 MCP,此时 MCP的接收模块 70直接接收时延情况，不需要启动确定模块 72。
本实施例的 MCP, 可以用于执行本发明网络时延测量方法实施例十三的 技术方案， 其实现原理和技术效果类似， 此处不再赘述。
再一方面，在图 14的基础上， 图 16为本发明 MCP实施例四的结构示意 图， 如图 16所示， 包括： 接收模块 70、 确定模块 72。
接收模块 70, 具体用于接收上游 TLP对应的 DCP发送的发送端时延测 量信息， 发送端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 发送 端业务流特征信息以及发送端分片重组标识， 以及接收下游 TLP对应的 DCP 发送的接收端时延测量信息， 接收端时延测量信息包括时戳信息、 业务流标 识、 TLP标识、 接收端业务流特征信息以及接收端分片重组标识。
确定模块 72包括： 第二匹配单元 721、 第二确定单元 723。
第二匹配单元 721 , 具体用于根据发送端业务流特征信息、发送端分片重 组标识、 接收端业务流特征信息以及接收端分片重组标识， 判断发送端时戳 信息和接收端时戳信息是否对应于同一数据包的时戳信息。 具体工作原理和 方法在本发明网络时延测量方法实施例五、 实施例十以及实施例十四已经进 行了详细说明， 此处不再赘述。
第二确定单元 723 , 具体用于， 若是， 则根据发送端时延测量信息和接 收端时延测量信息， 确定网络单次时延情况。 具体工作原理和方法在本发明 网络时延测量方法实施例五、实施例十以及实施例十四已经进行了详细说明， 此处不再赘述。
本实施例的 MCP , 可以用于执行本发明网络时延测量方法实施例十四的 技术方案， 其实现原理和技术效果类似， 此处不再赘述。
图 17为本发明网络时延测量系统实施例一的结构示意图，如图 17所示， 本实施例的系统包括： DCP、 TLP以及 MCP, 可选的， 在网络两侧以任意个 数进行 TLP的部署， 进而相应的部署 DCP, 本发明实施例对 TLP 以及相应 DCP的个数不做限定。 其中， DCP可以釆用图 8〜图 10的结构其对应地， 可 以执行本发明网络时延测量方法实施例一〜实施例五的技术方案； TLP可以釆 用图 11〜图 13的结构其对应地， 可以执行本发明网络时延测量方法实施例六 〜实施例十的技术方案； MCP可以釆用图 14〜图 16的结构其对应地， 可以执 行本发明提供的时延测量方法实施例十一〜实施例十四的技术方案，其实现原 理和技术效果类似， 此处不再赘述。
图 18为本发明网络时延测量系统实施例二的示意图， 参照图 17、 图 18, 下面对于本发明提供的网络时延测量方法、 装置和系统进行整体的说明。
参照图 18可知，在上游发送端和下游接收端的网络侧部署 TLP,可选的， 也可以将 TLP部署在用户侧， 在上游发送端和下游接收端的设备上分别部署 DCP,参照图 18,当目标业务流方向为由左向右时，基站侧网关 CSG( Cell Site Gateway,以下简称 CSG ) 为上游发送端设备， 无线网络控制器侧网关 RSG ( Radio Network Controller Site Gateway, 以下简称 RSG ) 1、 RSG2为下游接 收端设备， 当目标业务流方向相反时， 则对于上游下游的进行反向设置即可， 其具体的技术方案在本发明网络时延测量方法实施例十二中已进行了相信说 明， 此处不再赘述。 并在整个网络中任意一个网元节点上部署 MCP, 例如图 18中选择在 RSG1上部署 MCP。 可选的， 将 MCP部署在一个功能较强的网 元节点上， 并且时延测量信息的发送路径与目标业务流的发送路径（如图 18 中以实线表示） 区分开进行了外带传送， 保证了可选的时延测量信息读取和 发送的独立性。 管理网络（如图 18中， 管理网络层的路径以虚线表示）可以 釆用三层测量 VPN、 DCN或者具备 IP可达的公网。
当网络时延使能时， 各 TLP、 各 DCP、 MCP配合对业务流进行直接时延 测量， 其具体方法、技术方案已在本发明网络时延测量方法实施例一〜实施例 十四， 本发明 DCP实施例一〜实施例四、 本发明 TLP实施例一〜实施例四、 本发明 MCP实施例一〜实施例四中已进行了详细说明 , 此处不再赘述。
对于某些网络， 其中包含由二层 VPN网络和三层 VPN网络混合而成， 由 于二层 VPN网络和三层 VPN网络的测量基准不同， 现有技术对于这种网络场 景的还没有一种行之可效的时延测量方法， 由实施例一可知， 本发明实施例 中的各 TLP、 各 DCP、 MCP之间时延测量信息的获取和发送通过在管理网络 发送实现了外带传送， 有效避免了数据包测量信息随业务流发送时， 由于二 层 VPN网络与三层 VPN网络的测量基准不同而导致时延测量信息的问题。
参照图 18, 网络中对于网络右侧的 RSG1、 RSG2设备存在的双规接入场 景， 当 RSG1、 RSG2发生路径切换时， 由于本发明实施例中提供的时延测量 信息包括时戳信息、 业务流标识以及 TLP标识， 可以在 RSG1、 RSG2发生切 换时，有新的接收设备上的 TLP对目标业务流的数据包进行时延测量。例如， 目标业务流从左侧用户出发， 由 CSG上的 TLP进行识别， 使能网络时延测 量， 随后该目标业务流经网络至 RSG1的 TLP, 由该 TLP进行相应的接收端 时延测量。 当 RSG1故障时， 目标业务流切换至 RSG2上， 此时 RSG2上的 TLP可以识别该目标业务流并且继续进行相应的时延测量。
本领域普通技术人员可以理解： 实现上述各方法实施例的全部或部分 S 可以通过程序指令相关的硬件来完成。 前述的程序可以存储于一计算机可读 取存储介质中。 该程序在执行时， 执行包括上述各方法实施例的 S; 而前述的 存储介质包括： ROM, RAM,磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是： 以上各实施例仅用以说明本发明的技术方案， 而非对 其限制； 尽管参照前述各实施例对本发明进行了详细的说明， 本领域的普通 技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改， 或者对其中部分或者全部技术特征进行等同替换； 而这些修改或者替换， 并 不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims

