WO2014101047A1 - 一种网络丢包类型识别方法、设备及系统 - Google Patents

一种网络丢包类型识别方法、设备及系统 Download PDF

Info

Publication number
WO2014101047A1
WO2014101047A1 PCT/CN2012/087673 CN2012087673W WO2014101047A1 WO 2014101047 A1 WO2014101047 A1 WO 2014101047A1 CN 2012087673 W CN2012087673 W CN 2012087673W WO 2014101047 A1 WO2014101047 A1 WO 2014101047A1
Authority
WO
WIPO (PCT)
Prior art keywords
packet
packet loss
network element
element device
statistical result
Prior art date
Application number
PCT/CN2012/087673
Other languages
English (en)
French (fr)
Inventor
蓝海清
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 CN201280002149.2A priority Critical patent/CN104322017B/zh
Priority to PCT/CN2012/087673 priority patent/WO2014101047A1/zh
Publication of WO2014101047A1 publication Critical patent/WO2014101047A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0823Errors, e.g. transmission errors
    • H04L43/0829Packet loss

Definitions

  • the present invention relates to the field of communications, and in particular, to a network packet loss type identification method, device and system.
  • the transmission of data packets between two devices through the intermediate network is the basic requirement of current information communication.
  • the basic function of the intermediate network is to make full use of its own bandwidth to transmit the data packets of the two devices as accurately and reliably as possible.
  • packets are lost when they are transmitted over the intermediate network, which affects the communication between the two devices.
  • packet loss is mainly caused by two reasons. One is that the intermediate bandwidth is insufficient, and the network is congested, causing packet loss. The other is that the intermediate network has transmitted a wrong packet, causing the network to actively drop packets, such as a bit error in microwave transmission. The product is wrongly packaged. Network packet loss will affect the quality of upper-layer services.
  • a certain method must be used to fully utilize network bandwidth to improve economic efficiency, reduce packet loss rate, and ensure service experience.
  • the user equipment at both ends can cancel the congestion by slowing down.
  • the packet loss caused by the second cause the generating end device can use the redundant transmission to transmit the error packet.
  • the client device must determine which packet is causing the packet loss. If the error is judged, it will cause further deterioration of the transmission. If the actual situation is error packet loss, but the wrong judgment is congestion packet loss, the device attempts to use the speed reduction to cancel the congestion, but in fact can not reduce the packet loss rate, causing the device to continue to slow down until the packet transmission speed is lost. bottom.
  • the two ends of the device use redundant transmission to resist packet loss, but the network load is increased, which causes more serious congestion, which may cause the intermediate network. collapse.
  • the embodiments of the present invention provide a network packet loss type identification method, device, and system, which can accurately and quickly determine the network packet loss type according to the network packet loss reason and the dynamic change of the packet loss reason.
  • embodiments of the present invention use the following technical solutions:
  • a method for identifying a network packet loss type including:
  • the first network element device sends a data packet to the second network element device; the first network element device classifies the data packet sent to the second network element device according to a first classification manner, where the first classification mode is And categorizing, according to a preset range of a packet length or a packet length of the data packet, the first network element device is configured to count the number of data packets of each packet length in the data packet sent to the second network element device, and generate First statistical result;
  • the first network element device sends the first statistical result to the second network element device, so that the second network element device determines the first network according to the first statistical result and the second statistical result.
  • a type of network packet loss between the metadevice and the second network element device or
  • the first network element device receives the second statistics result sent by the second network element device, and the first network element device determines, according to the first statistical result and the second statistical result, the first The type of network packet loss between the network element device and the second network element device;
  • the second statistical result is that, after the second network element device classifies the received data packet according to the first classification manner, the number of data packets of each type of packet length in the received data packet is statistical results.
  • the first network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second network
  • the type of network packet loss between the meta-devices includes: acquiring at least one of the first statistical result and the second statistical result The statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length, wherein the average packet length of the data packet of the first type of packet length is greater than the average packet length of the data packet of the second type of packet length ;
  • a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length where the packet loss rate is the first statistical result and the second statistic a ratio of a difference between the number of packets of the same packet length and the number of packets of the same packet length in the first statistical result; when the packet of the first packet length is lost When the rate of the packet loss rate of the second packet length is greater than the preset number of times, it is determined that the packet loss type of the current network is an error packet loss; when the packet of the first packet length is lost, When the rate of the packet loss rate of the second packet length is less than or equal to the preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the first network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second network
  • the type of network packet loss between the meta-devices includes: obtaining, in the first statistical result and the second statistical result, a statistical result of a data packet of a preset different packet length category according to the first classification manner, Calculating a packet loss ratio of each packet length category in the preset different packet length categories, where the packet loss rate is the same type of packet length in the first statistical result and the second statistical result. a ratio of a difference between the number of packets and a number of packets corresponding to the same packet length in the first statistical result;
  • the data packets of the preset different packet length categories are sorted in ascending order according to the packet length category, and the packet loss rate of the preset different packet length categories includes at least a preset number of data packet loss ratios. In ascending order, it is determined that the packet loss type of the current network is an error packet loss;
  • the packet loss type of the current network is a congestion packet loss.
  • the method further includes:
  • the method includes:
  • the delay of the second network element device currently receiving at least one data packet and the average delay of the second network element device receiving the data packet are obtained. Determining that the current network delay of receiving at least one data packet is greater than the average time delay of the received data packet, determining that the current network has both error packet loss and congestion packet loss;
  • the second aspect provides a network packet loss type identification method, where the second network element device receives a data packet sent by the first network element device, and classifies the received data packet according to a first classification manner, where the first classification is performed.
  • the method is classified according to a preset range of a packet length or a packet length of the data packet;
  • the second network element device counts the number of data packets of each type of packet length in the received data packet, and generates a second statistical result
  • the second network element device sends the second statistical result to the first network element device, so that the first network element device determines the first network according to the first statistical result and the second statistical result.
  • a type of network packet loss between the metadevice and the second network element device or,
  • the second network element device receives the first statistics result sent by the first network element device, and the second network element device determines, according to the first statistical result and the second statistical result, the first The type of network packet loss between the network element device and the second network element device;
  • the first statistical result is that the first network element device classifies the data packet sent to the second network element device according to the first classification manner, and sends the data packet to the second network element.
  • the second network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second network
  • the type of the network packet loss between the meta-devices includes: obtaining, in the first statistical result and the second statistical result, the statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length at least once The average packet length of the data packet of the first type of packet length is greater than the average packet length of the data packet of the second type of packet length;
  • a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length where the packet loss rate is the first statistical result and the second statistic a ratio of a difference between the number of packets of the same packet length and the number of packets of the same packet length in the first statistical result; when the packet of the first packet length is lost When the rate of the packet loss rate of the second packet length is greater than the preset number of times, it is determined that the packet loss type of the current network is an error packet loss; when the packet of the first packet length is lost, When the rate of the packet loss rate of the second packet length is less than or equal to the preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the second network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second network
  • the type of network packet loss between the meta-devices includes: in the first statistical result and in the second statistical result, according to the first
  • the categorization method is configured to obtain a statistical result of a data packet of a preset different packet length category, and calculate a packet loss rate of each packet length category in the preset different packet length categories, where the data packet loss rate is the a ratio of a difference between a number of packets of the same packet length and a number of packets corresponding to the packet length of the first packet in the first statistical result in the first statistical result;
  • the data packets of the preset different packet length categories are sorted in ascending order according to the packet length category, and the data packet loss rate of the preset different packet length categories includes at least a preset number of data packet loss ratios in ascending order.
  • the packet loss type of the current network is a congestion packet loss.
  • the method further includes:
  • the packet loss type of the current network is an error packet loss
  • linearly fitting the average packet length of the at least one packet length category packet with the packet loss rate of the at least one packet length category If the slope of the obtained curve is an upward trend or the slope of the curve is a fixed value greater than the preset slope, it is determined that there is an error packet loss in the current network; when all the packet length categories corresponding to the packet length category are lost in the fitted curve
  • the packet rate is greater than the preset packet loss rate threshold, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is in the coordinate space in the packet length category.
