WO2018119790A1 - 网络性能检测方法及装置 - Google Patents

网络性能检测方法及装置 Download PDF

Info

Publication number
WO2018119790A1
WO2018119790A1 PCT/CN2016/112734 CN2016112734W WO2018119790A1 WO 2018119790 A1 WO2018119790 A1 WO 2018119790A1 CN 2016112734 W CN2016112734 W CN 2016112734W WO 2018119790 A1 WO2018119790 A1 WO 2018119790A1
Authority
WO
WIPO (PCT)
Prior art keywords
test
packet
group
link
path
Prior art date
Application number
PCT/CN2016/112734
Other languages
English (en)
French (fr)
Inventor
张绿川
蔡宏坚
刘欣超
王宇飞
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 PCT/CN2016/112734 priority Critical patent/WO2018119790A1/zh
Priority to EP16925546.0A priority patent/EP3474499B1/en
Priority to CN201680069142.0A priority patent/CN108513703B/zh
Publication of WO2018119790A1 publication Critical patent/WO2018119790A1/zh

Links

Images

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
    • 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
    • H04L43/50—Testing arrangements
    • 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
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/12—Network monitoring probes

Definitions

  • the present application relates to Internet technologies, and in particular, to a network performance detecting method and apparatus.
  • IP Internet Protocol
  • IP networks carry more and more new Internet services (such as various video, voice, games, etc.), end users have higher and higher requirements for the perception of Internet services, so the more operators have requirements for IP network performance. The higher the value, the measurement and optimization of IP network performance is one of the operators' current and future challenges. Therefore, IP network performance monitoring technology is born.
  • FIG. 1 is a schematic diagram of an application scenario of performance monitoring of an existing IP network.
  • the detection time of the test path L 1, 2 and the test path L 1, 2, 3 is usually inconsistent, which may result in a measurement error of the packet loss rate of the link ⁇ P2, P3>.
  • the packet loss rate of the link ⁇ P1, P2> is a, but when detecting the packet loss rate of the test path L 1, 2, 3 , the chain The actual packet loss rate of the path ⁇ P1, P2> becomes a1. Therefore, when calculating b, if the packet loss rate of ⁇ P1, P2> is still considered to be a, the calculated b will have a large error.
  • the embodiment of the present application provides a network performance detecting method and device, which improves the accuracy of the measurement result of the network performance of the link to be tested.
  • the embodiment of the present application provides a network performance detection method, including:
  • the network performance of the link to be tested is determined according to the received first test packet and the second test packet.
  • the network performance detection method provided by the first aspect, by sending a first group of test messages for testing the first test path and a second group of test messages for testing the second test path, where the first test path includes a second test path and a link to be tested; further, receiving a first set of test packets returned on the first test path (the first set of test packets oscillate on the link to be tested according to a preset number of oscillations), And the second group of test packets returned on the second test path, and determining network performance of the link to be tested according to the received first group of test messages and the second group of test packets.
  • the proportion of the link to be tested in the entire first test path is increased by the manner in which the first test packet oscillates on the link to be tested (that is, the proportion of the link to be tested is greater than the upstream chain).
  • the proportion of the road is reduced, thereby reducing the impact of the fluctuation of the network performance of the upstream link (such as the network performance of the second test path) on the performance measurement of the measured link network, so that the measurement result of the network performance of the link to be tested is more accurate.
  • Each of the first test packets of the first group of test packets carries: at least one first link label group and at least one second link label group corresponding to the link to be tested; wherein each first chain The road label group includes M identical first link labels, and each second link label group includes M identical second link labels; the value of M is equal to a preset number of oscillations, and the first link label is used for Indicates that the first test message is in the next hop target node in the forward transmission direction, and the second link label is used to Indicates the next hop target node of the first test message in the reverse transmission direction.
  • the network performance of the link to be tested is determined according to the received first test packet and the second test packet, including:
  • the network performance of the link to be tested is determined according to the received first test packet and the second test packet, including:
  • the packet loss rate of the link to be tested is determined according to the packet loss rate of the first test path, the packet loss rate of the second test path, and the preset number of oscillations.
  • the network performance of the link to be tested is determined according to the received first test packet and the second test packet, including:
  • the out-of-order rate of the link to be tested is determined according to the out-of-order rate of the first test path, the out-of-order rate of the second test path, and the preset number of oscillations.
  • determining the delay or jitter of the first test path according to the time information carried in each of the first test packets in the received first test packet including:
  • each of the first test messages is assigned a receive timestamp according to the out-of-order order of the first set of test messages, including:
  • i is the packet identifier of the currently received first test packet;
  • j is the pending report. Message identification of the text;
  • i is equal to j, it is determined whether the j+1th first test message is saved in the cache sequence; wherein the cache sequence is used to save the out-of-order first test message, and the out-of-order first test message is reported.
  • the text identifier is greater than j;
  • the initial receiving timestamp of the j+1th first test packet is changed to the receiving timestamp of the ith first test packet, and Adding j to 1 and returning to execute the step of determining whether the j+1th first test message is saved in the cache sequence until the j+1th first test message is not saved in the cache sequence, and then continuing to receive the first The first test packet that arrives in the next test packet.
  • the ith first test packet is saved in the cache sequence, and the first test packet that arrives in the first group of test packets is continued to be received;
  • the jth first test packet is not saved in the cache sequence, it is determined that the jth first test packet is lost, according to the first test packet whose packet identifier is less than j in the first group of test packets.
  • the average delay determines the delay of the jth first test packet, and performs a step of determining whether the j+1th first test packet is stored in the cache sequence.
  • the method further includes:
  • the j+1th first test message in the cache sequence is deleted.
  • the network performance detecting method provided by the embodiment is configured to allocate a receiving timestamp for each first test message according to the out-of-order order of the first group of test messages; further, according to the receiving timestamp of each first test message. (that is, according to the out-of-order sequence of the first test packet, the appropriate receiving timestamp allocated for each first test packet) and the transmission timestamp, and the measurement result of the obtained first test path delay or jitter is calculated. More accurate, which further improves the accuracy of the delay or jitter measurement of the link to be tested.
  • the embodiment of the present application provides a network performance detecting apparatus, including:
  • the sending module is configured to send the first test packet and the second test packet, where the first test packet is used to test the first test path, and the second test packet is used to test the second test path.
  • the first test path includes: a second test path and a link to be tested;
  • a receiving module configured to receive a first group of test packets returned on the first test path, and a second group of test packets returned on the second test path, where the first group of test packets is based on a preset number of oscillations Oscillation on the link to be tested;
  • the determining module is configured to determine network performance of the link to be tested according to the received first group of test packets and the second group of test packets.
  • the device also includes:
  • a generating module configured to generate a first group of test packets according to a preset number of oscillations
  • Each of the first test packets of the first group of test packets carries: at least one first link label group and at least one second link label group corresponding to the link to be tested; wherein each first chain The road label group includes M identical first link labels, and each second link label group includes M identical second link labels; the value of M is equal to a preset number of oscillations, and the first link label is used for The next hop target node indicating the first test message in the forward transmission direction, and the second link label is used to indicate the next hop target node of the first test message in the reverse transmission direction.
  • the determination module includes:
  • a first determining submodule configured to determine a delay or a jitter of the first test path according to the time information carried in each first test packet in the received first test packet;
  • a second determining submodule configured to determine a delay or jitter of the second test path according to the time information carried in each second test packet of the received second test packet;
  • a third determining submodule configured to use the delay according to the first test path and the time of the second test path
  • the delay of the link to be tested is determined by the delay and the preset number of oscillations.
  • the jitter of the link to be tested is determined according to the jitter of the first test path, the jitter of the second test path, and the preset number of oscillations.
  • the determination module includes:
  • the fourth determining sub-module is configured to determine the first test path according to the number of the first test packet in the first group of test packets sent and the number of the first test packet in the first test packet received Packet loss rate;
  • the fifth determining submodule is configured to determine the second test path according to the number of the second test packet in the sent second test packet and the number of the second test packet in the received second test packet Packet loss rate;
  • the sixth determining submodule is configured to determine a packet loss rate of the link to be tested according to a packet loss rate of the first test path, a packet loss rate of the second test path, and a preset number of oscillations.
  • the determination module includes:
  • the seventh determining sub-module is configured to determine the first group of tests according to the packet identifier ID and the first preset packet identifier of the test packet carried in each of the first test packets in the received first test packet.
  • the eighth determining submodule is configured to determine the second group of test packets according to the packet ID of the test packet and the second preset packet identifier carried in each second test packet in the received second test packet The out-of-order number, and determining the out-of-order rate of the second test path according to the quotient of the number of out-of-order and the number of second test messages sent on the second test path;
  • the ninth determining sub-module is configured to determine an out-of-order rate of the link to be tested according to the out-of-order rate of the first test path, the out-of-order rate of the second test path, and the preset number of oscillations.
  • the first determining sub-module includes:
  • An allocating unit configured to allocate a receiving timestamp for each first test packet according to the out-of-order sequence of the first group of test packets
  • a determining unit configured to determine a delay or jitter of the first test path according to the receiving timestamp and the sending timestamp of each first test packet.
  • the allocation unit is specifically used to:
  • the receiving module When the receiving module receives the ith first test packet in the first group of test packets, it determines whether i is equal to j; where i is the packet identifier of the currently received first test packet; j is to be At The message identifier of the message;
  • i is equal to j, it is determined whether the j+1th first test message is saved in the cache sequence; wherein the cache sequence is used to save the out-of-order first test message, and the out-of-order first test message is reported.
  • the text identifier is greater than j;
  • the initial receiving timestamp of the j+1th first test packet is changed to the receiving timestamp of the ith first test packet, and Adding j to 1, returns to the execution of the step of determining whether the j+1th first test message is stored in the cache sequence, until the j+1th first test message is not saved in the cache sequence, the receiving module continues to receive The first test message that arrives in the next test packet.
  • the allocation unit is also used to:
  • the ith first test packet is saved to the cache sequence, and the receiving module continues to receive the first test packet that arrives in the first group of test packets;
  • the jth first test packet is not saved in the cache sequence, it is determined that the jth first test packet is lost, according to the first test packet whose packet identifier is less than j in the first group of test packets.
  • the average delay determines the delay of the jth first test packet, and performs a step of determining whether the j+1th first test packet is stored in the cache sequence.
  • the allocation unit modifies the initial receiving timestamp of the j+1th first test message to the receiving timestamp of the ith first test message, and is further used to: The j+1th first test message is deleted.
  • the embodiment of the present application provides a network performance detection method, including:
  • the delay or jitter of the test path is determined according to the receiving timestamp and the sending timestamp of each test packet.
  • the network performance detection method provided by the first aspect, after receiving a set of test packets returned on the test path, assigning a receive timestamp to each test message according to the out-of-order order of the set of test messages; further, according to The receiving timestamp of each test packet (that is, the appropriate receiving timestamp assigned to each test packet according to the out-of-order of the group of test packets), and the sending timestamp, so that the network of the obtained test path is calculated Performance measurements are more accurate.
  • the receiving timestamp is assigned to each test packet in the group test packet according to the out-of-order of the group test message, including:
  • i is the packet identifier of the currently received test packet
  • j is the packet identifier of the to-be-processed packet
  • i is equal to j, it is determined whether the j+1th test message is stored in the cache sequence; wherein the cache sequence is used to save the out-of-order test message, and the message identifier of the out-of-order test message is greater than j;
  • the initial receiving timestamp of the j+1th test packet is changed to the receiving timestamp of the i th test packet, and j is incremented by 1, and returned.
  • the step of determining whether the j+1th test packet is saved in the cache sequence is performed until the j+1th test packet is not saved in the cache sequence, and then the next test packet in the group test packet is continuously received. .
  • the ith test packet is saved in the cache sequence, and the next test packet in the group test packet is continuously received;
  • the j-th test packet is not saved in the cache sequence, it is determined that the j-th test packet is lost, and the j-th is determined according to the average delay of each test packet whose packet identifier is less than j in the group test packet. The delay of the packet is tested, and a step of determining whether the j+1th test packet is saved in the cache sequence is performed.
  • the method further includes:
  • the embodiment of the present application provides a network performance detecting apparatus, including:
  • a receiving module configured to receive a set of test messages returned on the test path
  • An allocating module configured to allocate a receiving timestamp for each test packet in the group test packet according to the out-of-order of the group test packet;
  • the determining module is configured to determine the delay or jitter of the test path according to the receiving timestamp and the sending timestamp of each test packet.
  • the allocation module is specifically used to:
  • the receiving module When the receiving module receives the ith test packet in the group test packet, it determines whether i is equal to j; where i is the packet identifier of the currently received test packet; j is the packet of the packet to be processed. Identification
  • i is equal to j, it is determined whether the j+1th test message is stored in the cache sequence; wherein the cache sequence is used to save the out-of-order test message, and the message identifier of the out-of-order test message is greater than j;
  • the initial receiving timestamp of the j+1th test packet is changed to the receiving timestamp of the i th test packet, and j is incremented by 1, and returned.
  • the step of determining whether the j+1th test message is saved in the cache sequence is performed until the j+1th test message is not saved in the cache sequence, and the receiving module continues to receive the next test in the group test message. Message.
  • the distribution module is also used to:
  • the ith test packet is saved in the cache sequence, and the receiving module continues to receive the next test packet in the group test packet;
  • the j-th test packet is not saved in the cache sequence, it is determined that the j-th test packet is lost, and the j-th is determined according to the average delay of each test packet whose packet identifier is less than j in the group test packet. The delay of the packet is tested, and a step of determining whether the j+1th test packet is saved in the cache sequence is performed.
  • the allocation module changes the initial receiving timestamp of the j+1th test message to the receiving timestamp of the i-th test message, it is also used to:
  • FIG. 1 is a schematic diagram of an application scenario of performance monitoring of an existing IP network
  • FIG. 2A is a schematic flowchart of Embodiment 1 of a network performance detecting method according to the present application
  • 2B is a schematic diagram of an application scenario of a network performance detection method according to the present application.
  • FIG. 3A is a schematic flowchart of Embodiment 2 of a network performance detecting method according to the present application
  • 3B is a schematic diagram of a format of a test packet of the present application.
  • Embodiment 4 is a schematic flowchart of Embodiment 3 of a network performance detecting method according to the present application.
  • Embodiment 4 is a schematic flowchart of Embodiment 4 of a network performance detecting method according to the present application.
  • Embodiment 5B is a schematic flowchart of Embodiment 5 of a network performance detecting method according to the present application.
  • FIG. 6A is a schematic flowchart of Embodiment 6 of a network performance detecting method according to the present application.
  • FIG. 6B is a schematic flowchart of a network performance detecting apparatus assigning an appropriate receiving timestamp to each first test packet according to an embodiment of the present application;
  • FIG. 7 is a schematic flowchart of a network performance detecting apparatus assigning an appropriate receiving timestamp to each test packet according to an embodiment of the present application
  • FIG. 8 is a schematic diagram of a test message sequence and a cache sequence received in an embodiment of the network performance detecting method of the present application
  • Embodiment 7 of a network performance detecting method is a schematic flowchart of Embodiment 7 of a network performance detecting method according to the present application.
  • FIG. 9B is a schematic flowchart 3 of a network performance detecting apparatus for assigning a suitable receiving timestamp to each test packet according to an embodiment of the present application;
  • FIG. 10 is a schematic structural diagram of Embodiment 1 of a network performance detecting apparatus according to the present application.
  • Embodiment 11 is a schematic structural diagram of Embodiment 2 of a network performance detecting apparatus according to the present application.
  • Embodiment 12 is a schematic structural diagram of Embodiment 3 of a network performance detecting apparatus according to the present application.
  • FIG. 13 is a schematic structural diagram of Embodiment 4 of a network performance detecting apparatus according to the present application.
  • prl represents the packet loss rate
  • N send and N recv respectively represent the number of test packets sent and the number of received packets.
  • the delay of a single test packet can be expressed as
  • the delay i represents the delay of the i-th test packet
  • T i, recv represents the reception time of the i-th test packet
  • T i, send represents the transmission time of the i-th test packet.
  • the delay of the entire test path can be obtained by taking the average of the delays of all test packets on the test path:
  • N represents the number of all test packets received on the test path, and N is a positive integer greater than or equal to 1.
  • the jitter between the two test packets can be defined as:
  • Jitter i abs(delay i+1 -delay i ) (4)
  • abs(_) represents the absolute value
  • delay i+1 represents the delay of the i+ 1th test message
  • delay i represents the delay of the i-th test message
  • jitter i represents the i+1th test. Jitter between the message and the ith test message.
  • the jitter on the entire test path is obtained by taking the average of the jitter of all test packets:
  • jitter represents the jitter of the test path.
  • the out-of-order rate is similar to the packet loss rate.
  • the out-of-order of a single test packet has no meaning.
  • the network performance of the link can only be calculated through the network performance of the test path. Therefore, the measurement of the network performance of the latter hop link depends on the previous hop.
  • Network performance, and the upstream link accounts for more than the entire test path Or equal to the proportion of the downstream tested link, that is, the fluctuation of the network performance of the upstream link may seriously affect the measurement of the performance of the downstream link network, so that the measurement result of the network performance of the downstream link is relatively large.
  • Embodiment 1 of a network performance detecting method is a schematic flowchart of Embodiment 1 of a network performance detecting method according to the present application.
  • the execution body of this embodiment may be a network performance detecting device configured in a server, and the device may be implemented by software and/or hardware.
  • the method in this embodiment may include:
  • Step S201 Send a first group of test messages and a second group of test messages.
  • FIG. 2B is a schematic diagram of an application scenario of the network performance detection method of the present application.
  • the network shown in FIG. 2B is an all-IP network, and a segment routing (SR) protocol is run in the network.
  • P0 represents a network performance detecting device
  • P1, P2 and P3 represent routers or repeaters.
  • the first test paths L1 ⁇ P0, P1, P2, P3, P2, P3, P2, P1, P0> and the second test path L2 ⁇ P0, P1, P2, P1, P0> are shown in FIG. 2B.
  • the first test path L1 includes a second test path L2 and a link L3 to be tested.
  • the network performance detection method of the present application network can also be applied to other forms of network structures, which are not limited in this embodiment of the present application.
  • the network performance detecting apparatus sends a first group of test packets and a second group of test packets, wherein the first group of test messages is used to test the first test path L1, and the second group of tests
  • the message is used to test the second test path L2;
  • the first test path L1 includes: a second test path L2 and a link L3 to be tested.
  • each first test packet in the first group of test packets carries a test path identifier 1 (a first test path L1 for indicating the first test packet detection), and a second group of test packets.
  • Each of the second test packets carries a test path identifier 2 (a second test path L2 for indicating the second test packet detection).
  • Step S202 Receive a first group of test packets returned on the first test path, and a second group of test packets returned on the second test path.
  • the network performance detecting apparatus receives the first group of test packets returned on the first test path L1 and the second group of test packets returned on the second test path L2; wherein the first group of test packets is based on the pre-test
  • the number of oscillations is set to oscillate on the link to be tested.
  • the first test packet in the first group of test packets is on the link L3 to be tested. Oscillation (such as P2->P3, P3->P2, P2->P3, P3->P2). It can be seen that, in the embodiment of the present application, the proportion of the link to be tested is far greater than the proportion of the upstream link, thereby reducing the upstream link. The impact of network performance fluctuations on the measurement of link network performance measurements.
  • Step S203 Determine network performance of the link to be tested according to the received first group of test packets and the second group of test packets.
  • the network performance detecting device determines the network performance of the link L3 to be tested according to the received first group of test packets and the second group of test packets.
  • the network performance detecting apparatus determines network performance of the first test path L1 according to the received first set of test messages, and determines network performance of the second test path L2 according to the received second set of test messages; further, The network performance of the link to be tested is determined according to the network performance of the first test path L1, the network performance of the second test path L2, and the preset number of oscillations.