权 利 要 求 书
1、 一种网络时延测量方法， 其特征在于， 包括：
获取至少一个目标逻辑端口 TLP对业务流进行测量得到的时延测量信 息， 所述时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识；
将所述时延测量信息发送给测量控制点 MCP, 以使所述 MCP根据所述 时戳信息、 业务流标识以及 TLP标识， 确定网络时延情况。
2、 根据权利要求 1所述的方法， 其特征在于， 所述获取至少一个 TLP 对业务流进行测量得到的时延测量信息， 包括：
管理上游 TLP的数据收集点 DCP获取至少一个上游 TLP对发送的业务 流进行测量得到的发送端时延测量信息；
管理下游 TLP的 DCP获取至少一个下游 TLP对接收的业务流进行测量 得到的接收端时延测量信息；
所述将所述时延测量信息发送给 MCP, 包括：
所述管理上游 TLP的 DCP将所述发送端时延测量信息发送给 MCP, 所 述发送端时延测量信息包括发送端时戳信息、 业务流标识以及 TLP标识； 所述管理下游 TLP的 DCP将所述接收端时延测量信息发送给 MCP, 所 述接收端时延测量信息包括接收端时戳信息、 业务流标识以及 TLP标识。
3、 根据权利要求 2所述的方法， 其特征在于， 还包括：
所述管理上游 TLP的 DCP在测量周期结束时， 获取测量周期标识， 并 将所述测量周期标识发送给所述 MCP;
所述管理下游 TLP的 DCP获取该测量周期的起始时间， 若所述起始时 间与所述时戳信息之差小于等于预设时长， 则所述接收端时延测量信息属 于所述测量周期标识对应的测量信息； 若所述起始时间与所述时戳信息之 差大于预设时长，则将测量周期标识加 1 ,所述时戳信息属于下一个测量周 期， 将所述测量周期标识发送给所述 MCP。
4、 根据权利要求 3所述的方法， 其特征在于， 所述预设时长为所述测 量周期的 2/3时长。
5、 根据权利要求 3所述的方法， 其特征在于， 还包括：
所述管理上游 TLP的 DCP釆用网络时间协议 NTP或 IEEE 1588v2时钟与 所述上游 TLP进行时间同步， 所述管理下游 TLP的 DCP釆用 NTP或 IEEE 1588v2时钟与所述下游 TLP进行时间同步， 并且所述管理上游 TLP的 DCP 现时间同步。
6、 根据权利要求 2所述的方法， 其特征在于， 还包括：
所述管理下游 TLP的 DCP获取至少一个下游 TLP接收的由所述上游 TLP发送的测量 4艮文以及所述测量 4艮文到达所述下游 TLP时产生的一个测 量报文的到达时戳信息， 所述测量报文包括： 发送端时戳信息；
所述管理下游 TLP的 DCP确定所述到达时戳信息和所述接收端时戳信 息是否同属于预设时长范围， 若是， 则确定所述发送端时戳信息和所述接 收端时戳信息属于同一数据包， 并将确定结果发送给所述 MCP。
7、 根据权利要求 2所述的方法， 其特征在于， 所述发送端时延测量信 息还包括： 发送端业务流特征信息以及发送端分片重组标识， 所述接收端 时延测量信息还包括： 接收端业务流特征信息以及接收端分片重组标识， 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端 业务流特征信息以及接收端分片重组标识， 确定所述发送端时戳信息和所 述接收端时戳信息是对应于同一业务流的时戳信息。
8、 一种网络时延测量方法， 其特征在于， 包括：
根据业务流特征信息对业务流进行识别， 确定所述业务流是否是目标 业务流；
若是， 则对所述业务流的数据包添加时延测量标志， 获取与该时延测 量标志对应的数据包时延测量信息；
确定时延测量信息， 所述时延测量信息包括： 时戳信息、 业务流标识 以及目标逻辑端口 TLP标识， 以使所述数据收集点 DCP在获取所述时延测 量信息后将所述时延测量信息发送给测量控制点 MCP。
9、 根据权利要求 8所述的方法， 其特征在于， 对所述业务流的数据包 添加时延测量标志， 获取与该时延测量标志对应的数据包时延测量信息， 包括：
上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与该时延 测量标志对应的数据包发送端时延测量信息， 所述发送端时延测量信息包 括时戳信息、业务流标识以及 TLP标识， 以使管理上游 TLP的 DCP获取所述 发送端时延测量信息后将所述发送端时延测量信息发送给 MCP;
下游 TLP在识别到所述添加时延测量标志的数据包时， 获取与该时延 测量标志对应的数据包接收端端时延测量信息。 所述接收端时延测量信息 包括时戳信息、业务流标识以及 TLP标识 , 以使管理下游 TLP的 DCP获取所 述接收端时延测量信息后将所述接收端时延测量信息发送给 MCP。
10、 根据权利要求 9所述的方法， 其特征在于， 还包括：
所述上游 TLP在目标业务流的数据包上添加时延测量标志之前， 釆用 游 TLP在识别到所述添加时延测量标志的数据包之前， 釆用网络时间协议 对所述业务流的数据包添加时延测量标志， 获取与该时延测量标志对 应的数据包时延测量信息， 还包括：
所述上游 TLP在目标业务流的数据包上添加时延测量标志， 获取与所 述时延测量标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取所述 测量周期标识后将所述测量周期标识发送给 MCP;
所述下游 TLP在每个测量周期获取该测量周期起始时间， 在每个测量 周期内识别到所述添加时延测量标志的数据包时， 获取与所述时延测量标 志对应的测量周期标识， 以使管理下游 TLP的 DCP获取所述起始时间和所 述测量周期标识后将所述测量周期标识发送给 MCP。