  • the packet loss rate is plotted on the ordinate.
  • the method further includes:
  • the packet loss type of the current network is an error packet loss
  • obtaining a delay of receiving at least one data packet and an average delay of receiving the data packet when the current delay of receiving at least one data packet is greater than the receiving
  • the average time delay of the data packet determines that the current network has both error packet loss and congestion packet loss;
  • the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet, it is determined that there is an error packet loss in the current network.
  • the third aspect provides a first network element device, including: a data sending unit, configured to send a data packet to the second network element device;
  • a data classification unit configured to classify data packets sent by the data sending unit to the second network element device according to a first classification manner, where the first classification manner is according to a packet length or a packet length of the data packet.
  • Preset range classification Preset range classification
  • a statistical unit configured to count the number of data packets of each type of packet length in the data packet after the data classification unit is classified, and generate a first statistical result
  • the data sending unit is further configured to send the first statistics result forwarded by the statistics unit to the second network element device, so that the second network element device is configured according to the first statistical result and the second The statistics result determines the type of network packet loss between the first network element device and the second network element device;
  • a packet loss identifying unit configured to receive the second statistical result sent by the second network element device, and determine, according to the first statistical result and the second statistical result that are sent by the statistical unit, the first network a type of network packet loss between the metadevice and the second network element device;
  • the second statistical result is that, after the second network element device classifies the received data packet according to the first classification manner, the number of data packets of each type of packet length in the received data packet is statistical results.
  • the packet loss identification unit includes:
  • a first subunit configured to obtain, at least one time, a statistical result of a data packet of a first type of packet length and a data packet of a second type of packet length in the first statistical result and the second statistical result, where the first The average packet length of the packet of the packet length is greater than the average packet length of the packet of the second packet length;
  • a processing subunit configured to calculate a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length, where the packet loss rate is the first statistical knot And a ratio of a difference between a number of packets of the same packet length and a number of packets of the same packet length in the first statistical result; and a discriminant subunit, When the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length is greater than a preset number of times, determining that the current network packet loss type is an error packet loss When the packet loss rate of the first packet length is less than or equal to the packet loss rate of the second packet length is greater than a preset number of times, determining that the current network packet loss type is congestion loss package.
  • the packet loss identification unit includes:
  • the statistic result of the data packet of the preset different packet length categories is obtained according to the first classification manner in the first statistic result and the second statistic result, and the preset is calculated.
  • Packet loss rate of each packet length category in different packet length categories the data packet loss rate is the number of data packets of the same type of packet length in the first statistical result and the second statistical result a ratio of a difference between the number of data packets corresponding to the same packet length in the first statistical result;
  • the comparison subunit is configured to sort the data packets of the preset different packet length categories in ascending order according to the packet length category, where the packet loss rate of the packet loss rate of the preset different packet length classes is an error If the number of packet loss ratios in the ascending order of the data packet loss rate of the preset different packet length classes is less than the preset number, it is determined that the current network packet loss type is congestion. package.
  • the first network element device further includes:
  • a fitting unit configured to linearize an average packet length of the at least one packet length class packet and a packet loss rate of the at least one packet length class when determining that the packet loss type of the current network is an error packet loss Fitting
  • the re-judging unit is configured to determine that there is an error packet loss in the current network when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope; If the packet loss rate corresponding to all packet length categories is greater than the preset packet loss rate, it is determined that the current network has error packet loss at the same time. And the congestion packet; wherein the curve obtained by the fitting takes the packet length category as the abscissa in the coordinate space, and the packet loss ratio is the ordinate.
  • the first network element device further includes: a delay re-judging unit, configured to: When it is determined that the packet loss type of the current network is an error packet loss, the delay of the second network element device currently receiving at least one data packet and the average delay of the second network element device receiving the data packet are obtained. Determining that the current delay of receiving at least one data packet is greater than the average time delay of the received data packet, determining that the current network has both error packet loss and congestion packet loss; when the current reception time of at least one data packet is less than Or equal to the average time delay of the received data packet, and determine that there is an error packet loss in the current network.
  • a delay re-judging unit configured to: When it is determined that the packet loss type of the current network is an error packet loss, the delay of the second network element device currently receiving at least one data packet and the average delay of the second network element device receiving the data packet are obtained. Determining that the current delay of receiving at least one data packet is greater than the average time delay of the received data packet, determining that the current network has both
  • a second network element device including: a data receiving unit, configured to receive a data packet sent by the first network element device; and a data classification unit, configured to: according to the data packet received by the data receiving unit, Classification is performed in a first classification manner, where the first classification manner is classified according to a preset range of a packet length or a packet length of the data packet;
  • a statistical unit configured to count the number of data packets of each type of packet length in the received data packet after the data classification unit is classified, to generate a second statistical result
  • a data sending unit configured to send the second statistical result that is forwarded by the statistic unit to the first network element device, so that the first network element device determines, according to the first statistical result and the second statistical result a type of network packet loss between the first network element device and the second network element device;
  • a packet loss identifying unit configured to receive the first statistics result sent by the first network element device, and determine the first network according to the first statistical result and the second statistical result forwarded by the statistical unit a type of network packet loss between the metadevice and the second network element device;
  • the first statistical result is that the first network element device classifies the data packets sent to the second network element device according to the first classification manner, and sends the data to the The statistical result of the number of packets of each packet length in the data packet of the second network element device.
  • the packet loss identifying unit includes: a sub-unit, configured to acquire the first time in the first statistical result and the second statistical result at least once The statistical result of the data packet of the class packet length and the data packet of the second class packet length, wherein the average packet length of the data packet of the first class packet length is greater than the average packet length of the data packet of the second class packet length;
  • a processing subunit configured to calculate a data packet loss rate of the first type of packet length and a data packet loss ratio of the second type of packet length, where the data packet loss rate is in the first statistical result And a ratio of a difference between a number of packets of the same packet length and a number of packets of the same packet length in the first statistical result; and a discriminating subunit, When the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length is greater than a preset number of times, it is determined that the packet loss type of the current network is an error packet loss; When the packet loss rate of the first packet length is less than or equal to the packet loss rate of the second packet length is greater than a preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the packet loss identifying unit includes: a sub-unit, configured to be in the first statistical result and the second statistical result according to the first
  • the categorization method is configured to obtain a statistical result of a data packet of a preset different packet length category, and calculate a packet loss rate of each packet length category in the preset different packet length categories, where the data packet loss rate is the a ratio of a difference between a number of data packets of the same type of packet length and a number of data packets corresponding to the same type of packet length in the first statistical result in the first statistical result;
  • the comparison subunit is configured to sort the data packets of the preset different packet length categories in ascending order according to the packet length category, where the packet loss rate of the packet loss rate of the preset different packet length classes is an error If the number of packet loss ratios in the ascending order of the data packet loss rate of the preset different packet length classes is less than the preset number, it is determined that the current network packet loss type is congestion. package.
  • the packet loss identification unit includes:
  • a fitting unit configured to linearize an average packet length of the at least one packet length class packet and a packet loss rate of the at least one packet length class when determining that the packet loss type of the current network is an error packet loss Fitting
  • the re-judging unit is configured to determine that there is an error packet loss in the current network when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope; If the packet loss rate of all the packet length categories is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is The packet length class is the abscissa in the coordinate space, and the packet loss rate is the ordinate.
  • the first network element device further includes: a delay re-judging unit, configured to: When it is determined that the packet loss type of the current network is an error packet loss, obtaining a delay of currently receiving at least one data packet and an average delay of receiving the data packet, when the current delay of receiving at least one data packet is greater than the receiving The average time delay of the data packet determines that the current network has both error packet loss and congestion packet loss; when the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet, determining There is an error packet loss on the current network.