  • the proportion of the link to be tested is far greater than the upstream link by using the first group of test packets to oscillate on the link to be tested according to the preset number of oscillations (for example, The proportion of the link P1->P2 and the link P2->P1) in FIG. 2B, thereby reducing the influence of the fluctuation of the network performance of the upstream link on the performance measurement of the link network to be tested, and reducing the link to be tested. Error in measurement results of network performance.
  • the network performance includes one or more of the following: delay, jitter, packet loss rate, and out-of-order rate.
  • the network performance of the present application may also include other network performances, which is not limited in the embodiment of the present application.
  • the first test path for testing the first test path and the second test message for testing the second test path are sent, where the first test path includes: the second test path And the link to be tested; further, receiving the first group of test packets returned on the first test path (the first group of test packets oscillate on the link to be tested according to a preset number of oscillations), and the second test path
  • the second group of test packets returned, and the network performance of the link to be tested is determined according to the received first group of test packets and the second group of test packets.
  • the proportion of the link to be tested in the entire first test path is increased by the manner in which the first test packet oscillates on the link to be tested (that is, the proportion of the link to be tested is greater than the upstream chain).
  • the proportion of the road is reduced, thereby reducing the impact of the fluctuation of the network performance of the upstream link (such as the network performance of the second test path) on the performance measurement of the measured link network, so that the measurement result of the network performance of the link to be tested is more accurate.
  • FIG. 3A is a schematic flowchart of Embodiment 2 of a network performance detection method of the present application
  • FIG. 3B is a schematic diagram of a format of a test packet of the present application.
  • the method further includes:
  • Step S204 Generate a first group of test messages according to a preset number of oscillations.
  • the network performance detecting device before sending the first group of test packets and the second group of test packets, the network performance detecting device generates a first group of test packets according to the preset number of oscillations; wherein, in the first group of test packets, Each first test packet carries at least one first link label group and at least one second link label group corresponding to the link to be tested; each first link label group includes M identical first links a label, each second link label group includes M identical second link labels; the value of M is equal to a preset number of oscillations, and the first link label is used to indicate that the first test packet is in a forward transmission direction ( That is, the next hop target node that sends the test packet along the network performance detecting device, and the second link label is used to indicate that the first test packet is in the reverse transmission direction (ie, the test packet is received along the network performance detecting device).
  • the first link label included in the first link label group corresponding to the link to be tested and the second link label included in the second link label group may be carried in the SR of the first test packet.
  • the node on the link to be tested forwards the first test packet according to the link label included in the SR header after receiving the first test packet, so that the first test packet is sent. Oscillation on the link to be tested increases the proportion of the link to be tested in the entire first test path (that is, the proportion of the link to be tested is greater than the proportion of the upstream link), thereby reducing the network performance of the upstream link. The impact of fluctuations on the measured performance of the link network.
  • the preset number of oscillations is 2, and the first test packet oscillates on the link L3 to be tested (for example, P2->P3, P3->P2, P2->P3, P3->P2).
  • Each first test packet carries: a first link label group including two identical first link labels (such as P2->P3 link labels), and includes two identical second link labels (eg, P3->P2 link label) The second link label group.
  • each first test packet carries: a first link label including two identical first link labels (such as P2->P3 link label) Group 1, the first link label group 2 including two identical first link labels (such as P3->P4 link labels), including two identical second link labels (such as P3->P2 link labels)
  • the second link label group 1 and the second link label group 2 including two identical second link labels (such as P4->P3 link labels).
  • each of the first test packet and the second test packet of the second group of test packets can be carried: the identifier of the detector; the identifier of the detector is used to indicate the network performance detecting device that sends the corresponding test message.
  • the identifier of the probe carried in the first test packet is used to indicate the network performance detecting device that sends the first test packet
  • the identifier of the probe carried in the second test packet is used to indicate the network performance check of the second test packet.
  • the test packet includes: a path information field (for controlling a forwarding path of the test packet), a detection information field, and an optional field.
  • the path information field includes: an SR header (including an SR label) and an IP header.
  • the detection information field includes information necessary for calculating network performance on the test path.
  • the test message identifier is used to uniquely mark the test message (ie, filter other irrelevant messages), and the probe identifier is used to identify that the test message is
  • the test packet path identifier sent by the network performance detection device is used to identify which test path the test packet is responsible for, and the identity of the test packet (Identity, referred to as ID) is used to distinguish different on the same test path.
  • the test packet and the sending timestamp are used to indicate the sending time of the test packet, and the receiving timestamp is used to indicate the receiving time of the test packet.
  • step S203 includes:
  • Step S203A1 Determine delay or jitter of the first test path according to time information carried in each first test packet in the received first test packet.
  • each first test packet carries a sending timestamp and a receiving timestamp, where the sending timestamp is used to indicate that the network performance detecting apparatus sends the sending time of the first test packet, and the receiving timestamp is used to indicate the network.
  • the performance detecting device receives the receiving time of the first test message.
  • the network performance detecting apparatus passes the above formula according to the time information (for example, the sending time stamp and the receiving timestamp) carried in each of the first test packets received in the first group of test messages. 3) Determine the delay of the first test path L1.
  • the network performance detecting apparatus passes the time information (for example, sending timestamp and receiving timestamp) carried in each first test packet in the received first group of test packets, by using the above formula (2) and ( 4) Determine the jitter of the first test path L1.
  • time information for example, sending timestamp and receiving timestamp
  • Step S203B1 Determine time delay or jitter of the second test path according to time information carried in each second test packet in the received second group of test packets.
  • each second test packet carries a sending timestamp and a receiving timestamp, where the sending timestamp is used to indicate that the network performance detecting apparatus sends the sending time of the second test packet, The timestamp is used to indicate the time when the network performance detecting device receives the second test message.
  • the network performance detecting apparatus passes the above formula according to time information (for example, sending time stamp and receiving timestamp) carried in each second test packet in the received second group of test messages. 3) Determine the delay of the second test path L2.
  • the network performance detecting apparatus passes the time information (for example, sending timestamp and receiving timestamp) carried in each second test packet of the received second group of test packets by using the formula (2), ( 4) and (5) determine the jitter of the second test path L2.
  • time information for example, sending timestamp and receiving timestamp
  • Step S203C1 determining a delay of the link to be tested according to the delay of the first test path, the delay of the second test path, and the preset number of oscillations; or, according to the jitter of the first test path, the second test path Jitter and the preset number of oscillations determine the jitter of the link to be tested.
  • the oscillating factor needs to be considered when calculating the network performance of the link L3 to be tested.
  • the first test packet oscillates n times on the link L3 to be tested (that is, the preset number of oscillations is n), according to the delay delay L1 of the first test path L1 and the second test path L2.
  • the delay delay L2 and the preset number of oscillations n determine the delay delay L3 of the link L3 to be tested by the following formulas (6) and (7):
  • Delay L1 2*delay L2 +2n*delay L3 (6)
  • the jitter of the jitter jitter L1 and the second test path L2 according to the first test path L1 is performed.
  • Jitter L2 and the preset number of oscillations n determine the jitter jitter L3 of the link L3 to be tested by the following formulas (8) and (9):
  • Jitter L1 2*jitter L2 +2n*jitter L3 (8)
  • the calculation of the delay or jitter of the link L3 to be tested needs to be averaged multiple times, thereby reducing the upstream.
  • the delay of the link or the jitter (such as the delay or jitter of the second test path L2) affects the network performance measurement of the link L3, so that the delay or jitter of the link L3 to be tested is measured. The result is more accurate.
  • FIG. 5A is a schematic flowchart of Embodiment 4 of a network performance detecting method according to the present application. As shown in FIG. 5A, based on the above embodiment, step S203 includes:
  • Step S203A2 Determine a packet loss rate of the first test path according to the number of the first test packets in the first group of test packets sent and the number of the first test packets in the first group of test packets.
  • the network performance detecting apparatus passes the above formula according to the number of the first test packet in the first group of test packets sent and the number of the first test packet in the first group of test packets received.
  • the packet loss rate of the first test path L1 is determined.
  • Step S203B2 Determine the packet loss rate of the second test path according to the number of the second test packet in the second group of test packets sent and the number of the second test packet in the second group of test packets.
  • the network performance detecting apparatus passes the above formula according to the number of the second test packet in the second group of test packets sent and the number of the second test packet in the second group of test packets received.
  • the packet loss rate of the second test path L2 is determined.
  • Step S203C2 Determine a packet loss rate of the link to be tested according to the packet loss rate of the first test path, the packet loss rate of the second test path, and the preset number of oscillations.
  • the oscillating factor needs to be considered when calculating the network performance of the link L3 to be tested.
  • the first test packet oscillates n times on the link L3 to be tested that is, the preset number of oscillations is n
  • the packet loss rate prl L2 and the preset number of oscillations n determine the packet loss rate prl L3 of the link L3 to be tested by the following formulas (10) and (11):
  • the calculation of the packet loss rate of the link L3 to be tested needs to be powered by n, thereby reducing the upstream chain.
  • the fluctuation of the packet loss rate of the path affects the network performance measurement of the link L3, so that the measurement result of the packet loss rate of the link L3 to be tested is more accurate.
  • FIG. 5B is a schematic flowchart of Embodiment 5 of a network performance detecting method according to the present application. As shown in Figure 5B As shown, based on the above embodiment, step S203 includes:
  • Step S203A3 Determine, according to the packet identification ID and the first preset packet identifier of the test packet carried in each of the first test packets, the out-of-order of the first group of test packets. And determining the out-of-order rate of the first test path according to the quotient of the number of out-of-order and the number of the first test packets sent on the first test path.
  • the network performance detecting device identifies the packet ID of the test packet and the first preset packet identifier (such as the pending test packet) of each test packet carried in the received first test packet.
  • the packet identifier is used to determine whether the packet ID of the first test packet is equal to the first preset packet identifier, and if yes, the first test packet received is a test packet that arrives in advance. Then, the currently received first test message is stored in a cache sequence (the cache sequence is used to save the out-of-order first test message).
  • the network performance detecting apparatus determines the out-of-order number of the first group of test packets according to the number of the first test packets included in the cache sequence, and performs all the first tests sent on the first test path according to the out-of-order number and the first test path.
  • the quotient of the number of messages determines the out-of-order rate of the first test path L1.
  • Step S203B3 Determine, according to the packet ID of the test packet and the second preset packet identifier carried in each second test packet of the received second test packet, the number of out-of-order numbers of the second group of test packets. And determining the out-of-order rate of the second test path according to the quotient of the number of out-of-order and the number of second test packets sent on the second test path.
  • the network performance detecting device identifies, according to the packet ID of the test packet and the second preset packet, which are carried in the second test packet of the received second test packet, such as the test packet to be processed.
  • the message identifier is used to determine whether the packet ID of the second test packet is equal to the second preset packet identifier, and if yes, that is, the currently received second test packet is a test packet that arrives in advance, Then, the currently received second test message is stored in a cache sequence (the cache sequence is used to save the out-of-order second test message).
  • the network performance detecting apparatus determines the out-of-order number of the second group of test packets according to the number of the second test packets included in the buffer sequence, and performs all the second tests according to the out-of-order number and the second test path.
  • the quotient of the number of messages determines the out-of-order rate of the second test path L2.
  • Step S203C3 Determine an out-of-order rate of the link to be tested according to the out-of-order rate of the first test path, the out-of-order rate of the second test path, and the preset number of oscillations.
  • the first test packet oscillates on the link L3 to be tested according to the preset number of oscillations. Therefore, the oscillating factor needs to be considered when calculating the network performance of the link L3 to be tested.
  • the first test packet oscillates n times on the link L3 to be tested (that is, the preset number of oscillations is n), according to the out-of-order rate disor L1 and the second test path L2 of the first test path L1.
  • the out-of-order rate disor L2 and the preset number of oscillations n determine the out-of-order rate disor L3 of the link L3 to be tested by the following formulas (12) and (13):
  • Disor L1 1-(1-disor L2 ) 2 (1-disor L3 ) 2n (12)
  • Disor L3 1-((1-disor L1 )/(1-disor L2 ) 2 ) 1/2n (13)
  • the calculation of the out-of-order rate of the link L3 to be tested needs to be powered by n, thereby reducing the upstream chain.
  • the fluctuation of the out-of-order rate of the path affects the network performance measurement of the link L3, so that the measurement result of the out-of-order rate of the link L3 to be tested is more accurate.
  • the delay of the test path and the calculation of the jitter depend entirely on the transmission time and the reception time of the test message. Since the test packets are sent by the network performance detection device one by one, it can be considered that the test packets are sent at an accurate time. Therefore, marking the test receiving timestamp for each test packet directly affects the test path. Delay and jitter calculations.
  • the application when the application processes the data, it usually stores all the received data into the buffer, and sorts the data in the buffer to process the data in order. For example, suppose that the video data of the third second is prioritized over the video data of the first second, and the application does not directly process the video data of the third second, but waits until the video data of the first second arrives, and processes the first in order. 1 second of video data, therefore, the delay perceived by the user is longer than the real delay of the video data (i.e., the time between the time when the video data is sent from the transmitting end and the time of reaching the receiving end).
  • the time at which the test message arrives at the receiving node is not necessarily the real processing time.
  • the delay In order to make the calculation of the delay closer to the user's real feeling, it is necessary to consider the effect of out-of-order on the delay and/or jitter.
  • the following describes how to apply the appropriate receiving time stamp to the test message in consideration of the out-of-order situation. Get more accurate measurements of delay and/or jitter.
  • FIG. 6 is a schematic flowchart of Embodiment 6 of a network performance detecting method according to the present application. As shown in FIG. 6A, based on the foregoing embodiment, step S203A1 includes:
  • Step SA11 Allocating each first test message according to the out-of-order order of the first group of test messages Receive timestamps.
  • the network performance detecting device allocates an appropriate receiving time stamp for each first test packet according to the out-of-order of the first group of test packets, so that the calculation of the delay and/or jitter is closer to the real feeling of the user.
  • FIG. 6B is a schematic flowchart of a network performance detecting apparatus assigning an appropriate receiving timestamp to each first test packet according to an embodiment of the present application.
  • the network performance detecting device determines whether i is equal to j; where i is the currently received first test packet.
  • the message identifier; j is the message identifier of the test packet to be processed (ie, the network performance detecting device has allocated a suitable receiving time stamp for j-1 first test messages).
  • i is an integer greater than 0 and j is an integer greater than one.
  • the network performance detecting apparatus determines whether the j+1th first test message is stored in the cache sequence.
  • the cache sequence is used to save the out-of-order first test packet (that is, the first test packet that arrives in advance), and the packet identifier of the out-of-order first test packet is greater than j.
  • the j+1th first test message is stored in the cache sequence (ie, the j+1th first test message arrives at the network performance detecting device earlier than the jth first test message), then the j+th The initial receiving timestamp of the first test packet is modified to the receiving timestamp of the i-th (ie, jth) first test packet, and j is incremented by one, and the execution judgment cache sequence is saved with the jth. The step of +1 the first test message, and so on, until the j+1th first test message is not saved in the cache sequence, and continues to receive the first test of the next arrival in the first group of test messages Message.
  • the j+1th first test report in the buffer sequence is to be cached.
  • the text is deleted, so that the first test message that arrives earlier can be saved in the cache sequence.
  • the network performance detecting device determines whether the cache sequence is full. 2a) If the cache sequence is not full, save the ith first test message to the cache sequence, and continue to receive the next test message that arrives in the first group of test messages. 2b) If the buffer sequence is full, it is determined whether the jth first test packet (ie, the pending test packet) is stored in the cache sequence. If the jth first test packet is not saved in the cache sequence, it is determined that the jth first test packet is lost, according to the first test packet whose packet identifier is less than j in the first group of test packets.
  • the network performance detecting apparatus determines whether the j+1th first test packet is saved in the cache sequence.
  • the network performance detecting apparatus as the jth number, the product of the average delay of each first test message whose message identifier is less than j in the first group of test messages and a preset coefficient (for example, 1.5) The delay of a test message.
  • the delay of the jth first test packet may be determined by other methods according to the average delay of the first test packet whose packet identifier is less than j in the first group of test packets. There is no limit to this.
  • the network performance detecting device applies a suitable receiving time stamp to each of the first test packets of the first group of test packets received, wherein a reasonable delay is given to the lost first test packet.
  • Step SA12 Determine a delay or jitter of the first test path according to the receiving timestamp and the sending timestamp of each first test packet.
  • the network performance detecting apparatus is configured according to the receiving timestamp of each first test packet (that is, an appropriate receiving timestamp allocated for each first test packet according to the out-of-order order of the first group of test packets), Send a timestamp to determine the delay or jitter of the first test path.
  • the receiving timestamp of each first test packet that is, an appropriate receiving timestamp allocated for each first test packet according to the out-of-order order of the first group of test packets
  • the receiving time stamp is allocated to each first test packet according to the out-of-order of the first group of test packets; further, according to the receiving time stamp of each first test packet (ie, according to the first group)
  • the out-of-order of the test packet is the appropriate receiving timestamp assigned to each first test packet, and the timestamp is sent, and the measurement result of the obtained first test path delay or jitter is more accurate, thereby further improving The delay of the link to be tested or the accuracy of the jitter measurement.
  • the step S203B1 in the third embodiment of the foregoing network performance detection method is used to determine the delay of the second test path or the time information carried in the second test packet according to the received second test packet.
  • the jitter is obtained by: assigning a receiving timestamp to each second test packet according to the out-of-order of the second group of test packets; determining the second test path according to the receiving timestamp and the sending timestamp of each second test packet Delay or jitter.
  • the measurement result of the delay or jitter of the obtained second test path is more accurate according to the above manner, thereby further improving the accuracy of the delay or jitter measurement result of the link to be tested.
  • FIG. 7 is a schematic flowchart of a network performance detecting apparatus assigning an appropriate receiving timestamp to each test packet according to an embodiment of the present application. It is assumed that the network performance detecting device starts from the first test packet, and assigns a suitable receiving time stamp to the test packet one by one, and j indicates the packet identifier of the test packet to be processed (that is, the network performance detecting device has determined that the packet identifier is less than or equal to The test packet of j-1 is assigned a suitable receiving timestamp. i indicates the packet identifier of the test packet currently received by the network performance detecting device, and m1 indicates the total number of test packets received by the network performance detecting device. , m2 represents the total number of test messages that have entered the buffer sequence.
  • the network performance detecting apparatus allocates an appropriate receiving timestamp for each test packet according to the following procedure:
  • the network performance detecting device When receiving the ith test packet, the network performance detecting device adds 1 to m1 and adds an initial receiving timestamp for the ith test packet (used to indicate the receiving time of the ith test packet) .
  • the network performance detecting device determines whether i is equal to j; if i is not equal to j (ie, the currently received i-th test message is a non-to-be-processed message), the i-th test message is stored in the buffer. Sequence and determine if the cache sequence is full. If the buffer sequence is not full, m2 will be +1 and continue to receive the next arriving test message. If the cache sequence is full, it is determined whether the j-th test packet is saved in the cache sequence; if the j-th test packet is not stored in the cache sequence, it is determined that the j-th test packet has been lost. The time of receiving the j-th test packet cannot be determined.
  • the average delay of each test packet whose identifier is less than j and the preset coefficient (for example, 1.5) are determined according to the arrived packet.
  • the product is used to determine the default delay of the jth test message; further, it is determined whether the j+1th test message is stored in the cache sequence.
  • the network performance detecting apparatus determines whether the j+1th test message is stored in the cache sequence (ie, determining Whether the j+1th test message arrives earlier than the jth test message). If the j+1th test message is stored in the cache sequence (ie, the j+1th test message arrives at the network performance detecting device earlier than the jth test message), the receiving time of the jth test message is received.
  • the stamp covers the initial receiving timestamp of the j+1th test message, and deletes the j+1th test message in the buffer sequence and j plus 1 to return to the execution judgment cache sequence whether the j+1th number is saved. a step of testing the message, and so on;