11、 根据权利要求 9所述的方法， 其特征在于， 还包括：
由所述上游 TLP向下游 TLP发送测量 4艮文， 所述测量 ^艮文包括： 发送 端时戳信息；
所述下游 TLP的接收模块接收到所述测量报文，产生一个测量报文的 到达时戳信息， 并将所述测量报文、 以及所述到达时戳信息发送给所述管 理下游 TLP的 DCP, 以使所述 DCP确定所述到达时戳信息和所述接收端 时戳信息是否同属于预设时长范围， 若是， 则确定所述发送端时戳信息和 所述接收端时戳信息属于同一数据包， 并将确定结果发送给所述 MCP。
12、根据权利要求 9所述的方法， 其特征在于， 所述发送端时延测量 信息还包括： 发送端业务流特征信息以及发送端分片重组标识； 所述接收 端时延测量信息还包括:接收端业务流特征信息以及接收端分片重组标识； 以使所述管理上游 TLP的 DCP获取所述发送端时延测量信息并发送给 所述 MCP , 所述管理下游 TLP的 DCP获取所述接收端时延测量信息并发送 给所述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组 标识、 接收端业务流特征信息以及接收端分片重组标识， 确定所述发送端 时戳信息和所述接收端时戳信息是对应于同一数据包的时戳信息。
13、 根据权利要求 8〜12中任一项所述的方法， 其特征在于， 所述上游 TLP在目标业务流的数据包上添加时延测量标志， 包括：
在所述数据包的 IP头中服务类型 TOS的保留位或者 Flags的保留位上添 加时延测量标志。
14、 根据权利要求 8〜12中任一项所述的方法， 其特征在于， 所述根据 业务流特征信息对业务流进行识别， 包括：
根据五元组中的至少两元信息， 对所述业务流进行识别。
15、 一种网络时延测量方法， 其特征在于， 包括：
接收上游目标逻辑端口 TLP对应的数据收集点 DCP发送的发送端时延 测量信息以及与下游 TLP对应的 DCP发送的接收端时延测量信息， 所述发 送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 所述接收端 时延测量信息包括时戳信息、 业务流标识以及 TLP标识；
根据所述发送端时延测量信息和所述接收端时延测量信息， 确定网络 单次时延情况。
16、 根据权利要求 15所述的方法， 其特征在于， 还包括：
测量控制点 MCP接收所述管理上游 TLP的 DCP发送的测量周期标识， 所述 MCP接收所述管理下游 TLP的 DCP发送的测量周期标识， 由所述 MCP 根据所述管理上游 TLP的 DCP发送的测量周期标识和所述管理下游 TLP的 DCP发送的测量周期标识， 判断所述发送端时延测量信息和所述接收端时 延测量信息是否属于同一测量周期， 若是， 则所述 MCP根据所述发送端时 延测量信息和所述接收端时延测量信息， 确定网络单次时延情况。
17、 根据权利要求 15所述的方法， 其特征在于， 还包括：
量信息属于同一数据包的所述接收端时延测量信息， 所述 MCP根据所述发 送端时延测量信息和所述接收端时延测量信息， 确定网络单次时延情况。
18、 根据权利要求 15所述的方法， 其特征在于， 还包括：
MCP接收所述上游 TLP对应的 DCP发送的发送端时延测量信息， 所述 发送端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 发送端业务 流特征信息以及发送端分片重组标识；
所述 MCP接收所述下游 TLP对应的 DCP发送的接收端时延测量信息， 所述接收端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 接收端 业务流特征信息以及接收端分片重组标识；
所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端 业务流特征信息以及接收端分片重组标识， 判断所述发送端时戳信息和所 述接收端时戳信息是否对应于同一数据包的时戳信息； 若是， 则所述 MCP 根据所述发送端时延测量信息和所述接收端时延测量信息， 确定网络单次 时延情况。
19、 一种数据收集点 DCP, 其特征在于， 包括：
获取模块， 用于获取至少一个目标逻辑端口 TLP对业务流进行测量得 到的时延测量信息， 所述时延测量信息包括： 时戳信息、 业务流标识以及 TLP标识
发送模块， 用于将所述时延测量信息发送给测量控制点 MCP, 以使所 述 MCP根据所述时戳信息、 业务流标识以及 TLP标识，确定网络时延情况。
20、 根据权利要求 19所述的 DCP, 其特征在于， 所述 DCP为管理上游 TLP^DCP;
所述获取模块， 具体用于获取至少一个上游 TLP对发送的业务流进行 测量得到的发送端时延测量信息；
或者，
所述 DCP为管理下游 TLP的 DCP；
所述获取模块， 具体用于获取至少一个下游 TLP对接收的业务流进行 测量得到的接收端时延测量信息；
所述 DCP将时延测量信息发送给 MCP, 包括：
所述 DCP为管理上游 TLP的 DCP；
所述发送模块， 具体用于将所述发送端时延测量信息发送给 MCP, 所 述发送端时延测量信息包括发送端时戳信息、 业务流标识以及 TLP标识； 或者，
所述 DCP为管理下游 TLP的 DCP；
所述发送模块， 具体用于将所述接收端时延测量信息发送给 MCP, 所 述接收端时延测量信息包括接收端时戳信息、 业务流标识以及 TLP标识。