  • a delay re-judging unit configured to: When it is determined that the packet loss type of the current network is an error packet loss, obtaining a delay of currently receiving at least one data packet and an average delay of receiving the data packet, when the current delay of receiving at least one data packet is greater than the receiving The average time delay of the data packet determines that the current network has both error packet loss and congestion packet loss; when the current delay of receiving at least one data packet is less than or equal to
  • the fifth aspect provides a network system, including at least one first network element device and one second network element device, where the first network element device and the second network element device are connected by using a switching network, where the first network
  • the meta-device is any one of the foregoing first network element devices
  • the second network element device is any of the foregoing second network element devices.
  • the network packet loss type identification method, device and system according to the embodiment of the present invention can accurately and quickly determine the network packet loss type by using the packet length of the data packet transmitted between the network element devices and the packet loss rate of the data packet of different packet lengths. .
  • FIG. 1 is a schematic flowchart of a network packet loss type identification method according to an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of another network packet loss type identification method according to an embodiment of the present invention
  • FIG. 3 is a schematic flowchart of a network packet loss type identification method according to another embodiment of the present invention
  • FIG. 4 is a schematic flowchart of a network packet loss type identification method according to another embodiment of the present invention
  • FIG. 6 is a schematic structural diagram of a first network element device according to an embodiment of the present invention
  • FIG. 7 is a schematic diagram of a first network element device according to an embodiment of the present invention. Another schematic diagram of the structure of the first network element device provided by the example;
  • FIG. 8 is a schematic structural diagram of still another first network element device according to an embodiment of the present invention.
  • FIG. 9 is a schematic structural diagram of still another first network element device according to an embodiment of the present invention.
  • FIG. 10 is a schematic structural diagram of a second network element device according to another embodiment of the present invention.
  • FIG. 1 is a schematic structural diagram of another second network element device according to another embodiment of the present invention.
  • FIG. 12 is a schematic structural diagram of still another first network element device according to another embodiment of the present invention.
  • FIG. 13 is a schematic structural diagram of still another first network element device according to another embodiment of the present invention.
  • FIG. 14 is a schematic structural diagram of a first network element device according to another embodiment of the present invention. Intention
  • FIG. 15 is a schematic structural diagram of a second network element device according to still another embodiment of the present invention.
  • FIG. 16 is a schematic structural diagram of a network system according to an embodiment of the present invention.
  • the technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. example. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
  • the transmission of data packets by two network element devices through the intermediate network is a basic requirement of current information communication.
  • the basic function of the intermediate network is to fully utilize the bandwidth of the two devices to exhaust the data of the two devices. Possible and accurate and reliable delivery.
  • a method for identifying a network packet loss type includes:
  • the first network element device sends a data packet to the second network element device.
  • the first network element device classifies the data packets sent to the second network element device according to the first classification manner.
  • the first classification mode is classified according to the packet length of the data packet or the preset range of the packet length.
  • the packet length category in the first classification manner is a specific packet length value
  • the packet length is
  • the preset range is classified, the average packet length value of the data packet within each preset packet length may be used as the packet length category, and the specific form of the packet length preset range may be classified according to the arithmetic progression.
  • the data packets are sorted according to the packet length, with the preset number of bits as the tolerance value, and the maximum or minimum packet length of the data packet as the initial value of the packet length, such as the maximum packet length of the data packet as the initial value of the packet length, and the tolerance value.
  • packets of all packet lengths smaller than the maximum packet length in the range of n bits are set to the same type of data packet, and each tolerance is n bits in accordance with the packet length reduction direction of the data packet.
  • the data packets in the range are set to one category, and the average packet length value of the data packets in each tolerance range is taken as the packet length category.
  • the first network element device collects the number of data packets of each packet length in the data packet sent to the second network element device, and generates a first statistical result.
  • the first network element device sends the first statistics result to the second network element device.
  • the second network element device determines, according to the first statistical result and the second statistical result, a type of network packet loss between the first network element device and the second network element device.
  • the first network element device receives the second statistical result sent by the second network element device, and the first network element device determines, according to the first statistical result and the second statistical result, the network between the first network element device and the second network element device.
  • the second statistical result is a statistical result of the number of data packets of various packet lengths in the received data packet after the second network element device classifies the received data packets according to the first classification manner.
  • the present invention provides a network packet loss type identification method, which is shown in FIG. 2 on the side of the second network element device as the data packet receiving side, and includes:
  • the second network element device receives the data packet sent by the first network element device.
  • the second network element device classifies the received data packets according to the first classification manner.
  • the first classification mode is classified according to the packet length of the data packet or the preset range of the packet length.
  • the packet length category in the first classification manner that is, the specific packet length value
  • the preset range of the packet length is classified, the average packet length value of the data packet within each preset packet length may be used as the packet length category, and the specific form of the preset range of the packet length may be the same as the equal difference.
  • the sequence sorts the data packets according to the packet length, with the preset number of bits as the tolerance value, and the maximum or minimum packet length of the data packet as the initial value of the packet length, for example, the maximum packet length of the data packet is the initial value of the packet length, If the difference is set to n bits, the packets of all packet lengths smaller than the maximum packet length in the range of n bits are set to the same type of data packet, and accordingly, each tolerance is sequentially changed according to the packet length of the data packet.
  • the data packets in the n-bit range are set to one class, and the data within each tolerance range will be The average packet length value of the packet is taken as the packet length category.
  • the second network element device collects the number of data packets of each type of packet length in the received data packet, and generates a second statistical result.
  • the second network element device sends the second statistics result to the first network element device.
  • the first network element device determines, according to the first statistical result and the second statistical result, a type of network packet loss between the first network element device and the second network element device.
  • the second network element device receives the first statistics result sent by the first network element device, and the second network element device determines, according to the first statistical result and the second statistical result, the network between the first network element device and the second network element device.
  • the first statistical result is that the data packet sent by the first network element device to the second network element device is classified according to the first classification manner, and the data packet of each packet length is sent to the data packet sent to the second network element device.
  • the statistical results of the quantity can accurately and quickly determine the network packet loss type by using the packet length of the data packet transmitted between the network element devices and the packet loss rate of the data packet of different packet lengths. .
  • an embodiment of the present invention provides a network packet loss type identification method, which provides a network element device according to a first statistical result and a second
  • the specific method for discriminating the packet loss type between network element devices includes the following steps:
  • the first network element device sends a data packet to the second network element device.
  • the first network element device classifies the data packets sent to the second network element device according to the first classification manner.
  • the first classification method is classified according to the packet length or the preset range of the packet length.
  • the first network element device collects the number of data packets of each type of packet length in the data packet sent to the second network element device, and generates a first statistical result.
  • the second network element device receives the data packet sent by the first network element device. 305. The second network element device classifies the received data packets according to the first classification manner.
  • the first classification mode is classified according to the packet length of the data packet or the preset range of the packet length.
  • the packet length category in the first classification manner is a specific packet length value
  • the packet length is
  • the preset range is classified, the average packet length value of the data packet within each preset packet length may be used as the packet length category, and the specific form of the packet length preset range may be classified according to the arithmetic progression.
  • the data packets are sorted according to the packet length, with the preset number of bits as the tolerance value, and the maximum or minimum packet length of the data packet as the initial value of the packet length, such as the maximum packet length of the data packet as the initial value of the packet length, and the tolerance value.
  • packets of all packet lengths smaller than the maximum packet length in the range of n bits are set to the same type of data packet, and each tolerance is n bits in accordance with the packet length reduction direction of the data packet.
  • the data packets in the range are set to one category, and the average packet length value of the data packets in each tolerance range is taken as the packet length category.
  • the second network element device collects the number of data packets of each type of packet length in the received data packet, and generates a second statistical result.
  • the second network element device sends the second statistical result to the first network element device, so that the first network element device determines, according to the first statistical result and the second statistical result, between the first network element device and the second network element device.
  • the type of network packet loss is the first network element device determines, according to the first statistical result and the second statistical result, between the first network element device and the second network element device.
  • the first network element device receives the second statistical result sent by the second network element device, and obtains the data packet of the first type of packet length and the second type of packet length at least once in the first statistical result and the second statistical result. According to the statistical result of the data packet, the average packet length of the packet of the first packet length is greater than the average packet length of the packet of the second packet length.