  • the network performance detecting device calculates the average delay of the test packets that have been received in real time. When no new test packet is received within five consecutive average delay times, it is determined that the test packet is received.
  • the network performance detecting apparatus allocates an appropriate receiving time stamp for each test message according to the out-of-order, so that the calculation of the delay is closer to the real feeling of the user, so as to obtain a more accurate time. Measurement of delay and/or jitter.
  • FIG. 8 is a schematic diagram of a test message sequence and a cache sequence received in an embodiment of the network performance detecting method of the present application. As shown in FIG. 2B and FIG. 8, the flow of the network performance detecting apparatus assigning an appropriate receiving time stamp to each test message is described in detail.
  • the text is stored in the buffer sequence (assuming that the cache sequence can store 5 messages);
  • the buffer sequence is full, and the network performance detecting device determines that the third test packet has been lost, and assigns a default delay to the third test packet (optionally, the default delay is equal to the first test packet) Further, 1.5 times of the average delay of the second test message); further, when the fourth test message is stored in the determined cache sequence, the receive time stamp of the eighth test message is overwritten with the fourth test.
  • the initial timestamp of the message is received, and the fourth test message in the buffer sequence is deleted, and j is incremented by 1; the fifth test message, the sixth test message, and the seventh in the cache sequence are sequentially The test packet and the eighth test packet are processed accordingly, until the test packet that arrives early is not saved in the cache sequence, and the next arriving test packet (ie, the ninth test packet) is continuously received;
  • j is equal to j
  • the eleventh test packet is stored in the cached sequence, and the receiving timestamp of the tenth test packet is overwritten with the initial receiving timestamp of the eleventh test packet.
  • the network performance detecting device puts a suitable receiving timestamp for each test packet received, and gives a reasonable delay to the lost test packet, so that the measurement result of the delay/jitter of the test path is more accurate.
  • the network performance detecting device does not affect the second test packet and the third test packet, and the third one by assigning a default delay to the third test packet. The calculation of jitter between the test message and the fourth test message.
  • the delay of the test path and the calculation of the jitter are completely dependent on the transmission time and reception time of the test message. Since the test packets are sent by the network performance detection device one by one, it can be considered that the test packets are sent at an accurate time. Therefore, marking the test receiving timestamp for each test packet directly affects the test path. Delay and jitter calculations.
  • the application when the application processes the data, it usually stores all the received data into the buffer, and sorts the data in the buffer to process the data in order. For example, suppose that the video data of the third second is prioritized over the video data of the first second, and the application does not directly process the video data of the third second, but waits until the video data of the first second arrives, and processes the first in order. 1 second of video data, therefore, the delay perceived by the user is longer than the real delay of the video data (i.e., the time between the time when the video data is sent from the transmitting end and the time of reaching the receiving end).
  • the time at which the test message arrives at the receiving node is not necessarily the real processing time.
  • the following describes how to apply the appropriate receiving timestamp to the test message in consideration of the out-of-order situation. In order to obtain more accurate measurements of delay and / or jitter.
  • FIG. 9A is a schematic flowchart of Embodiment 7 of a network performance detecting method according to the present application.
  • the execution body of this embodiment may be a network performance detecting device configured in a server, and the device may be implemented by software and/or hardware.
  • the method in this embodiment may include:
  • Step S901 Receive a set of test messages returned on the test path.
  • the network performance detecting device receives a set of test packets returned on a certain test path.
  • each test packet in the group of test packets carries: a probe identifier; the identifier of the probe is used to indicate a network performance detecting device that sends a corresponding test packet; the specific packet format can be seen in Figure 3B. Shown.
  • Step S902 Allocating a receiving timestamp for each test packet in the group of test packets according to the out-of-order of the group of test packets.
  • the network performance detecting device allocates a receiving time stamp for each test packet in the group of test packets according to the out-of-order condition of the group of test packets, so that the calculation of delay and/or jitter is closer to the reality of the user. Feel.
  • FIG. 9B is a schematic flowchart 3 of a network performance detecting apparatus for allocating a suitable receiving timestamp for each test packet according to an embodiment of the present application.
  • the network performance detecting device determines whether i is equal to j+1; wherein i is the packet of the currently received test packet.
  • the identifier is a packet identifier of the test packet to be processed (that is, the network performance detecting device has allocated a suitable receiving time stamp for the j-1 first test packets).
  • the network performance detecting apparatus determines whether the j+1th test message is stored in the cache sequence, wherein the cache The sequence is used to save out-of-order test packets (that is, test packets arriving in advance), and the packet identifier of the out-of-order test packet is greater than j.
  • the j+1th test message is stored in the cache sequence (ie, the j+1th test message arrives at the network performance detecting device earlier than the jth test message), the j+1th test message is The initial receiving timestamp is modified to the receiving timestamp of the i-th (ie, the jth) test packet, and j is incremented by one, and the step of performing the judgment of the j+1th test packet in the cache sequence is returned until the step of executing the judgment If the j+1th test packet is not saved in the cache sequence, the message continues to be received. The next test packet arriving in the test packet.
  • the j+1th test packet in the cache sequence is deleted, so as to cache the sequence.
  • Other test messages arriving in advance can be saved.
  • the network performance detecting device determines whether the cache sequence is full. 2a) If the cache sequence is not full, the ith test packet is saved to the cache sequence, and the next test packet arriving in the test packet is continuously received. 2b) If the buffer sequence is full, it is determined whether the j-th test (ie, the pending test message) is stored in the cache sequence. If the j-th test packet is not saved in the cache sequence, it is determined that the j-th test packet has been lost. According to the average delay of each test packet whose packet identifier is less than j in the test packet, the j-th is determined.
  • the test packet delays and performs a step of determining whether the j+1th test message is stored in the cache sequence.
  • the network performance detecting apparatus as the product of the jth test packet, the product of the average delay of each test packet whose message identifier is less than j in the set of test packets is compared with a preset coefficient (for example, 1.5) Delay.
  • a preset coefficient for example, 1.5
  • the delay of the j-th test packet may be determined by other methods, which is not limited in the embodiment of the present application. .
  • the network performance detecting device puts a suitable receiving time stamp for each test packet of the received test packet, wherein a reasonable delay is given for the lost test packet.
  • Step S903 Determine a delay or jitter of the test path according to the receiving timestamp and the sending timestamp of each test packet.
  • the network performance detecting apparatus sends a timestamp according to the receiving timestamp of each test packet, that is, the appropriate receiving timestamp allocated for each test packet according to the out-of-order of the group of test packets. Determine the network performance delay or jitter of the test path.
  • a timestamp according to the receiving timestamp of each test packet, that is, the appropriate receiving timestamp allocated for each test packet according to the out-of-order of the group of test packets. Determine the network performance delay or jitter of the test path.
  • the receiving time stamp is allocated for each test message according to the out-of-order order of the group of test messages; further, according to each test message Receive timestamps (that is, appropriate receiving timestamps assigned to each test packet according to the out-of-order of the test packets), and send a timestamp, so that the measurement result of the network performance of the obtained test path is more accurate.
  • the network performance detecting apparatus allocates a suitable receiving timestamp for each test packet, and may also refer to the network performance detecting apparatus shown in FIG. 7 to allocate an appropriate one for each test packet.
  • the process of receiving the timestamp is shown in FIG. 2, and the network performance detecting apparatus described in FIG. 8 allocates a suitable receiving timestamp for each test packet, and details are not described herein again.
  • FIG. 10 is a schematic structural diagram of Embodiment 1 of a network performance detecting apparatus according to the present application.
  • the terminal network performance detecting apparatus 100 provided by this embodiment includes:
  • the sending module 1001 is configured to send a first group of test messages and a second group of test messages, where the first group of test messages is used to test the first test path, and the second group of test messages is used for testing.
  • a second test path the first test path includes: the second test path and a link to be tested;
  • the receiving module 1002 is configured to receive a first group of test packets returned on the first test path, and a second group of test packets returned on the second test path, where the first group of test packets Oscillation on the link to be tested according to a preset number of oscillations;
  • the determining module 1003 is configured to determine network performance of the link to be tested according to the received first group of test messages and the second group of test messages.
  • the device further includes:
  • a generating module configured to generate the first group of test packets according to the preset number of oscillations
  • Each of the first test packets of the first group of test packets carries: at least one first link label group and at least one second link label group corresponding to the link to be tested;
  • the first link label group includes M identical first link labels, each of the second link label groups includes M identical second link labels; the value of the M is equal to the pre-
  • the first link label is used to indicate the next hop target node of the first test packet in the forward transmission direction, and the second link label is used to indicate the first test packet. The next hop target node in the reverse direction of transmission.
  • the determining module 1003 includes:
  • a first determining submodule configured to determine a delay or jitter of the first test path according to time information carried in each first test packet of the received first test packet;
  • a second determining submodule configured to determine a delay or jitter of the second test path according to the time information carried in each second test packet in the received second test packet;
  • a third determining submodule configured to determine a delay of the link to be tested according to a delay of the first test path, a delay of the second test path, and the preset number of oscillations; or Determining the jitter of the link to be tested according to the jitter of the first test path, the jitter of the second test path, and the preset number of oscillations.
  • the determining module 1003 includes:
  • the fourth determining sub-module is configured to determine the first test path according to the number of the first test packet in the first group of test packets sent and the number of the first test packet in the first test packet received Packet loss rate;
  • the fifth determining submodule is configured to determine the second test path according to the number of the second test packet in the sent second test packet and the number of the second test packet in the received second test packet Packet loss rate;
  • a sixth determining submodule configured to determine, according to the packet loss rate of the first test path, the packet loss rate of the second test path, and the preset number of oscillations, a packet loss rate of the link to be tested .
  • the determining module 1003 includes:
  • a seventh determining sub-module configured to determine, according to the packet identifier ID and the first preset packet identifier of the test packet carried in each first test packet in the received first test packet, The out-of-order number of the test packet is determined, and the out-of-order rate of the first test path is determined according to the quotient of the number of the out-of-order and the number of the first test packets sent on the first test path.
  • An eighth determining sub-module configured to determine, according to the packet ID of the test packet and the second preset packet identifier carried in each second test packet of the received second test packet, the second group of tests An out-of-order number of the packet, and determining an out-of-order rate of the second test path according to the quotient of the number of the out-of-order and the number of the second test packet sent on the second test path;
  • a ninth determining submodule configured to determine an out-of-order rate of the link to be tested according to the out-of-order rate of the first test path, the out-of-order rate of the second test path, and the preset number of oscillations .
  • the first determining submodule includes:
  • An allocating unit configured to allocate a receiving timestamp for each first test packet according to the out-of-order of the first group of test messages
  • a determining unit configured to determine a delay or a jitter of the first test path according to a receiving timestamp and a sending timestamp of each of the first test messages.
  • the allocating unit is specifically configured to:
  • the receiving module When the receiving module receives the ith first test packet in the first group of test packets, it is determined whether i is equal to j; wherein i is the currently received first test packet Packet identifier; j is the packet identifier of the packet to be processed;
  • i is equal to j, it is determined whether the j+1th first test message is saved in the cache sequence; wherein the cache sequence is used to save the out-of-order first test message, the first test of the out-of-order
  • the message identifier of the message is greater than j;
  • the initial receiving timestamp of the j+1th first test packet is modified to the ith first test packet. Receiving a timestamp, and adding 1 to j, returning to perform the step of determining whether the j+1th first test message is saved in the cache sequence until the j+1 is not saved in the cache sequence. The first test packet, the receiving module continues to receive the first test packet that arrives in the first group of test packets.
  • the allocating unit is further configured to:
  • the allocating unit after modifying the initial receiving timestamp of the j+1th first test packet to the receiving timestamp of the ith first test packet, is further used to: The j+1th first test message in the cache sequence is deleted.
  • the network performance detecting apparatus of this embodiment may be used to perform the technical solution of any of the foregoing network performance detecting methods in the foregoing embodiments, and the implementation principles and technical effects are similar, and details are not described herein again.
  • FIG. 11 is a schematic structural diagram of Embodiment 2 of a network performance detecting apparatus according to the present application.
  • the network performance detecting apparatus 110 provided in this embodiment may include a processor 1101, a memory 1102, a receiver 1103, and a transmitter 1104.
  • the memory 1102, the receiver 1103, and the transmitter 1104 are all connected to the processor 1101.
  • the transmitter 1104 is configured to send a first group of test packets and a second group of test packets, where the first group of test packets is used to test a first test path, and the second group of test packets is used to Testing the second test path; the first test path includes: the second test path and the link to be tested.
  • the receiver 1103 is configured to receive a first group of test packets returned on the first test path, and a second group of test packets returned on the second test path, where the first group of test packets is based on The preset number of oscillations oscillates on the link to be tested.
  • the memory 1102 is configured to store the execution instruction
  • the processor 1101 is configured to execute the execution instruction in the memory 1102, to perform the following operations: determining, according to the received first set of test messages and the second set of test messages, Describe the network performance of the link to be tested.
  • processor 1101 is further configured to:
  • Each of the first test packets of the first group of test packets carries: at least one first link label group and at least one second link label group corresponding to the link to be tested;
  • the first link label group includes M identical first link labels, each of the second link label groups includes M identical second link labels; the value of the M is equal to the pre-
  • the first link label is used to indicate the next hop target node of the first test packet in the forward transmission direction, and the second link label is used to indicate the first test packet. The next hop target node in the reverse direction of transmission.
  • processor 1101 is further configured to:
  • the jitter, the jitter of the second test path, and the preset number of oscillations determine the jitter of the link to be tested.
  • processor 1101 is further configured to:
  • processor 1101 is further configured to:
  • the out-of-order rate of the link to be tested is determined according to the out-of-order rate of the first test path, the out-of-order rate of the second test path, and the preset number of oscillations.
  • processor 1101 is further configured to:
  • Determining the delay or jitter of the first test path according to the receiving timestamp and the sending timestamp of each of the first test packets.
  • processor 1101 is further configured to:
  • i is the currently received packet identifier of the first test packet ;
  • j is the message identifier of the message to be processed;
  • i is equal to j, it is determined whether the j+1th first test message is saved in the cache sequence; wherein the cache sequence is used to save the out-of-order first test message, the first test of the out-of-order
  • the message identifier of the message is greater than j;
  • the initial receiving timestamp of the j+1th first test packet is modified to the ith first test packet. Receiving a timestamp, and adding 1 to j, returning to perform the step of determining whether the j+1th first test message is saved in the cache sequence until the j+1 is not saved in the cache sequence. The first test packet continues to receive the first test packet that arrives in the first group of test packets.
  • processor 1101 is further configured to:
  • the processor 1101 is configured to: after the initial receiving timestamp of the j+1th first test packet is changed to the receiving timestamp of the ith first test packet, The j+1th first test message in the cache sequence is deleted.
  • the network performance detecting apparatus of this embodiment may be used to perform the technical solution of any of the foregoing network performance detecting methods in the foregoing embodiments, and the implementation principles and technical effects are similar, and details are not described herein again.
  • FIG. 12 is a schematic structural diagram of Embodiment 3 of a network performance detecting apparatus according to the present application.
  • the terminal network performance detecting apparatus 120 provided in this embodiment includes:
  • the receiving module 1201 is configured to receive a set of test messages returned on the test path;
  • the allocating module 1202 is configured to allocate a receiving timestamp for each test packet in the group test packet according to the out-of-order of the group test message;
  • the determining module 1203 is configured to determine a delay or jitter of the test path according to the receiving timestamp and the sending timestamp of each test packet.
  • the allocation module 1202 is specifically configured to:
  • the receiving module When the receiving module receives the ith test packet in the group test packet, it determines whether i is equal to j; where i is the packet identifier of the currently received test packet; j is the packet of the packet to be processed. Identification
  • i is equal to j, it is determined whether the j+1th test message is stored in the cache sequence; wherein the cache sequence is used to save the out-of-order test message, and the message identifier of the out-of-order test message is greater than j;
  • the initial receiving timestamp of the j+1th test packet is changed to the receiving timestamp of the i th test packet, and j is incremented by 1, and returned.
  • the step of determining whether the j+1th test message is saved in the cache sequence is performed until the j+1th test message is not saved in the cache sequence, and the receiving module continues to receive the next test in the group test message. Message.
  • the allocation module 1202 is further configured to:
  • the ith test packet is saved in the cache sequence, and the receiving module continues to receive the next test packet in the group test packet;
  • the j-th test packet is not saved in the cache sequence, it is determined that the j-th test packet is lost, and the j-th is determined according to the average delay of each test packet whose packet identifier is less than j in the group test packet. The delay of the packet is tested, and a step of determining whether the j+1th test packet is saved in the cache sequence is performed.
  • the allocation module 1202 modifies the initial receiving timestamp of the j+1th test packet to the receiving timestamp of the ith test packet, the allocation module 1202 is further configured to:
  • the network performance detecting apparatus of this embodiment may be used to implement the technical solution of the sixth embodiment of the network performance detecting method in the foregoing application, and the implementation principle and technical effects are similar, and details are not described herein again.
  • FIG. 13 is a schematic structural diagram of Embodiment 4 of a network performance detecting apparatus according to the present application.
  • the network performance detecting apparatus 130 provided in this embodiment may include a processor 1301, a memory 1302, and a transceiver 1303.
  • the memory 1302 and the transceiver 1303 are both connected to the processor 1301.
  • the transceiver 1303 is configured to receive a set of test messages returned on the test path.
  • Storage The processor 1301 is configured to execute the execution instruction in the memory 1302, and is configured to perform the following operations: each test message in the group test message according to the out-of-order of the group test message The receiving timestamp is allocated; determining the delay or jitter of the test path according to the receiving timestamp and the sending timestamp of each of the test messages.
  • processor 1301 is further configured to:
  • i is the packet identifier of the currently received test packet
  • j is a to-be-processed packet Message identification of the text
  • i is equal to j, it is determined whether the j+1th test message is stored in the cache sequence, wherein the cache sequence is used to save the out-of-order test message, and the message identifier of the out-of-order test packet is Greater than j;
  • the initial receiving timestamp of the j+1th test packet is modified to the receiving timestamp of the ith test packet, and Adding j to 1, and returning to perform the step of determining whether the j+1th test message is stored in the cache sequence, until the j+1th test message is not stored in the cache sequence, and then continuing to receive The next arriving test message in the group test message.
  • processor 1301 is further configured to:
  • the j-th test packet is not saved in the cache sequence, it is determined that the j-th test packet is lost, and the packet identifier in the group test packet is smaller than the test packet of the j.
  • the average delay of the text determines the delay of the jth test message, and performs a step of determining whether the j+1th test message is stored in the cache sequence.
  • the processor 1301 is further configured to: The j+1th test message is deleted.
  • the network performance detecting apparatus of this embodiment may be used to perform the foregoing network performance detection of the present application.
  • the technical solution of the sixth embodiment of the method has similar implementation principles and technical effects, and details are not described herein again.
  • the disclosed apparatus and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of cells is only a logical function division.
  • multiple units or components may be combined or integrated. Go to another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units.
  • the above-described integrated unit implemented in the form of a software functional unit can be stored in a computer readable storage medium.
  • the software functional unit described above is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform the methods described in various embodiments of the present application. Part of the steps.
  • the aforementioned program can be stored in a computer readable storage medium.
  • the program when executed, performs the steps including the foregoing method embodiments; and the foregoing storage medium includes various media 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)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本申请实施例提供一种网络性能检测方法及装置。该方法包括：发送第一组测试报文以及第二组测试报文；接收第一测试路径上返回的第一组测试报文，以及第二测试路径上返回的第二组测试报文；其中，第一组测试报文根据预设的振荡次数在待测链路上振荡；进一步地，根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。可见，本申请实施例中，通过第一测试报文在待测链路上振荡的方式，增加了待测链路在整个第一测试路径中的比重(即待测链路的比重大于上游链路的比重)，从而减小了上游链路的网络性能的波动对待测链路网络性能测量的影响，使得待测链路的网络性能的测量结果更加准确。