21、 根据权利要求 20所述的 DCP, 其特征在于， 所述获取模块包括： 第一获取单元， 用于获取至少一个上游 TLP对发送的业务流进行测量 得到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务 流进行测量得到的接收端时延测量信息；
周期标识获取单元， 用于在所述管理上游 TLP的 DCP在测量周期结束 时， 获取测量周期标识， 并将所述测量周期标识发送给所述 MCP, 或者， 在所述管理下游 TLP的 DCP在测量周期开始时， 所述周期标识获取单元获 取该测量周期的起始时间， 若所述起始时间与所述时戳信息之差小于等于 预设时长， 则所述接收端时延测量信息属于所述测量周期标识对应的测量 信息； 若所述起始时间与所述时戳信息之差大于预设时长， 则将测量周期 标识加 1 , 所述时戳信息属于下一个测量周期， 获取管理下游 TLP的 DCP在 该测量周期的测量周期标识；
所述发送模块包括：
第一发送单元， 将所述发送端时延测量信息发送给 MCP, 或者， 将所 述接收端时延测量信息发送给所述 MCP；
第二发送单元， 将所述管理上游 TLP的 DCP在测量周期结束时将周期 标识获取单元获取的所述测量周期标识发送给所述 MCP, 或者， 将所述管 理下游 TLP的 DCP的周期标识获取单元获取的所述测量周期标识发送给 MCP。
22、 根据权利要求 21所述的 DCP, 其特征在于， 所述预设时长为所述 测量周期的 2/3时长。
23、 根据权利要求 21所述的 DCP, 其特征在于， 还包括：
时间同步模块， 用于在所述获取模块获取至少一个 TLP对业务流进行 测量得到的时延测量信息之前，釆用网络时间协议 NTP或 IEEE 1588v2时钟 与所述 TLP进行时间同步， 以及釆用所述 NTP或 IEEE 1588v2时钟所述管理
24、 根据权利要求 20所述的 DCP, 其特征在于， 所述获取模块包括： 第二获取单元， 用于获取至少一个上游 TLP对发送的业务流进行测量 得到的发送端时延测量信息， 或者， 获取至少一个下游 TLP对发送的业务 流进行测量得到的接收端时延测量信息；
测量报文获取单元， 用于获取至少一个下游 TLP接收的由所述上游 TLP发送的测量 4艮文以及所述测量 4艮文到达所述下游 TLP时产生的一个测 量报文的到达时戳信息， 所述测量报文包括： 发送端时戳信息；
所述确定模块， 具体用于确定所述到达时戳信息和所述接收端时戳信 息是否同属于预设时长范围， 若是， 则确定所述发送端时戳信息和所述接 收端时戳信息属于同一数据包；
所述发送模块， 具体用于确定结果发送给所述 MCP。
25、 根据权利要求 20所述的 DCP, 其特征在于， 所述 DCP为管理上游 TLP的数据收集点；
所述获取模块， 具体用于获取所述发送端时延测量信息， 所述发送端 时延测量信息还包括： 发送端业务流特征信息以及发送端分片重组标识； 所述发送模块， 具体用于将所述发送端时延测量信息给所述 MCP; 所述 DCP为管理下游 TLP的数据收集点；
所述获取模块， 具体用于获取所述接收端时延测量信息， 所述接收端 时延测量信息还包括： 接收端业务流特征信息以及接收端分片重组标识； 所述发送模块， 具体用于将所述接收端时延测量信息发送给所述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组标识、 接收端业务流特征信息以及接收端分片重组标识， 确定所述发送端时戳信 息和所述接收端时戳信息是对应于同一业务流的时戳信息。
26、 一种目标逻辑端口 TLP, 其特征在于， 包括：
识别模块， 用于根据业务流特征信息对业务流进行识别， 确定所述业 务流是否是目标业务流；
时戳获取模块， 用于若是， 则对所述业务流的数据包添加时延测量标 志， 获取与该时延测量标志对应的数据包时延测量信息；
确定模块， 用于确定时延测量信息， 所述时延测量信息包括： 时戳信 息、 业务流标识以及 TLP标识， 以使数据收集点 DCP在获取所述时延测量 信息后将所述时延测量信息发送给测量控制点 MCP。
27、 根据权利要求 26所述的 TLP, 其特征在于， 所述时戳获取模块对 所述业务流的数据包添加时延测量标志， 获取与该时延测量标志对应的数 据包时延测量信息， 包括：
上游 TLP的所述时戳获取模块， 具体用于在目标业务流的数据包上添 加时延测量标志， 获取与该时延测量标志对应的数据包发送端时延测量信 息， 所述发送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 以使管理上游 TLP的 DCP获取所述发送端时延测量信息后将所述发送端时 延测量信息发送给 MCP；
下游 TLP的所述时戳获取模块， 具体用于在所述识别模块识别到所述 添加时延测量标志的数据包时， 所述时戳获取模块获取与该时延测量标志 对应的数据包接收端端时延测量信息； 所述接收端时延测量信息包括时戳 信息、业务流标识以及 TLP标识， 以使管理下游 TLP的 DCP获取所述接收端 时延测量信息后将所述接收端时延测量信息发送给 MCP。
28、 根据权利要求 27所述的 TLP, 其特征在于， 还包括：
所述 TLP为上游 TLP;
时间同步模块， 具体用于在所述上游 TLP的所述时戳获取模块在目标 业务流的数据包上添加时延测量标志之前 , 釆用网络时间协议 NTP或 IEEE 所述 TLP为下游 TLP;
所述时间同步模块， 具体用于在所述下游 TLP的所述识别模块在识别 到所述添加时延测量标志的数据包之前，釆用 NTP或 IEEE 1588v2时钟与所 述管理下游 TLP的 DCP进行时间同步；