  • the average packet length is the value of the packet length in the classification, and if the classification is classified by a plurality of preset packet length ranges,
  • the average packet length here is the average of the packet length of the preset packet length range.
  • the first network element device calculates a packet loss rate of the first type of packet length and a second type.
  • the packet loss ratio is the ratio of the difference between the number of packets of the same packet length in the first statistical result and the second statistical result to the number of packets of the same packet length in the first statistical result. Specifically, according to the number of data packets M of the first type of packet length in the first statistical result and the number of data packets M of the first type of packet length in the second statistical result, the packet loss of the data packet M of the first type of packet length is obtained.
  • the rate X is obtained according to the N number of the packet of the second type packet length in the first statistical result and the N number of the packet of the second type packet length in the second statistical result, and the packet loss of the packet N of the second type packet length is obtained. Rate X.
  • the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length is greater than a preset number of times, determining that the current network packet loss type is an error packet loss;
  • the packet loss rate of a packet with a packet length is less than or equal to the packet loss rate of the packet length of the second packet is greater than a preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the packet loss rate is independent of the packet length of the data packet, and the packet loss rate of each packet length packet is balanced, and the error is caused by the packet loss rate.
  • the packet loss rate is larger as the packet length is longer.
  • the statistics of the packet of the first packet length and the packet of the second packet length are obtained at least once, and only once.
  • Error packet loss if X is less than or equal to Y, it determines that the packet loss type of the current network is congestion packet loss; however, since the first statistical result and the second statistical result are dynamically changed at each stage during data transmission, It is also possible to obtain the statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length K times, according to the result of each sampling, when the packet loss rate of the first type of packet length M is greater than the second Class packet length N packet lost When the preset number of times greater than the rate, packet loss given the type determination of the current network packet loss error; otherwise, determines the current network packet loss is a type of congestion loss.
  • step 307 may be replaced by: the first network element device sends the first statistical result to the second network element device, so that the second network element device receives the first statistical result.
  • the method performed by the first network element device in steps 308 ⁇ 3 10 may be performed according to the first statistical result and the second statistical result, which is shown in the broken line in FIG. 3, and details are not described herein again.
  • the network packet loss type identification method provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet sent by the sending device and the data packet received by the receiving device. .
  • an embodiment of the present invention provides a network packet loss type identification method, which provides a network element device according to a first statistical result and a second
  • the specific method for discriminating the packet loss type between network element devices includes the following steps:
  • the first network element device sends a data packet to the second network element device.
  • the first network element device classifies the data packets sent to the second network element device according to a first classification manner.
  • the first classification mode is classified according to the packet length of the data packet or the preset range of the packet length.
  • the packet length category in the first classification manner is a specific packet length value
  • the packet length is
  • the preset range is classified, the average packet length value of the data packet within each preset packet length may be used as the packet length category, and the specific form of the packet length preset range may be classified according to the arithmetic progression.
  • the data packets are sorted according to the packet length, with the preset number of bits as the tolerance value, and the maximum or minimum packet length of the data packet as the initial value of the packet length, such as the maximum packet length of the data packet as the initial value of the packet length, and the tolerance value.
  • packets of all packet lengths smaller than the maximum packet length in the range of n bits are set to the same type of data packet, and each tolerance is n bits in accordance with the packet length reduction direction of the data packet.
  • the data packets in the range are set to one category, and the average packet length value of the data packets in each tolerance range is taken as the packet length category.
  • the first network element device collects the number of data packets of each packet length in the data packet sent to the second network element device, and generates a first statistical result.
  • the second network element device receives the data packet sent by the first network element device.
  • the second network element device classifies the received data packets according to the first classification manner.
  • the first classification method is classified according to the packet length or the preset range of the packet length.
  • the second network element device collects the number of data packets of each type of packet length in the received data packet, and generates a second statistical result.
  • the second network element device sends the second statistical result to the first network element device, so that the first network element device determines, according to the first statistical result and the second statistical result, between the first network element device and the second network element device.
  • the type of network packet loss is the first network element device determines, according to the first statistical result and the second statistical result, between the first network element device and the second network element device.
  • the first network element device receives the second statistical result sent by the second network element device, and obtains data of a preset different packet length category according to the first classification manner in the first statistical result and the second statistical result.
  • the packet statistics result calculates the packet loss rate of each packet length category in a preset different packet length category.
  • the packet loss ratio is the ratio of the number of packets of the same packet length in the first statistical result to the second statistical result and the number of packets corresponding to the same packet length in the first statistical result;
  • the statistical results of K packets of different packet length classes may be obtained, and the packet loss rate of K different packet length classes is calculated according to the statistical result, and the corresponding packet loss rate of K packets is obtained.
  • the first network element device sorts the preset different packet length categories in ascending order according to the packet length, and presets a packet loss rate of the different packet length categories to include at least a preset number of data packet loss ratios in ascending order.
  • Arranging determining that the packet loss type of the current network is an error packet loss; if the number of packet loss ratios in the ascending order of the packet loss rate of the different packet length classes is less than the preset number, Determine the packet loss type of the current network as congestion packet loss.
  • the K packet length categories in the statistical result are sorted, and then the packet packet loss rate corresponding to the packet length category of the data packet is detected, and at least a predetermined number of packet loss packets are included in the K packet loss rate.
  • the rate is in ascending order, it is determined that the packet loss type of the current network is an error packet loss, otherwise it is determined that the packet loss type of the current network is a congestion packet loss.
  • the packet loss rate is independent of the packet length of the data packet, and the packet loss rate of each packet length is balanced, and the packet loss caused by the error is lost.
  • the packet loss rate is larger as the packet length is longer, so the number of packet loss ratios in ascending order according to the sorted result exceeds a preset number (of course, the preset number here is less than or The more the K is equal to K, the more the current network packet loss is caused by the error.
  • the probability of a packet is greater.
  • the step 407 may be replaced by: the first network element device sends the first statistical result to the second network element device, so that the second network element device may also receive the first statistical result according to the first statistical result and The second statistical result is performed by the first network element device in steps 408 to 409, which is shown in the broken line in FIG. 4, and details are not described herein again.
  • the network packet loss type identification method provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet sent by the sending device and the data packet received by the receiving device. . Further, optionally, after determining, according to the foregoing embodiment, that the packet loss type of the current network is an error packet loss, the solution further includes: a, the first network element device, and the packet length of the at least one packet length category packet The packet loss rate of the data packet of the at least one packet length category is linearly fitted;
  • the relationship between the packet loss rate and the packet length category is linearly fitted, and the fitting method is not limited.
  • the packet loss rate is J.
  • the slope value of the curve 1 obtained by the fitting is an upward trend, it is determined that there is only an error packet loss in the current network; or the curve obtained by the fitting is a fixed slope K1, that is, the fitted slope is K1.
  • the i and j are respectively the abscissa (packet length category) and the ordinate (packet)
  • the corresponding packet loss rate is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss;
  • the linear fitting is a curve, since the slope is an upward trend, as long as the J at the coordinate point (0, J) is larger than the preset packet loss rate, it is determined that the current network has both error packet loss and Congestion is lost.
  • the solution further includes:
  • the second network element device obtains a delay of receiving, by the second network element device, the at least one data packet, and an average delay of receiving the data packet by the second network element device, where a delay of receiving at least one data packet is greater than an average of the received data packet.
  • the time delay determines that the current network has both error packet loss and congestion packet loss; when the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet, it is determined that the current network has error packet loss.
  • the delay of receiving the at least one data packet by the second network element device in a time period is compared with the global average delay, and the delay of receiving at least one data packet in the one time period is greater than the global average time delay. Indicates the possibility of congestion and packet loss on the network.
  • the step a2 is performed by the first network element device to obtain the relevant delay from the second network element device.
  • the packet loss type of the current network is determined to be an error packet loss
  • the second network element is determined.