Description

网络性能检测方法及装置 技术领域
本申请涉及互联网技术，尤其涉及一种网络性能检测方法及装置。
背景技术
随着互联网技术的发展，运营商为了降低网络运营成本和满足业务发展需求，对整网进行网络之间互连的协议(Internet Protocol，简称IP)化改造(全IP是目前和未来网络的发展趋势)。但IP技术是一种简单的无连接协议，并不能保证传输的可靠性。
随着IP网络承载越来越多的新型互联网业务(例如各种视频、语音、游戏等)，终端用户对互联网业务感知的体验要求越来越高，从而运营商对IP网络性能的要求也越来越高，如何测量和优化IP网络性能是运营商当前及未来面临的挑战之一。因此，IP网络性能监测技术应用而生。
由于检测点数量与位置的限制，链路的网络性能只能通过测试路径的网络性能计算得到，因此，后一跳链路的网络性能的测量会依赖于前一跳的网络性能，导致后一跳链路的网络性能的测量结果会存在较大误差。图1为现有IP网络性能监测的应用场景示意图。如图1所示，假设节点P1与P2之间双向链路的丢包率均为a，P2与P3之间双向链路的丢包率为b；根据测试路径L1,2＝<P0,P1,P2,P1,P0>确定a、根据
Figure PCTCN2016112734-appb-000001
确定测试路径L1,2，3＝<P0,P1,P2,P3,P2,P1,P0>的丢包率；进一步地，根据a以及测试路径L1,2，3的丢包率确定链路<P2,P3>的丢包率
Figure PCTCN2016112734-appb-000002
由于实际网络的网络性能是实时变化的，测试路径L1,2与测试路径L1,2，3的检测时间通常不一致，会导致链路<P2,P3>的丢包率的测量结果误差较大；例如，当检测测试路径L1,2的丢包率时，链路<P1,P2>的丢包率为a，但在检测测试路径L1,2，3的丢包率时，链路<P1,P2>的实际丢包率变为a1，因此，在计算b时，若仍认为<P1,P2>的丢包率为a，则计算得到的b会存在较大误差。
发明内容
本申请实施例提供一种网络性能检测方法及装置，提高了待测链路的网络性能的测量结果的准确度。
第一方面，本申请实施例提供一种网络性能检测方法，包括：
发送第一组测试报文以及第二组测试报文；其中，第一组测试报文用于测试第一测试路径，第二组测试报文用于测试第二测试路径；第一测试路径包括：第二测试路径以及待测链路；
接收第一测试路径上返回的第一组测试报文，以及第二测试路径上返回的第二组测试报文；其中，第一组测试报文根据预设的振荡次数在待测链路上振荡；
根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。
通过第一方面提供的网络性能检测方法，通过发送用于测试第一测试路径的第一组测试报文以及用于测试第二测试路径的第二组测试报文，其中，第一测试路径包括：第二测试路径以及待测链路；进一步地，接收第一测试路径上返回的第一组测试报文(第一组测试报文根据预设的振荡次数在待测链路上振荡)，以及第二测试路径上返回的第二组测试报文，并根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。可见，本申请实施例中，通过第一测试报文在待测链路上振荡的方式，增加了待测链路在整个第一测试路径中的比重(即待测链路的比重大于上游链路的比重)，从而减小了上游链路的网络性能(如第二测试路径的网络性能)的波动对待测链路网络性能测量的影响，使得待测链路的网络性能的测量结果更加准确。
在一个可能的设计中，发送第一组测试报文以及第二组测试报文之前，还包括：
根据预设的振荡次数生成第一组测试报文；
其中，第一组测试报文中的每个第一测试报文携带：待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个第一链路标签组包括M个相同的第一链路标签，每个第二链路标签组包括M个相同的第二链路标签；M的数值等于预设的振荡次数，第一链路标签用于指示第一测试报文在正向传输方向的下一跳目标节点，第二链路标签用于 指示第一测试报文在反向传输方向的下一跳目标节点。
在一个可能的设计中，根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能，包括：
根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定第一测试路径的时延或抖动；
根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定第二测试路径的时延或抖动；
根据第一测试路径的时延、第二测试路径的时延以及预设的振荡次数，确定待测链路的时延；或者，根据第一测试路径的抖动、第二测试路径的抖动以及预设的振荡次数，确定待测链路的抖动。
在一个可能的设计中，根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能，包括：
根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
根据第一测试路径的丢包率、第二测试路径的丢包率以及预设的振荡次数，确定待测链路的丢包率。
在一个可能的设计中，根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能，包括：
根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定第一组测试报文的乱序数，并根据乱序数与第一测试路径上发送的第一测试报文的个数的商，确定第一测试路径的乱序率；
根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定第二组测试报文的乱序数，并根据乱序数与第二测试路径上发送的第二测试报文的个数的商，确定第二测试路径的乱序率；
根据第一测试路径的乱序率、第二测试路径的乱序率以及预设的振荡次数，确定待测链路的乱序率。
在一个可能的设计中，根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定第一测试路径的时延或抖动，包括：
根据第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
根据每个第一测试报文的接收时间戳、发送时间戳，确定第一测试路径的时延或抖动。
在一个可能的设计中，根据第一组测试报文的乱序为每个第一测试报文分配接收时间戳，包括：
当接收到第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，i为当前接收到的第一测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，缓存序列用于保存乱序的第一测试报文，乱序的第一测试报文的报文标识大于j；
若缓存序列中保存有第j+1个第一测试报文，则将第j+1个第一测试报文的初始接收时间戳修改为第i个第一测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤，直至缓存序列中未保存有第j+1个第一测试报文，则继续接收第一组测试报文中下一个到达的第一测试报文。
在一个可能的设计中，判断i是否等于j之后，还包括：
若i不等于j，则判断缓存序列中是否已满；
若缓存序列中未满，则将第i个第一测试报文保存至缓存序列中，并继续接收第一组测试报文中下一个到达的第一测试报文；
若缓存序列中已满，则判断缓存序列中是否保存有第j个第一测试报文；
若缓存序列中未保存有第j个第一测试报文，则确定第j个第一测试报文已丢失，根据第一组测试报文中报文标识小于j的各第一测试报文的平均时延，确定第j个第一测试报文的时延，并执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤。
在一个可能的设计中，将第j+1个第一测试报文的初始接收时间戳修改为第i个第一测试报文的接收时间戳之后，还包括：
将缓存序列中的第j+1个第一测试报文删除。
通过该实施方式提供的网络性能检测方法，通过根据第一组测试报文的乱序为每个第一测试报文分配接收时间戳；进一步地，根据每个第一测试报文的接收时间戳(即根据第一组测试报文的乱序为每个第一测试报文所分配的合适的接收时间戳)、发送时间戳，计算所得到的第一测试路径的时延或抖动的测量结果更加精确，从而进一步提高了待测链路的时延或抖动测量结果的准确度。
第二方面，本申请实施例提供一种网络性能检测装置，包括：
发送模块，用于发送第一组测试报文以及第二组测试报文；其中，第一组测试报文用于测试第一测试路径，第二组测试报文用于测试第二测试路径；第一测试路径包括：第二测试路径以及待测链路；
接收模块，用于接收第一测试路径上返回的第一组测试报文，以及第二测试路径上返回的第二组测试报文；其中，第一组测试报文根据预设的振荡次数在待测链路上振荡；
确定模块，用于根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。
在一个可能的设计中，装置还包括：
生成模块，用于根据预设的振荡次数生成第一组测试报文；
其中，第一组测试报文中的每个第一测试报文携带：待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个第一链路标签组包括M个相同的第一链路标签，每个第二链路标签组包括M个相同的第二链路标签；M的数值等于预设的振荡次数，第一链路标签用于指示第一测试报文在正向传输方向的下一跳目标节点，第二链路标签用于指示第一测试报文在反向传输方向的下一跳目标节点。
在一个可能的设计中，确定模块，包括：
第一确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定第一测试路径的时延或抖动；
第二确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定第二测试路径的时延或抖动；
第三确定子模块，用于根据第一测试路径的时延、第二测试路径的时 延以及预设的振荡次数，确定待测链路的时延；或者，根据第一测试路径的抖动、第二测试路径的抖动以及预设的振荡次数，确定待测链路的抖动。
在一个可能的设计中，确定模块，包括：
第四确定子模块，用于根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
第五确定子模块，用于根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
第六确定子模块，用于根据第一测试路径的丢包率、第二测试路径的丢包率以及预设的振荡次数，确定待测链路的丢包率。
在一个可能的设计中，确定模块，包括：
第七确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定第一组测试报文的乱序数，并根据乱序数与第一测试路径上发送的第一测试报文的个数的商，确定第一测试路径的乱序率；
第八确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定第二组测试报文的乱序数，并根据乱序数与第二测试路径上发送的第二测试报文的个数的商，确定第二测试路径的乱序率；
第九确定子模块，用于根据第一测试路径的乱序率、第二测试路径的乱序率以及预设的振荡次数，确定待测链路的乱序率。
在一个可能的设计中，第一确定子模块，包括：
分配单元，用于根据第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
确定单元，用于根据每个第一测试报文的接收时间戳、发送时间戳，确定第一测试路径的时延或抖动。
在一个可能的设计中，分配单元具体用于：
当接收模块接收到第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，i为当前接收到的第一测试报文的报文标识；j为待处 理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，缓存序列用于保存乱序的第一测试报文，乱序的第一测试报文的报文标识大于j；
若缓存序列中保存有第j+1个第一测试报文，则将第j+1个第一测试报文的初始接收时间戳修改为第i个第一测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤，直至缓存序列中未保存有第j+1个第一测试报文，则接收模块继续接收第一组测试报文中下一个到达的第一测试报文。
在一个可能的设计中，分配单元还用于：
若i不等于j，则判断缓存序列中是否已满；
若缓存序列中未满，则将第i个第一测试报文保存至缓存序列中，接收模块继续接收第一组测试报文中下一个到达的第一测试报文；
若缓存序列中已满，则判断缓存序列中是否保存有第j个第一测试报文；
若缓存序列中未保存有第j个第一测试报文，则确定第j个第一测试报文已丢失，根据第一组测试报文中报文标识小于j的各第一测试报文的平均时延，确定第j个第一测试报文的时延，并执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤。
在一个可能的设计中，分配单元将第j+1个第一测试报文的初始接收时间戳修改为第i个第一测试报文的接收时间戳之后，还用于：将缓存序列中的第j+1个第一测试报文删除。
上述第二方面的实施方式所提供的网络性能检测装置，其有益效果可以参见上述第一方面的各可能的实施方式所带来的有益效果，在此不再赘述。
第三方面，本申请实施例提供一种网络性能检测方法，包括：
接收测试路径上返回的一组测试报文；
根据组测试报文的乱序为组测试报文中每个测试报文分配接收时间戳；
根据每个测试报文的接收时间戳、发送时间戳，确定测试路径的时延或抖动。
通过第一方面提供的网络性能检测方法，在接收测试路径上返回的一组测试报文后，通过根据该组测试报文的乱序为每个测试报文分配接收时间戳；进一步地，根据每个测试报文的接收时间戳(即根据该组测试报文的乱序为每个测试报文所分配的合适的接收时间戳)、发送时间戳，从而使得计算所得到的测试路径的网络性能的测量结果更加精确。
在一个可能的设计中，根据组测试报文的乱序为组测试报文中每个测试报文分配接收时间戳，包括：
当接收到组测试报文中的第i个测试报文时，判断i是否等于j；其中，i为当前接收到的测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个测试报文；其中，缓存序列用于保存乱序的测试报文，乱序的测试报文的报文标识大于j；
若缓存序列中保存有第j+1个测试报文，则将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个测试报文的步骤，直至缓存序列中未保存有第j+1个测试报文，则继续接收组测试报文中下一个到达的测试报文。
在一个可能的设计中，i是否等于j之后，还包括：
若i不是否等于j，则判断缓存序列中是否已满；
若缓存序列中未满，则将第i个测试报文保存至缓存序列中，并继续接收组测试报文中下一个到达的测试报文；
若缓存序列中已满，则判断缓存序列中是否保存有第j个测试报文；
若缓存序列中未保存有第j个测试报文，则确定第j个测试报文已丢失，根据组测试报文中报文标识小于j的各测试报文的平均时延，确定第j个测试报文的时延，并执行判断缓存序列中是否保存有第j+1个测试报文的步骤。
在一个可能的设计中，将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳之后，还包括：
将缓存序列中的第j+1个测试报文删除。
第四方面，本申请实施例提供一种网络性能检测装置，包括：
接收模块，用于接收测试路径上返回的一组测试报文；
分配模块，用于根据组测试报文的乱序为组测试报文中每个测试报文分配接收时间戳；
确定模块，用于根据每个测试报文的接收时间戳、发送时间戳，确定测试路径的时延或抖动。
在一个可能的设计中，分配模块具体用于：
当接收模块接收到组测试报文中的第i个测试报文时，判断i是否等于j；其中，i为当前接收到的测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个测试报文；其中，缓存序列用于保存乱序的测试报文，乱序的测试报文的报文标识大于j；
若缓存序列中保存有第j+1个测试报文，则将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个测试报文的步骤，直至缓存序列中未保存有第j+1个测试报文，则接收模块继续接收组测试报文中下一个到达的测试报文。
在一个可能的设计中，分配模块还用于：
若i不是否等于j，则判断缓存序列中是否已满；
若缓存序列中未满，则将第i个测试报文保存至缓存序列中，接收模块继续接收组测试报文中下一个到达的测试报文；
若缓存序列中已满，则判断缓存序列中是否保存有第j个测试报文；
若缓存序列中未保存有第j个测试报文，则确定第j个测试报文已丢失，根据组测试报文中报文标识小于j的各测试报文的平均时延，确定第j个测试报文的时延，并执行判断缓存序列中是否保存有第j+1个测试报文的步骤。
在一个可能的设计中，分配模块将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳之后，还用于：
将缓存序列中的第j+1个测试报文删除。
上述第四方面的实施方式所提供的网络性能检测装置，其有益效果可以参见上述第三方面的各可能的实施方式所带来的有益效果，在此不再赘述。
附图说明
图1为现有IP网络性能监测的应用场景示意图；
图2A为本申请网络性能检测方法实施例一的流程示意图；