所述 TLP, 还包括：
测量周期标识获取模块， 具体用于所述上游 TLP的所述测量周期标识 获取模块获取与所述时延测量标志对应的测量周期标识， 以使管理上游 TLP的 DCP获取所述测量周期标识后将所述测量周期标识发送给 MCP; 所 述下游 TLP的所述测量周期标识获取模块获取与所述时延测量标志对应的 测量周期标识以及每个测量周期起始时间， 以使管理下游 TLP的 DCP获取 所述起始时间和所述测量周期标识后进行匹配确定， 再将所述测量周期标 识发送给 MCP。
29、 根据权利要求 27所述的 TLP, 其特征在于， 还包括：
发送模块， 具体用于由所述上游 TLP的发送模块向下游 TLP发送测量 报文， 所述测量报文包括： 发送端时戳信息；
接收模块， 具体用于所述下游 TLP的接收模块接收到所述测量报文， 产生一个测量报文的到达时戳信息， 并将所述测量报文以及所述到达时戳 信息发送给所述管理下游 TLP的 DCP , 以使所述 DCP确定所述到达时戳信 息和所述接收端时戳信息是否同属于预设时长范围， 若是， 则确定所述发 送端时戳信息和所述接收端时戳信息属于同一数据包， 并将确定结果发送 给所述 MCP。
30、 根据权利要求 27所述的 TLP, 其特征在于， 所述上游 TLP的所述 时戳获取模块获取所述发送端时延测量信息还包括： 发送端业务流特征信 息以及发送端分片重组标识； 所述下游 TLP的所述时戳获取模块获取所述 接收端时延测量信息还包括： 接收端业务流特征信息以及接收端分片重组 标识；
以使所述管理上游 TLP的 DCP获取所述发送端时延测量信息并发送给 所述 MCP , 所述管理下游 TLP的 DCP获取所述接收端时延测量信息并发送 给所述 MCP, 以使所述 MCP根据发送端业务流特征信息、 发送端分片重组 标识、 接收端业务流特征信息以及接收端分片重组标识， 确定所述发送端 时戳信息和所述接收端时戳信息是对应于同一业务流的时戳信息。
31、 根据权利要求 26〜30中任一项所述的 TLP, 其特征在于， 所述上游 TLP的所述时戳获取模块在目标业务流的数据包上添加时延测量标志， 包 括：
所述时戳获取模块在所述数据包的 IP头中服务类型 TOS的保留位或者 Flags的保留位上添加时延测量标志。
32、 根据权利要求 26〜30中任一项所述的 TLP, 其特征在于， 所述识别 模块， 具体用于根据五元组中的至少两元信息， 对业务流进行识别。
33、 一种测量控制点 MCP, 其特征在于， 包括：
接收模块， 用于接收上游目标逻辑端口 TLP对应的数据收集点 DCP发 送的发送端时延测量信息以及与下游 TLP对应的 DCP发送的接收端时延测 量信息， 所述发送端时延测量信息包括时戳信息、 业务流标识以及 TLP标 识， 所述接收端时延测量信息包括时戳信息、 业务流标识以及 TLP标识； 确定模块， 用于根据所述发送端时延测量信息和所述接收端时延测量 信息， 确定网络单次时延情况。
34、根据权利要求 33所述的 MCP, 其特征在于， 所述接收模块， 包括： 第一接收单元， 具体用于接收上游 TLP对应的 DCP发送的发送端时延 测量信息以及与下游 TLP对应的 DCP发送的接收端时延测量信息， 所述发 送端时延测量信息包括时戳信息、 业务流标识以及 TLP标识， 所述接收端 时延测量信息包括时戳信息、 业务流标识以及 TLP标识；
第二接收单元， 具体用于接收所述管理上游 TLP的 DCP发送的测量周 期标识， 以及接收所述管理下游 TLP的 DCP发送的测量周期标识；
所述确定模块， 还包括：
第一匹配单元， 具体用于根据所述管理上游 TLP的 DCP发送的测量周 期标识和所述管理下游 TLP的 DCP发送的测量周期标识， 判断所述发送端 时延测量信息和所述接收端时延测量信息是否属于同一测量周期；
确定单元， 具体用于， 若是， 则根据所述发送端时延测量信息和所述 接收端时延测量信息， 确定网络单次时延情况。
35、 根据权利要求 33所述的 MCP, 其特征在于， 所述接收模块， 具体
于同一数据包的所述接收端时延测量信息；
所述确定模块， 具体用于根据所述发送端时延测量信息和所述接收端 时延测量信息， 确定网络单次时延情况。
36、 根据权利要求 33所述的 MCP, 其特征在于， 所述接收模块， 具 体用于接收所述上游 TLP对应的 DCP发送的发送端时延测量信息， 所述发 送端时延测量信息包括时戳信息、 业务流标识、 TLP标识、 发送端业务流 特征信息以及发送端分片重组标识， 以及接收所述下游 TLP对应的 DCP发 送的接收端时延测量信息， 所述接收端时延测量信息包括时戳信息、 业务 流标识、 TLP标识、 接收端业务流特征信息以及接收端分片重组标识； 所述确定模块包括：
第二匹配单元， 具体用于根据发送端业务流特征信息、 发送端分片重 组标识、 接收端业务流特征信息以及接收端分片重组标识， 判断所述发送 端时戳信息和所述接收端时戳信息是否对应于同一业务流的时戳信息； 确定单元， 具体用于， 若是， 则根据所述发送端时延测量信息和所述 接收端时延测量信息， 确定网络单次时延情况。
37、 一种网络时延测量系统， 其特征在于， 包括权利要求 19〜25中任 一项所述的 DCP、 权利要求 26〜32中任一项所述的 TLP以及权利要求 33〜36 中任一项所述的 MCP。
PCT/CN2012/082490 2012-09-29 2012-09-29 网络时延测量方法、装置和系统 WO2014047941A1 (zh)