  • the step a2 can also be directly executed by the second network element device. Since the congestion packet loss is mainly caused by the network condition, if the packet loss factor of the current network includes congestion packet loss, the delay of receiving the data packet by the receiving end increases, and therefore, when the receiving at least one data packet is delayed If the average time delay of the received data packet is greater than the average time delay of the received data packet, it is determined that the current network has both error packet loss and congestion packet loss; otherwise, it is determined that the current network has error packet loss.
  • An embodiment of the present invention provides a first network element device, where the first network element device may be any node network element device in a network architecture, such as a switching device, a remote radio unit, a base station, a mobile terminal, or a base station side. Any device capable of transmitting and receiving data packets in a network architecture, such as a unit, a module, or an entity for transmitting and receiving a data packet, in the present invention
  • the specific form of the network element device is not specifically limited, and the method for identifying the network packet loss type provided by the embodiment of the present invention may be implemented. For details, refer to FIG.
  • the data sending unit 61 is configured to send a data packet to the second network element device
  • the data classification unit 62 is configured to classify the data packet sent by the data sending unit 61 to the second network element device according to the first classification manner.
  • the categorization method is classified according to the packet length or the preset range of the packet length
  • the statistic unit 63 is configured to collect the number of packets of each packet length in the data packet classified by the statistic data classification unit 62, and generate the first statistical result. ;
  • the data sending unit 61 is further configured to send the first statistical result forwarded by the statistic unit 63 to the second network element device, so that the second network element device determines the first network element device according to the first statistical result and the second statistical result. Type of network packet loss with the second network element device; or
  • the packet loss identifying unit 64 is configured to receive the second statistic result sent by the second network element device, and determine, according to the first statistic result and the second statistic result forwarded by the statistic unit 63, between the first network element device and the second network element device.
  • the type of network packet loss; wherein, the second statistical result is a statistical result of the number of packets of each packet length in the received data packet after the second network element device classifies the received data packet according to the first classification manner. .
  • the packet loss identification unit 64 includes:
  • a sub-segment unit 641 configured to obtain, in at least one of the first statistical result and the second statistical result, a statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length, the first type of packet length
  • the average packet length of the data packet is greater than the average packet length of the second packet length packet
  • the processing sub-unit 642 is configured to calculate a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length, where the packet loss rate is the first statistical result and the second The ratio of the number of packets of the same packet length in the statistical result to the number of packets of the same packet length in the first statistical result;
  • the determining sub-unit 643 is configured to determine that the packet loss type of the current network is an error when the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length is greater than a preset number of times. Packet loss; When the packet loss rate of the first packet length is less than or equal to the packet loss rate of the second packet length is greater than the preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the packet loss identification unit 64 includes:
  • the sub-segment unit 644 is configured to obtain, according to the first classification manner, the statistical result of the preset data packet of different packet length categories in the first statistical result and the second statistical result, and calculate each of the preset different packet length categories.
  • the packet loss rate of the packet length class, the packet loss rate of the packet is the same as the packet of the same packet length in the first statistical result and the second statistical result, and the same packet in the first statistical result.
  • the ratio of the number of data packets corresponding to the length; the comparison sub-unit 645 is configured to sort the data packets of the preset different packet length categories in ascending order according to the packet length category, and preset the packet loss rate of different packet length classes.
  • the packet loss rate of at least the preset number of packets is in ascending order, it is determined that the packet loss type of the current network is an error packet loss; if the packet loss rate of the different packet length classes is preset, the packet loss rate is in ascending order. If the number of packet loss ratios is less than the preset number, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the first network element device 6 when the first network element device 6 includes the packet loss identification unit 64, the first network element device 6 further includes: a fitting unit 65, configured to determine a packet loss type of the current network. When packet loss occurs for an error, linearly fitting an average packet length of at least one packet length class packet with a packet loss rate of at least one packet length class;
  • the re-judging unit 66 is configured to determine that there is an error packet loss in the current network when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope; If the packet loss rate corresponding to the packet length category is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is in the packet length category.
  • the absc issa in the coordinate space, with the packet loss rate as the ordinate.
  • the first network element device 6 when the first network element device 6 includes a packet loss identification unit At 64 o'clock, the first network element device 6 further includes: a delay re-judging unit 67, configured to: when determining that the packet loss type of the current network is an error packet loss, acquiring, when the second network element device currently receives at least one data packet Extending the average delay of receiving the data packet by the second network element device, when the current delay of receiving at least one data packet is greater than the average time delay of the received data packet, determining that the current network has both error packet loss and congestion packet loss; When the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet, it is determined that there is an error packet loss in the current network.
  • the first network element device provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet received by the receiving device and the receiving device.
  • An embodiment of the present invention provides a second network element device, where the second network element device may be any node network element device in a network architecture, such as a switching device, a remote radio unit, a base station, a mobile terminal, or a base station side.
  • the device, the module, the entity, and the like, which are used for the data transmission and reception of the data packet are not limited to the specific form of the network element device in the embodiment of the present invention, so that the present invention can be implemented.
  • the above-mentioned network packet loss type identification method provided by the embodiment is applicable. As shown in FIG.
  • the network element device includes: a data receiving unit 71, configured to receive a data packet sent by the first network element device;
  • the data classification unit 72 is configured to classify the data packets received by the data receiving unit 71 according to a first classification manner, where the first classification manner is classified according to a preset range of a packet length or a packet length of the data packet;
  • the statistic unit 73 is configured to generate, by the statistical data classification unit 72, the number of data packets of each type of packet length in the received data packet, and generate a second statistical result;
  • the data sending unit 74 is configured to send the second statistical result forwarded by the statistic unit 73 to the first network element device, so that the first network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second Type of network packet loss between NE devices; or,
  • the packet loss identifying unit 75 is configured to receive the first statistical result sent by the first network element device, and determine the first statistical result and the second statistical result that are forwarded by the statistical unit 73.
  • the packet loss identification unit 75 includes:
  • the sub-unit 75 1 is configured to obtain, in the first statistical result and the second statistical result, the statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length at least once, the first type of packet length
  • the average packet length of the data packet is greater than the average packet length of the second packet length packet
  • the processing subunit 752 is configured to calculate a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length, where the packet loss rate is the same as the second statistical result in the first statistical result.
  • the determining sub-unit 753 is configured to determine that the packet loss type of the current network is an error when the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length by a preset number of times. Packet loss; When the packet loss rate of the first packet length is less than or equal to the packet loss rate of the second packet length is greater than the preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the packet loss identification unit 75 includes:
  • the sub-unit 754 is configured to obtain, according to the first classification manner, the statistical result of the data packet of the preset different packet length categories in the first statistical result and the second statistical result, and calculate each of the preset different packet length categories.
  • the packet loss rate of the packet length class, the packet loss rate of the packet is the same as the number of packets of the same packet length in the first statistical result and the second statistical result.
  • the comparison subunit 755 is configured to sort the data packets of the preset different packet length categories in ascending order according to the packet length category, and preset the packet loss rate of the different packet length categories to include at least a preset number of data packet loss. If the packet rate is in ascending order, it is determined that the packet loss type of the current network is an error packet loss; if the packet loss rate of the packet length of the different packet length classes is preset, the number of packet loss ratios in the ascending order is smaller than the packet loss rate. Predetermine the number to determine the current network The type of packet loss is congestion and packet loss.
  • the second network element device further includes: a fitting unit 76, configured to: when determining that the packet loss type of the current network is an error packet loss, at least one packet of the packet length category The average packet length is linearly fitted to the packet loss rate of at least one packet length class;
  • the re-judging unit 77 is configured to determine that there is an error packet loss in the current network when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope; If the packet loss rate corresponding to the packet length category is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is in the packet length category.
  • the absc issa in the coordinate space, with the packet loss rate as the ordinate.
  • the second network element device further includes: a delay re-judging unit 78, configured to acquire the second network element when determining that the packet loss type of the current network is an error packet loss The delay of receiving the at least one data packet by the device and the average delay of receiving the data packet by the second network element device.
  • a delay re-judging unit 78 configured to acquire the second network element when determining that the packet loss type of the current network is an error packet loss The delay of receiving the at least one data packet by the device and the average delay of receiving the data packet by the second network element device.