图2B为本申请网络性能检测方法的应用场景示意图；
图3A为本申请网络性能检测方法实施例二的流程示意图；
图3B为本申请测试报文的格式示意图；
图4为本申请网络性能检测方法实施例三的流程示意图；
图5A为本申请网络性能检测方法实施例四的流程示意图；
图5B为本申请网络性能检测方法实施例五的流程示意图；
图6A为本申请网络性能检测方法实施例六的流程示意图；
图6B为本申请实施例中网络性能检测装置为每个第一测试报文分配合适的接收时间戳的流程示意图；
图7为本申请实施例中网络性能检测装置为每个测试报文分配合适的接收时间戳的流程示意图；
图8为本申请网络性能检测方法一实施例中接收到的测试报文序列以及缓存序列的示意图；
图9A为本申请网络性能检测方法实施例七的流程示意图；
图9B为本申请实施例中网络性能检测装置为每个测试报文分配合适的接收时间戳的流程示意图三；
图10为本申请网络性能检测装置实施例一的结构示意图；
图11为本申请网络性能检测装置实施例二的结构示意图；
图12为本申请网络性能检测装置实施例三的结构示意图；
图13为本申请网络性能检测装置实施例四的结构示意图。
具体实施方式
首先对现有技术中计算测试路径的丢包率、时延、抖动以及乱序等网络性能的方式进行详细介绍：
1)丢包率：
由于丢包率是一个统计值，因此，无法从单个测试报文中获得测试路径上的丢包率，需要从一组测试报文中获得：
plr＝1-Nrecv/Nsend        (1)
其中，plr代表丢包率，Nsend、Nrecv分别代表测试报文的发送数量和接收数量。
2)时延：
单个测试报文的时延可以表示为
delayi＝Ti,recv-Ti,send        (2)
其中，delayi代表第i个测试报文的时延，Ti,recv代表第i个测试报文的接收时间，Ti,send代表第i个测试报文的发送时间。
整条测试路径的时延可以由该条测试路径上所有测试报文的时延取均值获得：
Figure PCTCN2016112734-appb-000003
其中，delay代表测试路径的时延，N代表测试路径上所接收到的所有测试报文的数量，N为大于等于1的正整数。
3)抖动
测试报文在传输路径上的时延存在差异，而这些差异就是测试报文的抖动，两个测试报文之间的抖动可以定义为：
jitteri＝abs(delayi+1-delayi)        (4)
其中，abs(_)代表取绝对值，delayi+1代表第i+1个测试报文的时延，delayi代表第i个测试报文的时延，jitteri代表第i+1个测试报文与第i个测试报文之间的抖动。
整条测试路径上的抖动通过对所有测试报文的抖动取均值获得：
Figure PCTCN2016112734-appb-000004
其中，jitter代表测试路径的抖动。
4)乱序率：
乱序率和丢包率类似，单个测试报文的乱序没有意义，通过定义缓存序列来获得乱序数，乱序数等于累积进入缓存序列的测试报文的数量；进一步地，乱序率＝乱序数/测试路径上发送的所有测试报文的个数。
如图1所示，由于检测点数量与位置的限制，链路的网络性能只能通过测试路径的网络性能计算得到，因此，后一跳链路的网络性能的测量会依赖于前一跳的网络性能，且上游链路在整个测试路径上所占的比重大于 或者等于下游被测链路所占的比重，即上游链路的网络性能的波动会严重影响下游链路网络性能的测量，使得下游链路的网络性能的测量结果误差较大。
图2A为本申请网络性能检测方法实施例一的流程示意图。本实施例的执行主体可以为配置在服务器中的网络性能检测装置，该装置可以通过软件和/或硬件实现。如图2A所示，本实施例的方法可以包括：
步骤S201、发送第一组测试报文以及第二组测试报文。
图2B为本申请网络性能检测方法的应用场景示意图，如图2B所示的网络为全IP网络，网络中运行分段路由(Segment routing，简称SR)协议。其中，P0代表网络性能检测装置，P1、P2和P3代表路由器或者转发器。为了便于描述，图2B中示出第一测试路径L1<P0、P1、P2、P3、P2、P3、P2、P1、P0>和第二测试路径L2<P0、P1、P2、P1、P0>，第一测试路径L1包括第二测试路径L2以及待测链路L3。当然，本申请网的络性能检测方法还可应用于其它形式的网络结构中，本申请实施例对此并不作限制。
本步骤中，如图2B所示，网络性能检测装置发送第一组测试报文以及第二组测试报文；其中，第一组测试报文用于测试第一测试路径L1，第二组测试报文用于测试第二测试路径L2；第一测试路径L1包括：第二测试路径L2以及待测链路L3。可选地，第一组测试报文中的每个第一测试报文中携带有测试路径标识1(用于指示第一测试报文检测的第一测试路径L1)，第二组测试报文中的每个第二测试报文中携带有测试路径标识2(用于指示第二测试报文检测的第二测试路径L2)。
步骤S202、接收第一测试路径上返回的第一组测试报文，以及第二测试路径上返回的第二组测试报文。
本步骤中，网络性能检测装置接收第一测试路径L1上返回的第一组测试报文，以及第二测试路径L2上返回的第二组测试报文；其中，第一组测试报文根据预设的振荡次数在待测链路上振荡，如图2B所示，以预设振荡次数等于2为例，第一组测试报文中每个第一测试报文会在待测链路L3上振荡(如P2->P3、P3->P2、P2->P3、P3->P2)。可见，本申请实施例中，待测链路的比重远远大于上游链路的比重，从而减小了上游链路 的网络性能的波动对待测链路网络性能测量的影响。
步骤S203、根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。
本步骤中，网络性能检测装置根据接收的第一组测试报文以及第二组测试报文，确定待测链路L3的网络性能。可选地，网络性能检测装置根据接收的第一组测试报文确定第一测试路径L1的网络性能，以及根据接收的第二组测试报文确定第二测试路径L2的网络性能；进一步地，根据第一测试路径L1的网络性能、第二测试路径L2的网络性能以及预设的振荡次数，确定待测链路的网络性能。可见，本申请实施例中，通过第一组测试报文根据预设的振荡次数在所述待测链路上振荡的方式，使得待测链路的比重远远大于上游链路(例如，如图2B中链路P1->P2、链路P2->P1)的比重，从而减小了上游链路的网络性能的波动对待测链路网络性能测量的影响，减小了待测链路的网络性能的测量结果误差。
可选地，网络性能包括以下一项或多项：时延、抖动、丢包率、乱序率。当然，本申请涉及的网络性能还可以包括其它网络性能，本申请实施例中对此并不作限制。
本申请实施例中，通过发送用于测试第一测试路径的第一组测试报文以及用于测试第二测试路径的第二组测试报文，其中，第一测试路径包括：第二测试路径以及待测链路；进一步地，接收第一测试路径上返回的第一组测试报文(第一组测试报文根据预设的振荡次数在待测链路上振荡)，以及第二测试路径上返回的第二组测试报文，并根据接收的第一组测试报文以及第二组测试报文，确定待测链路的网络性能。可见，本申请实施例中，通过第一测试报文在待测链路上振荡的方式，增加了待测链路在整个第一测试路径中的比重(即待测链路的比重大于上游链路的比重)，从而减小了上游链路的网络性能(如第二测试路径的网络性能)的波动对待测链路网络性能测量的影响，使得待测链路的网络性能的测量结果更加准确。
图3A为本申请网络性能检测方法实施例二的流程示意图，图3B为本申请测试报文的格式示意图。如图3A所示，在上述实例的基础上，步骤S201之前，还包括：
步骤S204、根据预设的振荡次数生成第一组测试报文。
本实施例中，在发送第一组测试报文以及第二组测试报文之前，网络性能检测装置根据预设的振荡次数生成第一组测试报文；其中，第一组测试报文中的每个第一测试报文携带：待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；每个第一链路标签组包括M个相同的第一链路标签，每个第二链路标签组包括M个相同的第二链路标签；M的数值等于预设的振荡次数，第一链路标签用于指示第一测试报文在正向传输方向(即沿着网络性能检测装置发送测试报文的方向)的下一跳目标节点，第二链路标签用于指示第一测试报文在反向传输方向(即沿着网络性能检测装置接收测试报文的方向)的下一跳目标节点。可选地，待测链路对应的第一链路标签组所包括的第一链路标签和第二链路标签组所包括的第二链路标签都可携带于第一测试报文的SR报头中(如图3B所示)，以便待测链路上的节点在收到第一测试报文后，根据SR报头中包括的链路标签转发第一测试报文，使得第一测试报文在待测链路上振荡，增加了待测链路在整个第一测试路径中的比重(即待测链路的比重大于上游链路的比重)，从而减小了上游链路的网络性能的波动对待测链路网络性能测量的影响。
例如：如图2B所示，预设的振荡次数为2，第一测试报文在待测链路L3上振荡(如P2->P3、P3->P2、P2->P3、P3->P2)，每个第一测试报文携带：包括2个相同第一链路标签(如P2->P3链路标签)的第一链路标签组，以及包括2个相同第二链路标签(如P3->P2链路标签)的第二链路标签组。又例如：预设的振荡次数为2，第一测试报文在待测链路上振荡(如P2->P3、P3->P4、P4->P3、P3->P2、P2->P3、P3->P4、P4->P3、P3->P2)，每个第一测试报文携带：包括2个相同第一链路标签(如P2->P3链路标签)的第一链路标签组1、包括2个相同第一链路标签(如P3->P4链路标签)的第一链路标签组2、包括2个相同第二链路标签(如P3->P2链路标签)的第二链路标签组1以及包括2个相同第二链路标签(如P4->P3链路标签)的第二链路标签组2。
为了区别多个网络性能检测装置所发出的测试报文，可选地，第一组测试报文中的每个第一测试报文和第二组测试报文中的每个第二测试报 文可携带：探测器标识；探测器标识用于指示发出对应测试报文的网络性能检测装置。例如，第一测试报文携带的探测器标识用于指示发出第一测试报文的网络性能检测装置，第二测试报文携带的探测器标识用于指示发出第二测试报文的网络性能检测装置。
如图3B所示，测试报文包括：路径信息字段(用于控制测试报文的转发路径)、检测信息字段以及可选字段。其中，路径信息字段包括：SR报头(包括SR标签)和IP报头。检测信息字段包括用于计算测试路径上网络性能所必须的信息，例如：测试报文标识用于唯一标志测试报文(即过滤其它无关报文)、探测器标识用于标识该测试报文是由哪个网络性能检测装置发出的、测试报文路径标识用于标识测试报文负责检测哪一条测试路径、测试报文的报文身份标识(Identity，简称ID)用于区分同一测试路径上的不同测试报文、发送时间戳用于指示测试报文的发送时间、接收时间戳用于指示测试报文的接收时间。
图4为本申请网络性能检测方法实施例三的流程示意图。如图4所示，在上述实施例的基础上，步骤S203包括：
步骤S203A1、根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定第一测试路径的时延或抖动。
可选地，每个第一测试报文都会携带发送时间戳和接收时间戳，其中，发送时间戳用于指示网络性能检测装置发送第一测试报文的发送时间，接收时间戳用于指示网络性能检测装置接收第一测试报文的接收时间。
本步骤中，可选地，网络性能检测装置根据接收的第一组测试报文中每个第一测试报文所携带的时间信息(例如，发送时间戳和接收时间戳)，通过上述公式(3)确定第一测试路径L1的时延。
可选地，网络性能检测装置根据接收的第一组测试报文中每个第一测试报文所携带的时间信息(例如，发送时间戳和接收时间戳)，通过上述公式(2)和(4)确定第一测试路径L1的抖动。
步骤S203B1、根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定第二测试路径的时延或抖动。
可选地，每个第二测试报文都会携带发送时间戳和接收时间戳，其中，发送时间戳用于指示网络性能检测装置发送第二测试报文的发送时间，接 收时间戳用于指示网络性能检测装置接收第二测试报文的接收时间。
本步骤中，可选地，网络性能检测装置根据接收的第二组测试报文中每个第二测试报文所携带的时间信息(例如，发送时间戳和接收时间戳)，通过上述公式(3)确定第二测试路径L2的时延。
可选地，网络性能检测装置根据接收的第二组测试报文中每个第二测试报文所携带的时间信息(例如，发送时间戳和接收时间戳)，通过上述公式(2)、(4)和(5)确定第二测试路径L2的抖动。
步骤S203C1、根据第一测试路径的时延、第二测试路径的时延以及预设的振荡次数，确定待测链路的时延；或者，根据第一测试路径的抖动、第二测试路径的抖动以及预设的振荡次数，确定待测链路的抖动。
本实施例中，由于第一测试报文根据预设的振荡次数会在待测链路L3上振荡，因此，在计算待测链路L3的网络性能时需要考虑振荡因素。可选地，假设第一测试报文在待测链路L3上振荡n次(即预设的振荡次数为n)，则根据第一测试路径L1的时延delayL1、第二测试路径L2的时延delayL2以及预设的振荡次数n，通过如下公式(6)和(7)确定待测链路L3的时延delayL3：
delayL1＝2*delayL2+2n*delayL3     (6)
Figure PCTCN2016112734-appb-000005
可选地，假设第一测试报文在待测链路L3上振荡n次(即预设的振荡次数为n)，则根据第一测试路径L1的抖动jitterL1、第二测试路径L2的抖动jitterL2以及预设的振荡次数n，通过如下公式(8)和(9)确定待测链路L3的抖动jitterL3：
jitterL1＝2*jitterL2+2n*jitterL3      (8)
Figure PCTCN2016112734-appb-000006
综上所述，由于第一测试报文在待测链路L3上振荡n次，因此，待测链路L3的时延或抖动的计算过程中，需要进行多次平均，从而减小了上游链路的时延或抖动(如第二测试路径L2的时延或抖动)的波动对待测链路L3的网络性能测量的影响，使得待测链路L3的时延或抖动的测量 结果更加准确。
图5A为本申请网络性能检测方法实施例四的流程示意图。如图5A所示，在上述实施例的基础上，步骤S203包括：
步骤S203A2、根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率。
本步骤中，网络性能检测装置根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，通过上述公式(1)确定第一测试路径L1的丢包率。
步骤S203B2、根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率。
本步骤中，网络性能检测装置根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，通过上述公式(1)确定第二测试路径L2的丢包率。
步骤S203C2、根据第一测试路径的丢包率、第二测试路径的丢包率以及预设的振荡次数，确定待测链路的丢包率。
本实施例中，由于第一测试报文根据预设的振荡次数会在待测链路L3上振荡，因此，在计算待测链路L3的网络性能时需要考虑振荡因素。可选地，假设第一测试报文在待测链路L3上振荡n次(即预设的振荡次数为n)，则根据第一测试路径L1的丢包率plrL1、第二测试路径L2的丢包率plrL2以及预设的振荡次数n，通过如下公式(10)和(11)确定待测链路L3的丢包率plrL3：
plrL1＝1-(1-plrL2)2(1-plrL3)2n       (10)
plrL3＝1-((1-plrL1)/(1-plrL2)2)1/2n       (11)
综上所述，由于第一测试报文在待测链路L3上振荡n次，因此，待测链路L3的丢包率的计算过程中，需要开n次幂，从而减小了上游链路的丢包率(如第二测试路径L2的丢包率)的波动对待测链路L3的网络性能测量的影响，使得待测链路L3的丢包率的测量结果更加准确。
图5B为本申请网络性能检测方法实施例五的流程示意图。如图5B 所示，在上述实施例的基础上，步骤S203包括：
步骤S203A3、根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定第一组测试报文的乱序数，并根据乱序数与第一测试路径上发送的第一测试报文的个数的商，确定第一测试路径的乱序率。
本步骤中，网络性能检测装置根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文ID以及第一预设报文标识(如待处理测试报文的报文标识)，判断当前接收到的第一测试报文的报文ID是否等于第一预设报文标识，若是(即当前接收到的第一测试报文为提前到达的测试报文)，则将当前接收到的第一测试报文存入缓存序列(缓存序列用于保存乱序的第一测试报文)。进一步地，网络性能检测装置根据缓存序列中所包括的第一测试报文的数量确定第一组测试报文的乱序数，并根据乱序数与第一测试路径上发送的所有第一测试报文的个数的商，确定第一测试路径L1的乱序率。
步骤S203B3、根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定第二组测试报文的乱序数，并根据乱序数与第二测试路径上发送的第二测试报文的个数的商，确定第二测试路径的乱序率。
本步骤中，网络性能检测装置根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识(如待处理测试报文的报文标识)，判断当前接收到的第二测试报文的报文ID是否等于第二预设报文标识，若是(即当前接收到的第二测试报文为提前到达的测试报文)，则将当前接收到的第二测试报文存入缓存序列(缓存序列用于保存乱序的第二测试报文)。进一步地，网络性能检测装置根据缓存序列中所包括的第二测试报文的数量确定第二组测试报文的乱序数，并根据乱序数与第二测试路径上发送的所有第二测试报文的个数的商，确定第二测试路径L2的乱序率。