Priority Applications (6)

Application Number Priority Date Filing Date Title
CN201280001472.8A CN103109501B (zh) 2012-09-29 2012-09-29 网络时延测量方法、装置和系统
ES12885357.9T ES2688696T3 (es) 2012-09-29 2012-09-29 Procedimiento y dispositivo de medición de un retardo de red
PCT/CN2012/082490 WO2014047941A1 (zh) 2012-09-29 2012-09-29 网络时延测量方法、装置和系统
EP12885357.9A EP2903212B1 (en) 2012-09-29 2012-09-29 Network delay measuring method and device
US14/669,513 US10374925B2 (en) 2012-09-29 2015-03-26 Method, apparatus, and system for measuring network delay
US16/524,570 US11133998B2 (en) 2012-09-29 2019-07-29 Method, apparatus, and system for measuring network delay

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2012/082490 WO2014047941A1 (zh) 2012-09-29 2012-09-29 网络时延测量方法、装置和系统

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US14/669,513 Continuation US10374925B2 (en) 2012-09-29 2015-03-26 Method, apparatus, and system for measuring network delay

Publications (1)

Publication Number Publication Date
WO2014047941A1 true WO2014047941A1 (zh) 2014-04-03

Family

ID=48316009

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/082490 WO2014047941A1 (zh) 2012-09-29 2012-09-29 网络时延测量方法、装置和系统

Country Status (5)

Country Link
US (2) US10374925B2 (zh)
EP (1) EP2903212B1 (zh)
CN (1) CN103109501B (zh)
ES (1) ES2688696T3 (zh)
WO (1) WO2014047941A1 (zh)