  • the current delay of receiving at least one data packet is greater than the average time delay of the received data packet, determining that the current network exists simultaneously Error packet loss and congestion packet loss;
  • the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet, it is determined that there is an error packet loss in the current network.
  • the second network element device provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet sent by the sending device and the received data packet.
  • An embodiment of the present invention provides a first network element device, where the first network element device may be any node network element device in a network architecture, such as a switching device, a remote radio unit, a base station, a mobile terminal, or a base station side.
  • the device, the module, the entity, or the like, which is used for the data transmission and reception of the data packet is not limited to the specific form of the network element device in the embodiment of the present invention, so that the present invention can be implemented.
  • the network packet loss type identification method provided by the embodiment is as follows.
  • the network element device includes: at least one processor 81, a memory 82, a communication port 83, and a bus 84.
  • the at least one processor 81, the memory 82, and the communication interface 83 are connected by a bus 84 and complete communication with each other.
  • the bus 84 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Wait.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component
  • EISA Extended Industry Standard Architecture
  • the bus 84 can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 14, but it does not mean that there is only one bus or one type of bus. among them:
  • Memory 82 is for storing executable program code, including computer operating instructions.
  • Memory 82 may contain high speed RAM memory and may also include non-volatile memory, such as at least one disk memory.
  • the storage device stores: an operating system, an application, and program code for implementing an embodiment of the present invention.
  • the operating system is for controlling and implementing processing functions performed by the processing unit.
  • the application contains program code, such as word processing software, email software.
  • the processor 81 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more configured to implement the embodiments of the present invention. integrated circuit.
  • CPU central processing unit
  • ASIC application specific integrated circuit
  • the communication interface 83 is mainly used to implement communication between devices in the embodiment.
  • the processor 81 is configured to send a data packet to the second network element device by using the communication port 83.
  • the data packets sent by the processor 81 to the second network element device through the communication port 83 are classified according to the first classification manner, and the first classification manner is classified according to the packet length of the data packet or the preset range of the packet length;
  • the processor 81 is configured to collect, according to the number of packets of various packet lengths in the classified data packet, to generate a first statistical result
  • the processor 81 is configured to send, by using the communication port 83, the first statistical result to the second network element device, so that the second network element device determines, according to the first statistical result and the second statistical result, the first network element device and the second network element.
  • Type of network packet loss between devices or, The processor 81 is configured to receive, by using the communication port 83, the second statistical result sent by the second network element device, and determine, according to the first statistical result and the second statistical result, network packet loss between the first network element device and the second network element device. type;
  • the second statistical result is a statistical result of the number of data packets of various packet lengths in the received data packet after the second network element device classifies the received data packets according to the first classification manner.
  • the processor 81 in the first statistical result and the second statistical result, obtain the statistical result of the data packet of the first type of packet length and the data packet of the second type of packet length at least once, and the data packet of the first type of packet length The average packet length of the data packet whose average packet length is greater than the packet length of the second packet;
  • the processor 81 is configured to calculate a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length, where the packet loss rate is the same in the first statistical result and the second statistical result.
  • the processor 81 is configured to determine that the packet loss type of the current network is an error when the packet loss rate of the first packet length is greater than the packet loss rate of the second packet length. When the packet loss rate of the first packet length is less than or equal to the packet loss rate of the second packet length is greater than a preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the processor 81 is configured to obtain, according to the first classification manner, the statistical result of the preset data packet of different packet length categories in the first statistical result and the second statistical result, and calculate a preset different packet length.
  • the packet loss rate of each packet length category in the category, the packet loss rate of the packet in the first statistical result and the same packet length in the second statistical result are the same as the first statistical result.
  • the ratio of the number of data packets corresponding to the packet length; the processor 81 is configured to sort the data packets of the preset different packet length categories in ascending order according to the packet length category, and preset the packet loss rate of different packet length classes.
  • the packet loss rate of at least the preset number of packets is in ascending order, and it is determined that the packet loss type of the current network is an error packet loss; if the packet loss rate of the packet of different packet lengths is preset, the packet loss rate is in ascending order. If the number of packet loss ratios of the column is less than the preset number, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the processor 81 is further configured to: when determining that the packet loss type of the current network is an error packet loss, the average packet length of the at least one packet length category packet and the at least one packet length category packet.
  • the packet loss rate is linearly fitted;
  • the processor 81 is configured to: when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope, determine that the current network has an error packet loss; when all the packets in the curve obtained by the fitting are obtained If the packet loss rate corresponding to the long category is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is in the packet length category.
  • the packet loss rate is optional on the ordinate.
  • the processor 81 is configured to: when determining that the packet loss type of the current network is an error packet loss, obtain, by using the communication interface 83, a delay that the second network element device currently receives at least one data packet, and the second network element device receives the data packet. Average delay, when the current delay of receiving at least one data packet is greater than the average time delay of the received data packet, it is determined that the current network has both error packet loss and congestion packet loss; when the current reception time of at least one data packet is less than Or equal to the average time delay of receiving the data packet, and determine that there is an error packet loss in the current network.
  • the first network element device provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet received by the receiving device and the receiving device.
  • An embodiment of the present invention provides a second network element device, where the network element device may be any one of the network element devices in the network architecture, such as a switching device, a remote radio unit, a base station, a mobile terminal, or a base station.
  • the specific form of the network element device is not specifically limited in the embodiment of the present invention, and the implementation of the present invention can be implemented. Any of the above network packet loss type identification methods provided by the example shall prevail, and the specific reference picture
  • the network element device includes: at least one processor 91, a memory 92, a communication port 93, and a bus 94, the at least one processor 91, the memory 92, and the communication
  • the interfaces 93 are connected by a bus 94 and complete communication with each other.
  • the bus 94 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Wait.
  • ISA Industry Standard Architecture
  • PCI Peripheral Component
  • EISA Extended Industry Standard Architecture
  • the bus 94 can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 15, but it does not mean that there is only one bus or one type of bus. among them:
  • Memory 92 is for storing executable program code, the program code including computer operating instructions.
  • Memory 92 may contain high speed RAM memory and may also include non-volatile memory, such as at least one disk memory.
  • the storage device stores: an operating system, an application, and program code for implementing an embodiment of the present invention.
  • the operating system is for controlling and implementing processing functions performed by the processing unit.
  • the application contains program code, such as word processing software, email software.
  • the processor 91 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more configured to implement the embodiments of the present invention. integrated circuit.
  • CPU central processing unit
  • ASIC application specific integrated circuit
  • the communication interface 93 is mainly used to implement communication between devices in this embodiment.
  • the processor 91 is configured to receive, by using the communication port 93, a data packet sent by the first network element device.
  • the processor 91 is configured to classify the data packets received through the communication port 93 according to the first classification manner, where the first classification manner is classified according to a preset range of a packet length or a packet length of the data packet;
  • the processor 91 is configured to generate, by using the number of data packets of each type of packet length in the received data packet, to generate a second statistical result
  • the processor 91 is configured to send the second statistical result to the first network element device by using the communication port 93, so that the first network element device determines the first network element device and the second network element according to the first statistical result and the second statistical result.
  • Type of network packet loss between devices or,
  • the processor 91 is configured to receive, by using the communication port 93, the first statistics result sent by the first network element device, and determine, according to the first statistical result and the second statistical result, network packet loss between the first network element device and the second network element device.
  • the first statistical result is that after the first network element device classifies the data packets sent to the second network element device according to the first classification manner, the packet lengths of the data packets sent to the second network element device are The statistical result of the number of packets.
  • the processor 91 is configured to obtain, according to at least one of the first statistical result and the second statistical result, a statistical result of the data packet of the first class packet length and the data packet of the second class packet length, the first class.