步骤S203C3、根据第一测试路径的乱序率、第二测试路径的乱序率以及预设的振荡次数，确定待测链路的乱序率。
本实施例中，由于第一测试报文根据预设的振荡次数会在待测链路 L3上振荡，因此，在计算待测链路L3的网络性能时需要考虑振荡因素。可选地，假设第一测试报文在待测链路L3上振荡n次(即预设的振荡次数为n)，则根据第一测试路径L1的乱序率disorL1、第二测试路径L2的乱序率disorL2以及预设的振荡次数n，通过如下公式(12)和(13)确定待测链路L3的乱序率disorL3：
disorL1＝1-(1-disorL2)2(1-disorL3)2n      (12)
disorL3＝1-((1-disorL1)/(1-disorL2)2)1/2n       (13)
综上所述，由于第一测试报文在待测链路L3上振荡n次，因此，待测链路L3的乱序率的计算过程中，需要开n次幂，从而减小了上游链路的乱序率(如第二测试路径L2的乱序率)的波动对待测链路L3的网络性能测量的影响，使得待测链路L3的乱序率的测量结果更加准确。
根据上述公式(2)-(5)可知，测试路径的时延以及抖动的计算，完全依赖于测试报文的发送时间和接收时间。由于测试报文是网络性能检测装置逐个按顺序发送的，可以认为测试报文的发送时间都是准确的，因此，为每个测试报文标记合适的接收时间戳，会直接影响测试路径的时延与抖动的计算。
由于网络中存在乱序情况，应用程序在处理数据时，通常将接收的所有数据存入缓冲区，并对缓冲区内的数据做排序处理，以便进一步按顺序依次处理数据。例如，假设第3秒的视频数据比第1秒的视频数据优先达到，应用程序不会直接处理第3秒的视频数据，而是在等到第1秒的视频数据到达后，按顺序先处理第1秒的视频数据，因此，用户感觉到的时延要比视频数据的真实时延(即视频数据从发送端发出的时间与到达接收端的时间之间的时延)要长。
与上述视频数据类似，测试报文到达接收节点的时间不一定是真实处理时间。为了使时延的计算更接近于用户的真实感受，需要考虑乱序对时延和/或抖动的影响，下面介绍在考虑乱序情况下，如何对测试报文打上合适的接收时间戳，以便获得更准确的时延和/或抖动的测量结果。
图6A为本申请网络性能检测方法实施例六的流程示意图。如图6A所示，在上述实施例的基础上，步骤S203A1，包括：
步骤SA11、根据第一组测试报文的乱序为每个第一测试报文分配接 收时间戳。
本步骤中，网络性能检测装置根据第一组测试报文的乱序为每个第一测试报文分配合适的接收时间戳，使得时延和/或抖动的计算更接近于用户的真实感受。
图6B为本申请实施例中网络性能检测装置为每个第一测试报文分配合适的接收时间戳的流程示意图。如图6B所示，当接收到第一组测试报文中的第i个第一测试报文时，网络性能检测装置判断i是否等于j；其中，i为当前接收到的第一测试报文的报文标识；j为待处理测试报文的报文标识(即网络性能检测装置已经为j-1个第一测试报文分配了合适的接收时间戳)。其中，i为大于0的整数，j为大于1的整数。
1)若i等于j(即当前所接收到的第i个第一测试报文为待处理报文)，则网络性能检测装置判断缓存序列中是否保存有第j+1个第一测试报文，其中，缓存序列用于保存乱序的第一测试报文(即提前到达的第一测试报文)，乱序的第一测试报文的报文标识大于j。若缓存序列中保存有第j+1个第一测试报文(即第j+1个第一测试报文比第j个第一测试报文提前到达网络性能检测装置)，则将第j+1个第一测试报文的初始接收时间戳修改为第i个(即第j个)第一测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤，以此类推，直至缓存序列中未保存有第j+1个第一测试报文，则继续接收第一组测试报文中下一个到达的第一测试报文。可选地，将第j+1个第一测试报文的初始接收时间戳修改为第i个第一测试报文的接收时间戳之后，将缓存序列中的第j+1个第一测试报文删除，以便缓存序列中可以保存其它提前到达的第一测试报文。
2)若i不等于j(即当前所接收到的第i个第一测试报文为非待处理报文)，则网络性能检测装置判断缓存序列中是否已满。2a)若缓存序列中未满，则将第i个第一测试报文保存至缓存序列中，并继续接收第一组测试报文中下一个到达的第一测试报文。2b)若缓存序列中已满，则判断缓存序列中是否保存有第j个第一测试报文(即待处理测试报文)。若缓存序列中未保存有第j个第一测试报文，则确定第j个第一测试报文已丢失，根据第一组测试报文中报文标识小于j的各第一测试报文的平均时延， 确定第j个第一测试报文的时延，并执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤。可选地，网络性能检测装置将第一组测试报文中报文标识小于j的各第一测试报文的平均时延与预设系数(例如，1.5)的乘积，作为该第j个第一测试报文的时延。当然，根据第一组测试报文中报文标识小于j的各第一测试报文的平均时延，还可通过其它方式确定该第j个第一测试报文的时延，本申请实施例中对此并不作限制。
可见，网络性能检测装置将接收到的第一组测试报文的每个第一测试报文都打上合适的接收时间戳，其中，对于丢失的第一测试报文赋予了合理的时延。
步骤SA12、根据每个第一测试报文的接收时间戳、发送时间戳，确定第一测试路径的时延或抖动。
本步骤中，网络性能检测装置根据每个第一测试报文的接收时间戳(即根据第一组测试报文的乱序为每个第一测试报文所分配的合适的接收时间戳)、发送时间戳，确定第一测试路径的时延或抖动。可选地，具体的计算方式参见本申请上述网络性能检测方法实施例三中关于步骤S203A1的相关描述，本申请实施例中，此处不再赘述。
本实施例中，通过根据第一组测试报文的乱序为每个第一测试报文分配接收时间戳；进一步地，根据每个第一测试报文的接收时间戳(即根据第一组测试报文的乱序为每个第一测试报文所分配的合适的接收时间戳)、发送时间戳，计算所得到的第一测试路径的时延或抖动的测量结果更加精确，从而进一步提高了待测链路的时延或抖动测量结果的准确度。
可选地，本申请上述网络性能检测方法实施例三中的步骤S203B1(即根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定第二测试路径的时延或抖动)，包括：根据第二组测试报文的乱序为每个第二测试报文分配接收时间戳；根据每个第二测试报文的接收时间戳、发送时间戳，确定第二测试路径的时延或抖动。根据上述方式计算所得到的第二测试路径的时延或抖动的测量结果更加精确，从而进一步提高了待测链路的时延或抖动测量结果的准确度。可选地，具体的实现方式可参见本申请上述网络性能检测方法实施例六中关于步骤SA11和步骤SA12的相关描述，本申请实施例中，此处不再赘述。
图7为本申请实施例中网络性能检测装置为每个测试报文分配合适的接收时间戳的流程示意图。假设网络性能检测装置从第1个测试报文开始，逐个为测试报文分配合适的接收时间戳，j表示待处理测试报文的报文标识(即网络性能检测装置已经将报文标识小于等于j-1的测试报文都分配了合适的接收时间戳)，i表示网络性能检测装置当前接收到的测试报文的报文标识，m1表示网络性能检测装置接收到测试报文的总个数，m2表示已经进入缓冲序列的测试报文的总个数。
如图7所示，在上述实施例的基础上，本实施例中网络性能检测装置为每个测试报文分配合适的接收时间戳的流程如下：
(1)当接收到第i个测试报文时，网络性能检测装置将m1加1，并为第i个测试报文打上初始接收时间戳(用于指示第i个测试报文的接收时间)。
(2)网络性能检测装置判断i是否等于j；若i不等于j(即当前所接收到的第i个测试报文为非待处理报文)，则将第i个测试报文存入缓冲序列，并判断缓存序列中是否已满。若缓存序列中未满，则将m2+1，并继续接收下一个到达的测试报文。若缓存序列中已满，则判断缓存序列中是否保存有第j个测试报文；若缓存序列中未保存有第j个测试报文，则确定第j个测试报文已丢失，此时由于无法确定第j个测试报文的接收时间，为了便于计算测试路径上的时延，根据已到达的报文标识小于j的各测试报文的平均时延与预设系数(例如，1.5)的乘积，确定第j个测试报文的默认时延；进一步地，判断缓存序列中是否保存有第j+1个测试报文。
(3)若i等于j(即当前所接收到的第i个测试报文为待处理报文)，则网络性能检测装置判断缓存序列中是否保存有第j+1个测试报文(即判断第j+1个测试报文是否比第j个测试报文提前到达)。若缓存序列中保存有第j+1个测试报文(即第j+1个测试报文比第j个测试报文提前到达网络性能检测装置)，则将第j个测试报文的接收时间戳覆盖第j+1个测试报文的初始接收时间戳，并删除缓存序列中的第j+1个测试报文以及j加1，返回执行判断缓存序列中是否保存有第j+1个第一测试报文的步骤，以此类推；
(4)接上述步骤“判断缓存序列中是否保存有第j+1个第一测试报文，以此类推”，直至缓存序列中未保存有第j+1个测试报文，则继续接收下一个到达的测试报文。可选地，在继续接收下一个测试报文之前，若对缓存序列 中的测试报文进行删除操作，则将m2加1；否则，无需加1。
(5)网络性能检测装置实时计算已经接收到的测试报文的平均时延，当连续5倍的平均时延时间内都未接收到新测试报文时，则确定测试报文接收完毕。
综上所述，本实施例中，网络性能检测装置通过根据乱序为每个测试报文分配合适的接收时间戳，使得时延的计算更接近于用户的真实感受，以便获得更准确的时延和/或抖动的测量结果。
图8为本申请网络性能检测方法一实施例中接收到的测试报文序列以及缓存序列的示意图。结合图2B和图8所示，对网络性能检测装置为每个测试报文分配合适的接收时间戳的流程进行详细说明。
1)初始化m1＝0，m2＝0，j＝1；
2)当接收到第1个测试报文时(即i＝1)，网络性能检测装置将m1加1(即m1＝1)，并为第1个测试报文打上初始接收时间戳(用于指示第1个测试报文的接收时间)；由于i等于j并且缓冲序列为空，则直接处理下一个测试报文(即第2个测试报文)，对第1个测试报文的处理同第1个测试报文，此处不再赘述；在处理完第2个测试报文后，j＝3，m1＝2，m2＝0；
3)当接收到第6个测试报文时(即i＝6)时，网络性能检测装置将m1加1(即m1＝3)，并为第6个测试报文打上初始接收时间戳；由于i不等于j并且缓冲序列为空，则将第6个测试报文存入缓冲序列，以及将m2加1(即m2＝1)，并直接处理下一个测试报文，直到第8个测试报文存入缓冲序列(假设缓存序列能够存放5个报文)；
4)此时缓冲序列已满，网络性能检测装置确定第3个测试报文已丢失，并为第3个测试报文分配默认时延(可选地，默认时延等于第1个测试报文与第2个测试报文的平均时延的1.5倍)；进一步地，在确定缓存序列中保存有第4个测试报文，则将第8个测试报文的接收时间戳覆盖第4个测试报文的初始接收时间戳，并删除缓存序列中的第4个测试报文，以及将j加1；依次对缓存序列中的第5个测试报文、第6个测试报文、第7个测试报文和第8个测试报文进行相应处理，直至缓存序列中未保存有提前到达的测试报文，则继续接收下一个到达的测试报文(即第9个测试报文)；
5)当接收到第9个测试报文时(即i＝9)时，j＝9，m1＝7以及m2＝5，网络 性能检测装置将m1加1(即m1＝8)，并为第9个测试报文打上初始接收时间戳；由于i等于j并且缓冲序列为空，则直接处理下一个测试报文(即第11个测试报文)；
6)当接收到第11个测试报文时(即i＝11)时，j＝10，m1＝8以及m2＝5，网络性能检测装置将m1加1(即m1＝9)，并为第11个测试报文打上初始接收时间戳；由于i不等于j，则将第11个测试报文存入缓冲序列，由于缓冲序列未满则将m2加1(即m2＝6)，并直接处理下一个测试报文(即第10个测试报文)；
7)当接收到第10个测试报文时(即i＝10)时，网络性能检测装置将m1加1(即m1＝10)，并为第10个测试报文打上初始接收时间戳(此时j＝9)；由于i等于j，在确定缓存序列中保存有第11个测试报文，则将第10个测试报文的接收时间戳覆盖第11个测试报文的初始接收时间戳，并删除缓存序列中的第11个测试报文(由于执行了删除操作，将m2加1)；此时j＝12，m1＝10以及m2＝7。
可见，网络性能检测装置将接收到的每个测试报文都打上合适的接收时间戳，并对于丢失的测试报文赋予了合理的时延，使得测试路径的时延/抖动的测量结果更加精确。例如，虽然第3个测试报文丢失了，网络性能检测装置通过为第3个测试报文分配默认时延，并不影响第2个测试报文与第3个测试报文，以及第3个测试报文与第4个测试报文之间抖动的计算。
根据上述公式公式(2)-(5)可知，测试路径的时延以及抖动的计算，完全依赖于测试报文的发送时间和接收时间。由于测试报文是网络性能检测装置逐个按顺序发送的，可以认为测试报文的发送时间都是准确的，因此，为每个测试报文标记合适的接收时间戳，会直接影响测试路径的时延与抖动的计算。
由于网络中存在乱序情况，应用程序在处理数据时，通常将接收的所有数据存入缓冲区，并对缓冲区内的数据做排序处理，以便进一步按顺序依次处理数据。例如，假设第3秒的视频数据比第1秒的视频数据优先达到，应用程序不会直接处理第3秒的视频数据，而是在等到第1秒的视频数据到达后，按顺序先处理第1秒的视频数据，因此，用户感觉到的时延要比视频数据的真实时延(即视频数据从发送端发出的时间与到达接收端的时间之间的时延)要长。
与上述视频数据类似，测试报文到达接收节点的时间不一定是真实处理时间。为了使时延的计算更接近于用户的真实感受，需要考虑乱序情况对时延和/或抖动的影响，下面介绍在考虑乱序情况下，如何对测试报文打上合适的接收时间戳，以便获得更准确的时延和/或抖动的测量结果。
图9A为本申请网络性能检测方法实施例七的流程示意图。本实施例的执行主体可以为配置在服务器中的网络性能检测装置，该装置可以通过软件和/或硬件实现。如图9A所示，本实施例的方法可以包括：
步骤S901、接收测试路径上返回的一组测试报文。
本步骤中，网络性能检测装置接收某条测试路径上返回的一组测试报文。可选地，该组测试报文中的每个测试报文携带：探测器标识；所述探测器标识用于指示发出对应测试报文的网络性能检测装置；具体的报文格式可参见图3B所示。
步骤S902、根据该组测试报文的乱序为该组测试报文中每个测试报文分配接收时间戳。
本步骤中，网络性能检测装置根据该组测试报文的乱序情况为该组测试报文中每个测试报文分配接收时间戳，使得时延和/或抖动的计算更接近于用户的真实感受。
图9B为本申请实施例中网络性能检测装置为每个测试报文分配合适的接收时间戳的流程示意图三。如图9B所示，当接收到一组测试报文中的第i个测试报文时，网络性能检测装置判断i是否等于j+1；其中，i为当前接收到的测试报文的报文标识；j为待处理测试报文的报文标识(即网络性能检测装置已经为j-1个第一测试报文分配了合适的接收时间戳)。
1)若i等于j(即当前所接收到的第i个测试报文为待处理报文)，则网络性能检测装置判断缓存序列中是否保存有第j+1个测试报文，其中，缓存序列用于保存乱序的测试报文(即提前到达的测试报文)，乱序的测试报文的报文标识大于j。若缓存序列中保存有第j+1个测试报文(即第j+1个测试报文比第j个测试报文提前到达网络性能检测装置)，则将第j+1个测试报文的初始接收时间戳修改为第i个(即第j个)测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个测试报文的步骤，直至缓存序列中未保存有第j+1个测试报文，则继续接收 该组测试报文中下一个到达的测试报文。可选地，将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳之后，将缓存序列中的第j+1个测试报文删除，以便缓存序列中可以保存其它提前到达的测试报文。
2)若i不等于j(即当前所接收到的第i个测试报文为非待处理报文)，则网络性能检测装置判断缓存序列中是否已满。2a)若缓存序列中未满，则将第i个测试报文保存至缓存序列中，并继续接收该组测试报文中下一个到达的测试报文。2b)若缓存序列中已满，则判断缓存序列中是否保存有第j个测试(即待处理测试报文)。若缓存序列中未保存有第j个测试报文，则确定第j个测试报文已丢失，根据该组测试报文中报文标识小于j的各测试报文的平均时延，确定第j个测试报文的时延，并执行判断缓存序列中是否保存有第j+1个测试报文的步骤。可选地，网络性能检测装置将该组测试报文中报文标识小于j的各测试报文的平均时延与预设系数(例如，1.5)的乘积，作为该第j个测试报文的时延。当然，根据该组测试报文中报文标识小于j的各测试报文的平均时延，还可通过其它方式确定第j个测试报文的时延，本申请实施例中对此并不作限制。
可见，网络性能检测装置将接收到的一组测试报文的每个测试报文都打上合适的接收时间戳，其中，对于丢失的测试报文赋予了合理的时延。
步骤S903、根据每个测试报文的接收时间戳、发送时间戳，确定测试路径的时延或抖动。
本步骤中，网络性能检测装置根据每个测试报文的接收时间戳(即根据该组测试报文的乱序为每个该测试报文所分配的合适的接收时间戳)、发送时间戳，确定测试路径的网络性能时延或抖动。可选地，具体的计算方式参见本申请上述网络性能检测方法实施例三中关于步骤S203A1的相关描述，本申请实施例中，此处不再赘述。
本实施例中，在接收测试路径上返回的一组测试报文后，通过根据该组测试报文的乱序为每个测试报文分配接收时间戳；进一步地，根据每个测试报文的接收时间戳(即根据该组测试报文的乱序为每个测试报文所分配的合适的接收时间戳)、发送时间戳，从而使得计算所得到的测试路径的网络性能的测量结果更加精确。