Families Citing this family (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103109501B (zh) * 2012-09-29 2016-11-09 华为技术有限公司 网络时延测量方法、装置和系统
CN102946330B (zh) * 2012-09-29 2017-03-15 华为技术有限公司 网络丢包测量方法、装置和系统
CN103501209B (zh) * 2013-09-25 2017-04-19 中国科学院声学研究所 一种异构多网协同传输的单业务分流方法和设备
US10555195B2 (en) * 2015-08-09 2020-02-04 Lg Electronics Inc. Method for performing uplink packet measurements in a wireless communication system and a device therefor
CN105119778B (zh) * 2015-09-09 2018-09-07 华为技术有限公司 测量时延的方法和设备
CN108141345B (zh) * 2015-10-08 2021-01-08 日本电信电话株式会社 传输系统、传输方法以及传输装置
US10687238B2 (en) * 2015-11-01 2020-06-16 Lg Electronics Inc. Method for transmitting an uplink packet delay measurement report in a wireless communication system and a device therefor
CN106817269B (zh) * 2015-12-01 2020-03-20 中国电信股份有限公司 网络层监测方法和系统以及相关设备
CN105515910B (zh) * 2015-12-17 2018-10-30 北京无线电计量测试研究所 一种群时延测量方法和设备
CN106936661B (zh) * 2015-12-31 2020-01-03 华为技术有限公司 一种网络监测方法、装置及系统
US10225161B2 (en) * 2016-10-31 2019-03-05 Accedian Networks Inc. Precise statistics computation for communication networks
CN108282811A (zh) * 2017-01-05 2018-07-13 中兴通讯股份有限公司 一种数据处理方法和装置
CN113542155B (zh) * 2017-04-11 2023-06-20 华为技术有限公司 用于处理业务流的方法及装置
CN110324162B (zh) * 2018-03-29 2021-10-15 华为技术有限公司 业务服务质量的检测方法、设备及系统
CN109039421A (zh) * 2018-06-28 2018-12-18 上海卫星工程研究所 用于多载荷数据相对传输时延测试的方法
CN110858811B (zh) * 2018-08-24 2022-01-18 华为技术有限公司 测量时延的方法和网络设备
IT201800010791A1 (it) * 2018-12-04 2020-06-04 Telecom Italia Spa Misura di prestazioni in una rete di comunicazioni a commutazione di pacchetto
CN112003757B (zh) * 2019-05-27 2022-06-17 杭州萤石软件有限公司 网络传输延迟确定方法、设备及系统
US11595308B2 (en) * 2019-06-13 2023-02-28 At&T Intellectual Property I, L.P. Closed loop prefix management and controller for whiteboxes
CN112398557B (zh) 2019-08-16 2022-06-28 华为技术有限公司 时延统计方法、装置、存储介质及系统
CN111030882A (zh) * 2019-11-11 2020-04-17 国电南瑞南京控制系统有限公司 一种基于智能变电站业务报文的网络质量测量方法及系统
CN114071487A (zh) * 2020-08-03 2022-02-18 中国移动通信有限公司研究院 一种接口时延测量方法、相关网络设备和存储介质
US11716268B2 (en) * 2020-12-01 2023-08-01 Cisco Technology, Inc. Telemetry data optimization for path tracing and delay measurement
CN112600735B (zh) * 2020-12-14 2022-08-09 北京信而泰科技股份有限公司 网络时延测试方法、装置及系统
CN115499335A (zh) * 2021-06-01 2022-12-20 中国移动通信有限公司研究院 一种网络性能测量方法、装置、通信设备和存储介质
CN115250243A (zh) * 2022-07-22 2022-10-28 中国电信股份有限公司 时延数据测量方法、装置、系统、电子设备及存储介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101056215A (zh) * 2006-04-14 2007-10-17 华为技术有限公司 一种网络性能测量方法及系统
CN101854268A (zh) * 2009-04-04 2010-10-06 华为技术有限公司 Ip网络性能测量、服务质量控制的方法、装置和系统
US20110292797A1 (en) * 2010-05-28 2011-12-01 Alcatel-Lucent Usa Inc. Dynamic delay budget allocation using single-point estimation of end-to-end delay