  • the average packet length of the packet length packet is greater than the average packet length of the packet of the second packet length;
  • the processor 91 is configured to calculate a packet loss rate of the first type of packet length and a packet loss rate of the second type of packet length, where the packet loss rate is the same in the first statistical result and the second statistical result. The ratio of the difference between the number of packets of the packet length and the number of packets of the same packet length in the first statistical result; the processor 91, configured to reduce the packet loss rate of the first packet length to be greater than the second packet length When the packet loss rate of the packet length is greater than the preset number of times, it is determined that the packet loss type of the current network is an error packet loss; when the packet loss rate of the first packet length is less than or equal to the second packet type When the number of long packet loss ratios is greater than the preset number of times, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the processor 91 is configured to obtain, according to the first classification manner, the statistics result of the preset data packets of different packet length categories in the first statistical result and the second statistical result, and calculate a preset different packet.
  • the packet loss rate of each packet length category in the long category, the packet loss rate of the packet is the same as the first statistical result in the first statistical result and the second packet in the second statistical result.
  • the processor 91 is configured to sort the data packets of different preset packet length categories in ascending order according to the packet length category, and preset a packet loss rate of the different packet length categories to include at least a preset number of data packet loss packets. If the rate is in ascending order, it is determined that the packet loss type of the current network is an error packet loss; if the packet loss rate of the different packet length classes is preset, the packet loss rate is in ascending order. If the number of packet loss ratios of the column is less than the preset number, it is determined that the packet loss type of the current network is a congestion packet loss.
  • the processor 91 is configured to: when determining that the packet loss type of the current network is an error packet loss, the average packet length of the at least one packet length class packet and the at least one packet length class packet.
  • the packet loss rate is linearly fitted;
  • the processor 91 is configured to: when the slope of the curve obtained by the fitting is an upward trend or the slope of the curve is a fixed value greater than a preset slope, determine that the current network has an error packet loss; when all the packets in the curve are obtained by fitting If the packet loss rate corresponding to the long category is greater than the preset packet loss rate, it is determined that the current network has both error packet loss and congestion packet loss; wherein the curve obtained by the fitting is in the packet length category. For the abscissa in the coordinate space, the packet loss rate is the ordinate.
  • the processor 91 is configured to: when determining that the packet loss type of the current network is an error packet loss, obtain a delay of receiving the at least one data packet and an average delay of receiving the data packet, when currently receiving at least one data If the delay of the packet is greater than the average time delay of the received data packet, it is determined that the current network has both error packet loss and congestion packet loss; when the current delay of receiving at least one data packet is less than or equal to the average time delay of the received data packet. Determines that there is an error packet loss on the current network.
  • the second network element device provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet sent by the sending device and the received data packet.
  • a network system includes at least one first network element device 101 and one second network element device 102, where the first network element device 101 and the second network element device 102 are provided.
  • the first network element device 101 is any one of the first network element devices provided by the foregoing device embodiment
  • the second network element device 102 is any second network element device provided by the foregoing device embodiment.
  • the network system provided by the embodiment of the present invention may apply the network element device provided by the embodiment of the present invention to implement the network packet loss type identification method provided by the embodiment of the present invention, where each network element included in the system
  • the specific structure of the device and the network packet loss type identification method implemented by each network element device are described in the above description of the method and device, and are not described here.
  • the network system provided by the embodiment of the present invention can accurately and quickly determine the network packet loss type according to the dynamic change of the packet length and the packet loss rate of the data packet sent by the sending device and the data packet received by the receiving device.
  • a person skilled in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by using hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed.