可选地，本申请实施例中，网络性能检测装置为每个测试报文分配合适的接收时间戳的流程，还可参见上述图7所示网络性能检测装置为每个测试报文分配合适的接收时间戳的流程示意图二，以及结合图8所述的网络性能检测装置为每个测试报文分配合适的接收时间戳的流程，此处不再赘述。
图10为本申请网络性能检测装置实施例一的结构示意图。如图10所示，本实施例提供的终端网络性能检测装置100，包括：
发送模块1001，用于发送第一组测试报文以及第二组测试报文；其中，所述第一组测试报文用于测试第一测试路径，所述第二组测试报文用于测试第二测试路径；所述第一测试路径包括：所述第二测试路径以及待测链路；
接收模块1002，用于接收所述第一测试路径上返回的第一组测试报文，以及所述第二测试路径上返回的第二组测试报文；其中，所述第一组测试报文根据预设的振荡次数在所述待测链路上振荡；
确定模块1003，用于根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能。
可选地，所述装置还包括：
生成模块，用于根据所述预设的振荡次数生成所述第一组测试报文；
其中，所述第一组测试报文中的每个第一测试报文携带：所述待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个所述第一链路标签组包括M个相同的第一链路标签，每个所述第二链路标签组包括M个相同的第二链路标签；所述M的数值等于所述预设的振荡次数，所述第一链路标签用于指示所述第一测试报文在正向传输方向的下一跳目标节点，所述第二链路标签用于指示所述第一测试报文在反向传输方向的下一跳目标节点。
可选地，所述确定模块1003，包括：
第一确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定所述第一测试路径的时延或抖动；
第二确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定所述第二测试路径的时延或抖动；
第三确定子模块，用于根据所述第一测试路径的时延、所述第二测试路径的时延以及所述预设的振荡次数，确定所述待测链路的时延；或者，根据所述第一测试路径的抖动、所述第二测试路径的抖动以及所述预设的振荡次数，确定所述待测链路的抖动。
可选地，所述确定模块1003，包括：
第四确定子模块，用于根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
第五确定子模块，用于根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
第六确定子模块，用于根据所述第一测试路径的丢包率、所述第二测试路径的丢包率以及所述预设的振荡次数，确定所述待测链路的丢包率。
可选地，所述确定模块1003，包括：
第七确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定所述第一组测试报文的乱序数，并根据所述乱序数与所述第一测试路径上发送的所述第一测试报文的个数的商，确定所述第一测试路径的乱序率；
第八确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定所述第二组测试报文的乱序数，并根据所述乱序数与所述第二测试路径上发送的所述第二测试报文的个数的商，确定所述第二测试路径的乱序率；
第九确定子模块，用于根据所述第一测试路径的乱序率、所述第二测试路径的乱序率以及所述预设的振荡次数，确定所述待测链路的乱序率。
可选地，所述第一确定子模块，包括：
分配单元，用于根据所述第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
确定单元，用于根据每个所述第一测试报文的接收时间戳、发送时间戳，确定所述第一测试路径的时延或抖动。
可选地，所述分配单元具体用于：
当所述接收模块接收到所述第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，所述i为当前接收到的所述第一测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，所述缓存序列用于保存乱序的第一测试报文，所述乱序的第一测试报文的报文标识大于j；
若所述缓存序列中保存有第j+1个第一测试报文，则将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳，并将j加1，返回执行所述判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤，直至所述缓存序列中未保存有第j+1个第一测试报文，则所述接收模块继续接收所述第一组测试报文中下一个到达的第一测试报文。
可选地，所述分配单元还用于：
若i不等于j，则判断所述缓存序列中是否已满；
若所述缓存序列中未满，则将所述第i个第一测试报文保存至所述缓存序列中，所述接收模块继续接收所述第一组测试报文中下一个到达的第一测试报文；
若所述缓存序列中已满，则判断所述缓存序列中是否保存有第j个第一测试报文；
若所述缓存序列中未保存有所述第j个第一测试报文，则确定所述第j个第一测试报文已丢失，根据所述第一组测试报文中报文标识小于所述j的各第一测试报文的平均时延，确定所述第j个第一测试报文的时延，并执行判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤。
可选地，所述分配单元将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳之后，还用于：将所述缓存序列中的所述第j+1个第一测试报文删除。
本实施例的网络性能检测装置，可以用于执行本申请上述网络性能检测方法实施例一至实施例六中任意实施例的技术方案，其实现原理和技术效果类似，此处不再赘述。
图11为本申请网络性能检测装置实施例二的结构示意图。如图11所 示，本实施例提供的网络性能检测装置110可以包括处理器1101、存储器1102、接收器1103以及发送器1104。其中，存储器1102、接收器1103以及发送器1104都与处理器1101相连。
其中，发送器1104用于发送第一组测试报文以及第二组测试报文；其中，所述第一组测试报文用于测试第一测试路径，所述第二组测试报文用于测试第二测试路径；所述第一测试路径包括：所述第二测试路径以及待测链路。接收器1103用于接收所述第一测试路径上返回的第一组测试报文，以及所述第二测试路径上返回的第二组测试报文；其中，所述第一组测试报文根据预设的振荡次数在所述待测链路上振荡。存储器1102用于存储执行指令；处理器1101用于执行存储器1102中的执行指令，用以执行以下操作：根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能。
可选地，所述处理器1101还用于：
根据所述预设的振荡次数生成所述第一组测试报文；
其中，所述第一组测试报文中的每个第一测试报文携带：所述待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个所述第一链路标签组包括M个相同的第一链路标签，每个所述第二链路标签组包括M个相同的第二链路标签；所述M的数值等于所述预设的振荡次数，所述第一链路标签用于指示所述第一测试报文在正向传输方向的下一跳目标节点，所述第二链路标签用于指示所述第一测试报文在反向传输方向的下一跳目标节点。
可选地，所述处理器1101还用于：
根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定所述第一测试路径的时延或抖动；
根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定所述第二测试路径的时延或抖动；
根据所述第一测试路径的时延、所述第二测试路径的时延以及所述预设的振荡次数，确定所述待测链路的时延；或者，根据所述第一测试路径的抖动、所述第二测试路径的抖动以及所述预设的振荡次数，确定所述待测链路的抖动。
可选地，所述处理器1101还用于：
根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
根据所述第一测试路径的丢包率、所述第二测试路径的丢包率以及所述预设的振荡次数，确定所述待测链路的丢包率。
可选地，所述处理器1101还用于：
根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定所述第一组测试报文的乱序数，并根据所述乱序数与所述第一测试路径上发送的所述第一测试报文的个数的商，确定所述第一测试路径的乱序率；
根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定所述第二组测试报文的乱序数，并根据所述乱序数与所述第二测试路径上发送的所述第二测试报文的个数的商，确定所述第二测试路径的乱序率；
根据所述第一测试路径的乱序率、所述第二测试路径的乱序率以及所述预设的振荡次数，确定所述待测链路的乱序率。
可选地，所述处理器1101还用于：
根据所述第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
根据每个所述第一测试报文的接收时间戳、发送时间戳，确定所述第一测试路径的时延或抖动。
可选地，所述处理器1101还用于：
当接收到所述第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，所述i为当前接收到的所述第一测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，所述缓存序列用于保存乱序的第一测试报文，所述乱序的第一测试报文的报文标识大于j；
若所述缓存序列中保存有第j+1个第一测试报文，则将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳，并将j加1，返回执行所述判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤，直至所述缓存序列中未保存有第j+1个第一测试报文，则继续接收所述第一组测试报文中下一个到达的第一测试报文。
可选地，所述处理器1101还用于：
若i不等于j，则判断所述缓存序列中是否已满；
若所述缓存序列中未满，则将所述第i个第一测试报文保存至所述缓存序列中，并继续接收所述第一组测试报文中下一个到达的第一测试报文；
若所述缓存序列中已满，则判断所述缓存序列中是否保存有第j个第一测试报文；
若所述缓存序列中未保存有所述第j个第一测试报文，则确定所述第j个第一测试报文已丢失，根据所述第一组测试报文中报文标识小于所述j的各第一测试报文的平均时延，确定所述第j个第一测试报文的时延，并执行判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤。
可选地，所述处理器1101将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳之后，还用于：将所述缓存序列中的所述第j+1个第一测试报文删除。
本实施例的网络性能检测装置，可以用于执行本申请上述网络性能检测方法实施例一至实施例六中任意实施例的技术方案，其实现原理和技术效果类似，此处不再赘述。
图12为本申请网络性能检测装置实施例三的结构示意图。如图12所示，本实施例提供的终端网络性能检测装置120，包括：
接收模块1201，用于接收测试路径上返回的一组测试报文；
分配模块1202，用于根据组测试报文的乱序为组测试报文中每个测试报文分配接收时间戳；
确定模块1203，用于根据每个测试报文的接收时间戳、发送时间戳，确定测试路径的时延或抖动。
可选地，分配模块1202具体用于：
当接收模块接收到组测试报文中的第i个测试报文时，判断i是否等于j；其中，i为当前接收到的测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个测试报文；其中，缓存序列用于保存乱序的测试报文，乱序的测试报文的报文标识大于j；
若缓存序列中保存有第j+1个测试报文，则将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳，并将j加1，返回执行判断缓存序列中是否保存有第j+1个测试报文的步骤，直至缓存序列中未保存有第j+1个测试报文，则接收模块继续接收组测试报文中下一个到达的测试报文。
可选地，分配模块1202还用于：
若i不是否等于j，则判断缓存序列中是否已满；
若缓存序列中未满，则将第i个测试报文保存至缓存序列中，接收模块继续接收组测试报文中下一个到达的测试报文；
若缓存序列中已满，则判断缓存序列中是否保存有第j个测试报文；
若缓存序列中未保存有第j个测试报文，则确定第j个测试报文已丢失，根据组测试报文中报文标识小于j的各测试报文的平均时延，确定第j个测试报文的时延，并执行判断缓存序列中是否保存有第j+1个测试报文的步骤。
可选地，分配模块1202将第j+1个测试报文的初始接收时间戳修改为第i个测试报文的接收时间戳之后，还用于：
将缓存序列中的第j+1个测试报文删除。
本实施例的网络性能检测装置，可以用于执行本申请上述网络性能检测方法实施例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。
图13为本申请网络性能检测装置实施例四的结构示意图。如图13所示，本实施例提供的网络性能检测装置130可以包括处理器1301、存储器1302以及收发器1303。其中，存储器1302以及收发器1303都与处理器1301相连。
其中，收发器1303用于接收测试路径上返回的一组测试报文。存储 器1302用于存储执行指令；处理器1301用于执行存储器1302中的执行指令，用以执行以下操作：根据所述组测试报文的乱序为所述组测试报文中每个测试报文分配接收时间戳；根据每个所述测试报文的接收时间戳、发送时间戳，确定所述测试路径的时延或抖动。
可选地，处理器1301还用于：
当接收到所述组测试报文中的第i个测试报文时，判断i是否等于j；其中，所述i为当前接收到的所述测试报文的报文标识；j为待处理报文的报文标识；
若i等于j，则判断缓存序列中是否保存有第j+1个测试报文；其中，所述缓存序列用于保存乱序的测试报文，所述乱序的测试报文的报文标识大于j；
若所述缓存序列中保存有第j+1个测试报文，则将所述第j+1个测试报文的初始接收时间戳修改为所述第i个测试报文的接收时间戳，并将j加1，返回执行所述判断所述缓存序列中是否保存有第j+1个测试报文的步骤，直至所述缓存序列中未保存有第j+1个测试报文，则继续接收所述组测试报文中下一个到达的测试报文。
可选地，处理器1301还用于：
若i不是否等于j，则判断所述缓存序列中是否已满；
若所述缓存序列中未满，则将所述第i个测试报文保存至所述缓存序列中，并继续接收所述组测试报文中下一个到达的测试报文；
若所述缓存序列中已满，则判断所述缓存序列中是否保存有第j个测试报文；
若所述缓存序列中未保存有所述第j个测试报文，则确定所述第j个测试报文已丢失，根据所述组测试报文中报文标识小于所述j的各测试报文的平均时延，确定所述第j个测试报文的时延，并执行判断所述缓存序列中是否保存有第j+1个测试报文的步骤。
可选地，处理器1301将所述第j+1个测试报文的初始接收时间戳修改为所述第i个测试报文的接收时间戳之后，还用于：将所述缓存序列中的所述第j+1个测试报文删除。
本实施例的网络性能检测装置，可以用于执行本申请上述网络性能检测 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。
在本申请所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
上述以软件功能单元的形式实现的集成的单元，可以存储在一个计算机可读取存储介质中。上述软件功能单元存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或处理器(processor)执行本申请各个实施例所述方法的部分步骤。
本领域技术人员可以清楚地了解到，为描述的方便和简洁，仅以上述各功能模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将装置的内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。上述描述的装置的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
本领域普通技术人员可以理解：本文中涉及的第一、第二等各种数字编号仅为描述方便进行的区分，并不用来限制本申请实施例的范围。
本领域普通技术人员可以理解，在本申请的各种实施例中，上述各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功 能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。
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步骤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步骤；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。