Family Cites Families (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4665519A (en) * 1985-11-04 1987-05-12 Electronic Systems Technology, Inc. Wireless computer modem
US7028133B1 (en) * 1999-04-30 2006-04-11 Daniel Kelvin Jackson Method and apparatus for extending communications over USB
US7639617B2 (en) * 2001-06-27 2009-12-29 Cisco Technology, Inc. Upstream physical interface for modular cable modem termination system
US7065086B2 (en) * 2001-08-16 2006-06-20 International Business Machines Corporation Method and system for efficient layer 3-layer 7 routing of internet protocol (“IP”) fragments
US7304995B2 (en) * 2001-08-29 2007-12-04 Texas Instruments Incorporated Systems and methods for packet flow control
US20050102412A1 (en) * 2003-11-10 2005-05-12 Jan Hirsimaki Transmission performance of a transport layer protocol connection
GB2422505A (en) * 2005-01-20 2006-07-26 Agilent Technologies Inc Sampling datagrams
JP2007180611A (ja) * 2005-12-26 2007-07-12 Toshiba Corp 通信システム及び通信方法
US7693190B2 (en) * 2006-11-22 2010-04-06 Cisco Technology, Inc. Lip synchronization for audio/video transmissions over a network
CN101237278A (zh) * 2007-01-30 2008-08-06 西门子通信技术（北京）有限公司 移动通信中传输数据的方法、系统、中继站及基站
US7702827B2 (en) * 2007-06-29 2010-04-20 International Business Machines Corporation System and method for a credit based flow device that utilizes PCI express packets having modified headers wherein ID fields includes non-ID data
CN101577631B (zh) * 2008-05-07 2012-04-25 华为技术有限公司 评价用户体验质量的方法、系统及网络装置
EP2335086A4 (en) * 2008-10-01 2012-09-05 Korea Electronics Telecomm APPARATUS AND METHOD FOR DETERMINING POSITION
US20100302967A1 (en) * 2009-05-29 2010-12-02 Electronics And Telecommunications Research Institute Method and apparatus for measuring network delay in ethernet ring network
JP2011081769A (ja) * 2009-09-14 2011-04-21 Ricoh Co Ltd データ転送装置、データ転送デバイスおよびデータ転送方法
US8724497B2 (en) * 2010-11-03 2014-05-13 Mediatek Inc. Method of uplink MDT measurement
WO2012059138A1 (en) * 2010-11-05 2012-05-10 Telecom Italia S.P.A. "measurement on a data flow in a communication network"
US8509072B2 (en) * 2011-03-07 2013-08-13 Comcast Cable Communications, Llc Network congestion analysis
US9430432B2 (en) * 2011-04-21 2016-08-30 Ineda Systems Pvt. Ltd. Optimized multi-root input output virtualization aware switch
US8958327B2 (en) * 2012-08-10 2015-02-17 Cisco Technology, Inc. Passive network latency monitoring
CN103109501B (zh) * 2012-09-29 2016-11-09 华为技术有限公司 网络时延测量方法、装置和系统
CN102946330B (zh) * 2012-09-29 2017-03-15 华为技术有限公司 网络丢包测量方法、装置和系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101056215A (zh) * 2006-04-14 2007-10-17 华为技术有限公司 一种网络性能测量方法及系统
CN101854268A (zh) * 2009-04-04 2010-10-06 华为技术有限公司 Ip网络性能测量、服务质量控制的方法、装置和系统
US20110292797A1 (en) * 2010-05-28 2011-12-01 Alcatel-Lucent Usa Inc. Dynamic delay budget allocation using single-point estimation of end-to-end delay

Also Published As

Publication number Publication date
CN103109501B (zh) 2016-11-09
US10374925B2 (en) 2019-08-06
CN103109501A (zh) 2013-05-15
ES2688696T3 (es) 2018-11-06
EP2903212A1 (en) 2015-08-05
US20190349279A1 (en) 2019-11-14
EP2903212A4 (en) 2016-04-20
EP2903212B1 (en) 2018-08-08
US11133998B2 (en) 2021-09-28
US20150207712A1 (en) 2015-07-23

Similar Documents

Publication Publication Date Title
WO2014047941A1 (zh) 网络时延测量方法、装置和系统
CN102904775B (zh) 网络丢包测量方法、设备和系统
WO2014048137A1 (zh) 网络丢包测量方法、装置和系统
KR101597255B1 (ko) 통신 네트워크에서 시간 측정 수행
US7315546B2 (en) Alignment of clock domains in packet networks
KR101459252B1 (ko) Traceroute_delay 진단 커맨드
JP5661941B2 (ja) 通信ネットワークにおけるデータ・フローに関する測定
Lv et al. An enhanced IEEE 1588 time synchronization for asymmetric communication link in packet transport network
WO2006102840A1 (fr) Procede de surveillance du taux de perte de paquets
WO2007095801A1 (fr) Procédé, appareil et système de mesure de la performance d&#39;un réseau
CN110971331B (zh) 一种逐跳时延测量方法和系统
WO2011079702A1 (zh) 丢包检测方法和装置及路由器
CN101753578B (zh) Ethernet/e1协议转换方法及协议转换器
CN102006157A (zh) 一种时间同步的方法和系统
WO2007056915A1 (en) A method for measuring mpls network performance parameter and device and system for transmitting packet
US6768748B2 (en) Flexible mapping of circuits into packets
EP2493126A1 (en) Method and system for clock recovery in packet switched network
WO2001088746A1 (en) Method and system for transmit time stamp insertion in a hardware time stamp system for packetized data networks
US11121938B2 (en) Performance measurement in a packet-switched communication network
JP5041815B2 (ja) 通信装置および基地局システム
KR20100048124A (ko) 근거리 통신망에서의 시간 동기화 방법
JP6341869B2 (ja) 通信装置および通信システム
WO2014013303A1 (en) Network service testing
JP5922047B2 (ja) 遅延測定方法および遅延測定システム
CN115038109A (zh) 时延测量方法及设备

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 201280001472.8

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12885357

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2012885357

Country of ref document: EP