Claims (18)

  1. 一种网络性能检测方法，其特征在于，包括：
    发送第一组测试报文以及第二组测试报文；其中，所述第一组测试报文用于测试第一测试路径，所述第二组测试报文用于测试第二测试路径；所述第一测试路径包括：所述第二测试路径以及待测链路；
    接收所述第一测试路径上返回的第一组测试报文，以及所述第二测试路径上返回的第二组测试报文；其中，所述第一组测试报文根据预设的振荡次数在所述待测链路上振荡；
    根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能。
  2. 根据权利要求1所述的方法，其特征在于，所述发送第一组测试报文以及第二组测试报文之前，还包括：
    根据所述预设的振荡次数生成所述第一组测试报文；
    其中，所述第一组测试报文中的每个第一测试报文携带：所述待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个所述第一链路标签组包括M个相同的第一链路标签，每个所述第二链路标签组包括M个相同的第二链路标签；所述M的数值等于所述预设的振荡次数，所述第一链路标签用于指示所述第一测试报文在正向传输方向的下一跳目标节点，所述第二链路标签用于指示所述第一测试报文在反向传输方向的下一跳目标节点。
  3. 根据权利要求1或2所述的方法，其特征在于，所述根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能，包括：
    根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定所述第一测试路径的时延或抖动；
    根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定所述第二测试路径的时延或抖动；
    根据所述第一测试路径的时延、所述第二测试路径的时延以及所述预设的振荡次数，确定所述待测链路的时延；或者，根据所述第一测试路径的抖动、所述第二测试路径的抖动以及所述预设的振荡次数，确定所述待 测链路的抖动。
  4. 根据权利要求1或2所述的方法，其特征在于，所述根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能，包括：
    根据发送的第一组测试报文中第一测试报文的个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
    根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
    根据所述第一测试路径的丢包率、所述第二测试路径的丢包率以及所述预设的振荡次数，确定所述待测链路的丢包率。
  5. 根据权利要求1或2所述的方法，其特征在于，所述根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能，包括：
    根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定所述第一组测试报文的乱序数，并根据所述乱序数与所述第一测试路径上发送的所述第一测试报文的个数的商，确定所述第一测试路径的乱序率；
    根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定所述第二组测试报文的乱序数，并根据所述乱序数与所述第二测试路径上发送的所述第二测试报文的个数的商，确定所述第二测试路径的乱序率；
    根据所述第一测试路径的乱序率、所述第二测试路径的乱序率以及所述预设的振荡次数，确定所述待测链路的乱序率。
  6. 根据权利要求3所述的方法，其特征在于，所述根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定所述第一测试路径的时延或抖动，包括：
    根据所述第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
    根据每个所述第一测试报文的接收时间戳、发送时间戳，确定所述第一测试路径的时延或抖动。
  7. 根据权利要求6所述的方法，其特征在于，所述根据所述第一组测试报文的乱序为每个第一测试报文分配接收时间戳，包括：
    当接收到所述第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，所述i为当前接收到的所述第一测试报文的报文标识；j为待处理测试报文的报文标识；
    若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，所述缓存序列用于保存乱序的第一测试报文，所述乱序的第一测试报文的报文标识大于j；
    若所述缓存序列中保存有第j+1个第一测试报文，则将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳，并将j加1，返回执行所述判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤，直至所述缓存序列中未保存有第j+1个第一测试报文，则继续接收所述第一组测试报文中下一个到达的第一测试报文。
  8. 根据权利7所述的方法，其特征在于，所述判断i是否等于j之后，还包括：
    若i不等于j，则判断所述缓存序列中是否已满；
    若所述缓存序列中未满，则将所述第i个第一测试报文保存至所述缓存序列中，并继续接收所述第一组测试报文中下一个到达的第一测试报文；
    若所述缓存序列中已满，则判断所述缓存序列中是否保存有第j个第一测试报文；
    若所述缓存序列中未保存有所述第j个第一测试报文，则确定所述第j个第一测试报文已丢失，根据所述第一组测试报文中报文标识小于所述j的各第一测试报文的平均时延，确定所述第j个第一测试报文的时延，并执行判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤。
  9. 根据权利要求7或8所述的方法，其特征在于，所述将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳之后，还包括：
    将所述缓存序列中的所述第j+1个第一测试报文删除。
  10. 一种网络性能检测装置，其特征在于，包括：
    发送模块，用于发送第一组测试报文以及第二组测试报文；其中，所述第一组测试报文用于测试第一测试路径，所述第二组测试报文用于测试第二测试路径；所述第一测试路径包括：所述第二测试路径以及待测链路；
    接收模块，用于接收所述第一测试路径上返回的第一组测试报文，以及所述第二测试路径上返回的第二组测试报文；其中，所述第一组测试报文根据预设的振荡次数在所述待测链路上振荡；
    确定模块，用于根据接收的所述第一组测试报文以及所述第二组测试报文，确定所述待测链路的网络性能。
  11. 根据权利要求10所述的装置，其特征在于，所述装置还包括：
    生成模块，用于根据所述预设的振荡次数生成所述第一组测试报文；
    其中，所述第一组测试报文中的每个第一测试报文携带：所述待测链路对应的至少一个第一链路标签组和至少一个第二链路标签组；其中，每个所述第一链路标签组包括M个相同的第一链路标签，每个所述第二链路标签组包括M个相同的第二链路标签；所述M的数值等于所述预设的振荡次数，所述第一链路标签用于指示所述第一测试报文在正向传输方向的下一跳目标节点，所述第二链路标签用于指示所述第一测试报文在反向传输方向的下一跳目标节点。
  12. 根据权利要求10或11所述的装置，其特征在于，所述确定模块，包括：
    第一确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的时间信息确定所述第一测试路径的时延或抖动；
    第二确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的时间信息确定所述第二测试路径的时延或抖动；
    第三确定子模块，用于根据所述第一测试路径的时延、所述第二测试路径的时延以及所述预设的振荡次数，确定所述待测链路的时延；或者，根据所述第一测试路径的抖动、所述第二测试路径的抖动以及所述预设的振荡次数，确定所述待测链路的抖动。
  13. 根据权利要求10或11所述的装置，其特征在于，所述确定模块，包括：
    第四确定子模块，用于根据发送的第一组测试报文中第一测试报文的 个数以及接收的第一组测试报文中第一测试报文的个数，确定第一测试路径的丢包率；
    第五确定子模块，用于根据发送的第二组测试报文中第二测试报文的个数以及接收的第二组测试报文中第二测试报文的个数，确定第二测试路径的丢包率；
    第六确定子模块，用于根据所述第一测试路径的丢包率、所述第二测试路径的丢包率以及所述预设的振荡次数，确定所述待测链路的丢包率。
  14. 根据权利要求10或11所述的装置，其特征在于，所述确定模块，包括：
    第七确定子模块，用于根据接收的第一组测试报文中每个第一测试报文携带的测试报文的报文身份标识ID以及第一预设报文标识，确定所述第一组测试报文的乱序数，并根据所述乱序数与所述第一测试路径上发送的所述第一测试报文的个数的商，确定所述第一测试路径的乱序率；
    第八确定子模块，用于根据接收的第二组测试报文中每个第二测试报文携带的测试报文的报文ID以及第二预设报文标识，确定所述第二组测试报文的乱序数，并根据所述乱序数与所述第二测试路径上发送的所述第二测试报文的个数的商，确定所述第二测试路径的乱序率；
    第九确定子模块，用于根据所述第一测试路径的乱序率、所述第二测试路径的乱序率以及所述预设的振荡次数，确定所述待测链路的乱序率。
  15. 根据权利要求12所述的装置，其特征在于，所述第一确定子模块，包括：
    分配单元，用于根据所述第一组测试报文的乱序为每个第一测试报文分配接收时间戳；
    确定单元，用于根据每个所述第一测试报文的接收时间戳、发送时间戳，确定所述第一测试路径的时延或抖动。
  16. 根据权利要求15所述的装置，其特征在于，所述分配单元具体用于：
    当所述接收模块接收到所述第一组测试报文中的第i个第一测试报文时，判断i是否等于j；其中，所述i为当前接收到的所述第一测试报文的报文标识；j为待处理报文的报文标识；
    若i等于j，则判断缓存序列中是否保存有第j+1个第一测试报文；其中，所述缓存序列用于保存乱序的第一测试报文，所述乱序的第一测试报文的报文标识大于j；
    若所述缓存序列中保存有第j+1个第一测试报文，则将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳，并将j加1，返回执行所述判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤，直至所述缓存序列中未保存有第j+1个第一测试报文，则所述接收模块继续接收所述第一组测试报文中下一个到达的第一测试报文。
  17. 根据权利16所述的装置，其特征在于，所述分配单元还用于：
    若i不等于j，则判断所述缓存序列中是否已满；
    若所述缓存序列中未满，则将所述第i个第一测试报文保存至所述缓存序列中，所述接收模块继续接收所述第一组测试报文中下一个到达的第一测试报文；
    若所述缓存序列中已满，则判断所述缓存序列中是否保存有第j个第一测试报文；
    若所述缓存序列中未保存有所述第j个第一测试报文，则确定所述第j个第一测试报文已丢失，根据所述第一组测试报文中报文标识小于所述j的各第一测试报文的平均时延，确定所述第j个第一测试报文的时延，并执行判断所述缓存序列中是否保存有第j+1个第一测试报文的步骤。
  18. 根据权利要求16或17所述的装置，其特征在于，所述分配单元将所述第j+1个第一测试报文的初始接收时间戳修改为所述第i个第一测试报文的接收时间戳之后，还用于：将所述缓存序列中的所述第j+1个第一测试报文删除。
PCT/CN2016/112734 2016-12-28 2016-12-28 网络性能检测方法及装置 WO2018119790A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
PCT/CN2016/112734 WO2018119790A1 (zh) 2016-12-28 2016-12-28 网络性能检测方法及装置
EP16925546.0A EP3474499B1 (en) 2016-12-28 2016-12-28 Network performance detection method and apparatus
CN201680069142.0A CN108513703B (zh) 2016-12-28 2016-12-28 网络性能检测方法及装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2016/112734 WO2018119790A1 (zh) 2016-12-28 2016-12-28 网络性能检测方法及装置

Publications (1)

Publication Number Publication Date
WO2018119790A1 true WO2018119790A1 (zh) 2018-07-05

Family

ID=62710846

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/112734 WO2018119790A1 (zh) 2016-12-28 2016-12-28 网络性能检测方法及装置

Country Status (3)

Country Link
EP (1) EP3474499B1 (zh)
CN (1) CN108513703B (zh)
WO (1) WO2018119790A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115086199A (zh) * 2021-03-15 2022-09-20 中国电信股份有限公司 网络质量测试方法及装置、存储介质、电子设备

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111262749B (zh) * 2018-11-30 2023-05-23 华为技术有限公司 一种检测网络可靠性的方法及设备
CN112787877B (zh) * 2019-11-07 2022-08-26 华为技术有限公司 一种网络时延检测方法及相关设备
CN111277454B (zh) * 2020-01-15 2021-06-25 Ut斯达康通讯有限公司 一种网络性能检测系统及方法
CN112565017A (zh) * 2020-12-04 2021-03-26 盛科网络（苏州）有限公司 一种基于SRv6 SID的时延测量方法及装置
CN115333980B (zh) * 2021-05-11 2023-10-20 中国联合网络通信集团有限公司 测试方法、第二边缘节点设备及边缘网关

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070081460A1 (en) * 2005-09-29 2007-04-12 Avaya Technology Llc Evaluating quality of service in an IP network with cooperating relays
CN101677289A (zh) * 2008-09-17 2010-03-24 华为技术有限公司 一种优化路由的方法和装置
CN102684947A (zh) * 2012-05-25 2012-09-19 四川迅游网络科技股份有限公司 一种网络加速系统的测速方法
CN104780095A (zh) * 2015-04-30 2015-07-15 杭州华三通信技术有限公司 一种sdn网络中的路径探测方法和装置

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1172481C (zh) * 2002-08-22 2004-10-20 陈鸣 互连网端到端性能监测方法及其系统
US7443801B2 (en) * 2004-10-28 2008-10-28 Telcordia Technologies, Inc. Remote estimation of round-trip delays in a data network
US7324913B2 (en) * 2006-02-01 2008-01-29 International Business Machines Corporation Methods and apparatus for testing a link between chips
US9571366B2 (en) * 2009-12-27 2017-02-14 At&T Intellectual Property I, L.P. Method and apparatus for detecting and localizing an anomaly for a network
US9369371B2 (en) * 2012-10-05 2016-06-14 Cisco Technologies, Inc. Method and system for path monitoring using segment routing
CN103139014B (zh) * 2013-01-28 2016-08-10 深信服网络科技（深圳）有限公司 基于旁路的网络质量评测方法及装置
CN106936656B (zh) * 2015-12-30 2020-01-03 华为技术有限公司 一种实现丢包检测的方法、装置和系统
CN108401490B (zh) * 2016-12-06 2020-02-21 华为技术有限公司 一种网络性能测量方法及探测设备

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070081460A1 (en) * 2005-09-29 2007-04-12 Avaya Technology Llc Evaluating quality of service in an IP network with cooperating relays
CN101677289A (zh) * 2008-09-17 2010-03-24 华为技术有限公司 一种优化路由的方法和装置
CN102684947A (zh) * 2012-05-25 2012-09-19 四川迅游网络科技股份有限公司 一种网络加速系统的测速方法
CN104780095A (zh) * 2015-04-30 2015-07-15 杭州华三通信技术有限公司 一种sdn网络中的路径探测方法和装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3474499A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115086199A (zh) * 2021-03-15 2022-09-20 中国电信股份有限公司 网络质量测试方法及装置、存储介质、电子设备

Also Published As

Publication number Publication date
CN108513703B (zh) 2020-07-28
EP3474499B1 (en) 2020-08-12
EP3474499A4 (en) 2019-08-07
EP3474499A1 (en) 2019-04-24
CN108513703A (zh) 2018-09-07

Similar Documents

Publication Publication Date Title
WO2018119790A1 (zh) 网络性能检测方法及装置
US10644977B2 (en) Scalable distributed end-to-end performance delay measurement for segment routing policies
JP4348124B2 (ja) ＱｏＳを推定する方法および通信装置
CN110505101B (zh) 一种网络切片编排方法及装置
US9379955B2 (en) Method for queuing data packets and node
JPWO2005067227A1 (ja) 負荷分散方法、ノード及び制御プログラム
CN106656679B (zh) 可用带宽测量方法
US20090310494A1 (en) Real-time network measurement
JPWO2005067227A6 (ja) 負荷分散方法、ノード及び制御プログラム
JP2013516807A (ja) パケットロス検出方法及び装置、並びにルータ
EP1424809A2 (en) Decentralized SLS monitoring in a differentiated service environment
CN106067854A (zh) 一种网络质量检测方法及设备
US20090310508A1 (en) Real-time network measurement
CN114553760B (zh) 路径权重分配方法以及装置
WO2020135429A1 (zh) 流量的分析、检测方法及装置
Wei et al. Shared bottleneck detection based on congestion interval variance measurement
US11088960B2 (en) Information processing apparatus and verification system
US20220247651A1 (en) System and method for network and computation performance probing for edge computing
CN110784378B (zh) 使用twamp实现精准流量均衡的方法及装置
JP2001069146A (ja) パケット交換ネットワーク区間空き帯域測定方法及び装置
CN107171887B (zh) 一种网络带宽测试方法、客户端及网络带宽测试系统
US10462032B2 (en) Probing a network
CN113812119A (zh) 用于性能测量的网络节点
Staehle et al. Measuring one-way delay in wireless mesh networks: an experimental investigation
Zhou Measuring available bandwidth for smart cyber-physical applications

Legal Events

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

Ref document number: 16925546

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2016925546

Country of ref document: EP

Effective date: 20190117

NENP Non-entry into the national phase

Ref country code: DE