  • the foregoing steps include the steps of the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

Landscapes

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

Abstract

本发明的实施例提供一种网络丟包类型识别方法、设备及系统,涉及通信领域,能够精确快速的判断出网络丟包类型。该方法包括:第一网元设备向第二网元设备发送数据包;第一网元设备将发送至第二网元设备的数据包按第一分类方式进行分类;第一网元设备统计发送至第二网元设备的数据包中各类包长的数据包数量,生成第一统计结果;第一网元设备将第一统计结果发送至第二网元设备,以便第二网元设备根据第一统计结果及第二统计结果判断第一网元设备与第二网元设备间的网络丟包的类型。本发明的实施例应用于网络数据包丟包类型识别。

Description

一种网络丢包类型识别方法、 设备及系统 技术领域 本发明涉及通信领域, 尤其涉及一种网络丟包类型识别方法、 设备 及系统。
背景技术
两个设备通过中间网络传输数据包是当前信息通信的基本需 求, 中间网络的基本功能就是充分利用 自 己的带宽将两个设备的数 据包尽可能多的且准确可靠的传送。 但是由于各种原因, 数据包在 中间网络传输时会出现丟包, 从而影响两个设备之前的通信。 目前丟包主要由两种原因引起的, 一种是中间带宽不足, 网络 出现拥塞而引起丟包; 另外一种是中间网络出现传输错包, 引起网 络主动丟包, 如微波传输出现误码, 产品错包。 网络丟包会影响上层业务质量,必须釆用一定的方法即要充分 利用网络带宽提高经济效益, 也要降低丟包率, 保障业务体验。 对 于第一种原因引起的丟包, 两端的用户设备可以通过降速解除拥 塞, 对于第二种原因引起的丟包, 发生端设备可以用冗余传输抗传 输错包。 前提用户端设备必须判断清楚是哪种原因引起的丟包, 如 果判断出错,反而引起进一步恶化传输。如果实际情况为误码丟包, 但错判为拥塞丟包, 设备企图釆用降速来解除拥塞, 但实际上仍不 能降低丟包率, 导致设备不断的降速, 直到数据包传输速掉底。 另 外, 如果实际情况为拥塞丟包, 但错判为传输错包丟包, 设备两端 釆用冗余传输抗丟包, 但反而加重网络负载, 从而引起更严重的拥 塞, 有可能引起中间网络崩溃。
然而 艮多情况下,两端设备无法与中间网络一起配合获取网络 丟包的原因, 因此目前主要还是釆用事前人工评估是拥塞引起的丟 包, 还是传输出错引起的丟包, 从配置相关的抗网络拥塞或抗错包 的功能, 在上述过程中发明人发现, 现有的技术不能准确的判断丟 包原因及丟包原因的动态变化, 因此不能精确快速判断网络丟包类 型。 发明内容 本发明的实施例提供一种网络丟包类型识别方法、 设备及系 统, 能够准确的根据网络丟包原因及丟包原因的动态变化, 精确 快速的判断出网络丟包类型。 为达到上述目 的, 本发明的实施例釆用如下技术方案:
一方面, 提供一种网络丟包类型识别方法, 包括:
第一网元设备向第二网元设备发送数据包; 所述第一网元设备将发送至所述第二网元设备的数据包按第 一分类方式进行分类,所述第一分类方式为按所述数据包的包长或 包长的预设范围分类; 所述第一网元设备统计所述发送至所述第二网元设备的数据 包中各类包长的数据包数量, 生成第一统计结果;
所述第一网元设备将所述第一统计结果发送至所述第二网元 设备,以便所述第二网元设备根据所述第一统计结果及第二统计结 果判断所述第一网元设备与所述第二网元设备间的网络丟包的类 型; 或者,
所述第一网元设备接收所述第二网元设备发送的所述第二统 计结果,所述第一网元设备根据所述第一统计结果和所述第二统计 结果判断所述第一网元设备与所述第二网元设备间的网络丟包的 类型;
其中,所述第二统计结果为所述第二网元设备将接收到的数据 包按所述第一分类方式分类后,对所述接收到的数据包中各类包长 的数据包数量的统计结果。
在第一种可能的实现方式中, 结合第一方面, 所述第一网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元 设备与所述第二网元设备间的网络丟包的类型包括: 在所述第一统计结果中和所述第二统计结果中至少一次获取 第一类包长的数据包和第二类包长的数据包的统计结果,所述第一 类包长的数据包的平均包长大于所述第二类包长的数据包的平均 包长;
计算所述第一类包长的数据包丟包率及所述第二类包长的数 据包丟包率,所述数据包丟包率为所述第一统计结果中和所述第二 统计结果中同一类包长的数据包的个数差值与所述第一统计结果 中所述同一类包长的数据包的个数的比值; 当所述第一类包长的数据包丟包率大于所述第二类包长的数 据包丟包率的次数大于预设的次数时,判定当前网络的丟包类型为 误码丟包; 当所述第一类包长的数据包丟包率小于或等于所述第二类包 长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟包 类型为拥塞丟包。
在第二种可能的实现方式中, 结合第一方面, 所述第一网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元 设备与所述第二网元设备间的网络丟包的类型包括: 在所述第一统计结果中和所述第二统计结果中根据所述第一 分类方式, 获取预设个不同包长类别的数据包的统计结果, 计算所 述预设个不同包长类别中各个包长类别的数据包丟包率,所述数据 包丟包率为所述第一统计结果中和所述第二统计结果中同一类包 长的数据包的个数差值与所述第一统计结果中所述同一类包长对 应的数据包的个数的比值;
将所述预设个不同包长类别的数据包按照包长类别进行升序 排序,所述预设个不同包长类别的数据包丟包率中至少包括预设个 数的数据包丟包率为升序排列,则判定当前网络的丟包类型为误码 丟包;
若所述预设个不同包长类别的数据包丟包率中按照升序排列 的数据包丟包率的个数小于所述预设个数,则判定当前网络的丟包 类型为拥塞丟包。
在第三种可能的实现方式中, 结合第一方面, 第一种或第二种 可能的实现方式中的任意一种, 所述方法还包括:
当判定当前网络的丟包类型为误码丟包时,将至少一个包长类 别的数据包的平均包长与所述至少一个包长类别的数据包丟包率 进行线性拟合; 当拟合得到的曲线的斜率为上升趋势或者所述曲线的斜率为 大于预设斜率的固定值时, 则判定当前网络存在误码丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中所述拟合得到的曲线以所述包长类别为坐标空间 中的横坐标, 以所述数据包丟包率为纵坐标。 在第四种可能的实现方式中, 结合第一方面, 第一种或第二种 可能的实现方式中的任意一种, 所述方法包括:
当判定当前网络的丟包类型为误码丟包时,获取所述第二网元 设备当前接收至少一个数据包的时延及所述第二网元设备接收数 据包的平均时延, 当所述当前接收至少一个数据包的时延大于所述 接收数据包的平均时延时,则判定当前网络同时存在误码丟包及拥 塞丟包;
当所述当前接收至少一个数据包的时延小于或等于所述接收 数据包的平均时延时, 判定当前网络存在误码丟包。
第二方面, 提供一种网络丟包类型识别方法, 包括 第二网元设备接收第一网元设备发送的数据包; 将接收到的数据包按照第一分类方式进行分类,所述第一分类 方式为按所述数据包的包长或包长的预设范围分类;
所述第二网元设备统计所述接收到的数据包中各类包长的数 据包数量, 生成第二统计结果;
所述第二网元设备将所述第二统计结果发送至所述第一网元 设备,以便所述第一网元设备根据所述第一统计结果及第二统计结 果判断所述第一网元设备与所述第二网元设备间的网络丟包的类 型; 或者,
所述第二网元设备接收所述第一网元设备发送的所述第一统 计结果,所述第二网元设备根据所述第一统计结果和所述第二统计 结果判断所述第一网元设备与所述第二网元设备间的网络丟包的 类型;
其中,所述第一统计结果为所述第一网元设备将发送至所述第 二网元设备的数据包按所述第一分类方式分类后,对所述发送至所 述第二网元设备的数据包中各类包长的数据包数量的统计结果。 在第一种可能的实现方式中, 结合第二方面, 所述第二网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元 设备与所述第二网元设备间的网络丟包的类型包括: 在所述第一统计结果中和所述第二统计结果中至少一次获取 第一类包长的数据包和第二类包长的数据包的统计结果,所述第一 类包长的数据包的平均包长大于所述第二类包长的数据包的平均 包长;
计算所述第一类包长的数据包丟包率及所述第二类包长的数 据包丟包率,所述数据包丟包率为所述第一统计结果中和所述第二 统计结果中同一类包长的数据包的个数差值与所述第一统计结果 中所述同一类包长的数据包的个数的比值; 当所述第一类包长的数据包丟包率大于所述第二类包长的数 据包丟包率的次数大于预设的次数时,判定当前网络的丟包类型为 误码丟包; 当所述第一类包长的数据包丟包率小于或等于所述第二类包 长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟包 类型为拥塞丟包。
在第二种可能的实现方式中, 结合第二方面, 所述第二网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元 设备与所述第二网元设备间的网络丟包的类型包括: 在所述第一统计结果中和所述第二统计结果中根据所述第一 分类方式, 获取预设个不同包长类别的数据包的统计结果, 计算所 述预设个不同包长类别中各个包长类别的数据包丟包率,所述数据 包丟包率为所述第一统计结果中和所述第二统计结果中同一类包 长的数据包的个数差值与所述第一统计结果中所述同一类包长对 应的数据包的个数的比值; 将所述预设个不同包长类别的数据包按照包长类别进行升序 排序,所述预设个不同包长类别的数据包丟包率中至少包括预设个 数的数据包丟包率为升序排列,则判定当前网络的丟包类型为误码 丟包;
若所述预设个不同包长类别的数据包丟包率中按照升序排列 的数据包丟包率的个数小于所述预设个数,则判定当前网络的丟包 类型为拥塞丟包。
在第三种可能的实现方式中, 结合第二方面, 第一种或第二种 可能的实现方式中的任意一种, 所述方法还包括:
当判定当前网络的丟包类型为误码丟包时,将至少一个包长类 别的数据包的平均包长与所述至少一个包长类别的数据包丟包率 进行线性拟合; 当拟合得到的曲线的斜率为上升趋势或者所述曲线的斜率为 大于预设斜率的固定值时, 则判定当前网络存在误码丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中所述拟合得到的曲线以所述包长类别为坐标空间 中的横坐标, 以所述数据包丟包率为纵坐标。
在第四种可能的实现方式中, 结合第一方面, 第一种或第二种 可能的实现方式中的任意一种, 所述方法还包括:
当判定当前网络的丟包类型为误码丟包时,获取当前接收至少 一个数据包的时延及接收数据包的平均时延, 当所述当前接收至少 一个数据包的时延大于所述接收数据包的平均时延时,则判定当前 网络同时存在误码丟包及拥塞丟包; 当所述当前接收至少一个数据包的时延小于或等于所述接收 数据包的平均时延时, 判定当前网络存在误码丟包。
第三方面, 提供一种第一网元设备, 包括: 数据发送单元, 用于向第二网元设备发送数据包;
数据分类单元,用于将所述数据发送单元发送至所述第二网元 设备的数据包按第一分类方式进行分类,所述第一分类方式为按所 述数据包的包长或包长的预设范围分类;
统计单元,用于统计所述数据分类单元分类后的所述数据包中 各类包长的数据包数量, 生成第一统计结果;
所述数据发送单元,还用于将所述统计单元转发的所述第一统 计结果发送至所述第二网元设备,以便所述第二网元设备根据所述 第一统计结果及第二统计结果判断所述第一网元设备与所述第二 网元设备间的网络丟包的类型; 或者,
丟包识别单元,用于接收所述第二网元设备发送的所述第二统 计结果,根据所述统计单元转发的所述第一统计结果和所述第二统 计结果判断所述第一网元设备与所述第二网元设备间的网络丟包 的类型;
其中,所述第二统计结果为所述第二网元设备将接收到的数据 包按所述第一分类方式分类后,对所述接收到的数据包中各类包长 的数据包数量的统计结果。
在第一种可能的实现方式中, 结合第三方面, 所述丟包识别单 元, 包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果 中至少一次获取第一类包长的数据包和第二类包长的数据包的统 计结果,所述第一类包长的数据包的平均包长大于所述第二类包长 的数据包的平均包长;
处理子单元,用于计算所述第一类包长的数据包丟包率及所述 第二类包长的数据包丟包率,所述数据包丟包率为所述第一统计结 果中和所述第二统计结果中同一类包长的数据包的个数差值与所 述第一统计结果中所述同一类包长的数据包的个数的比值; 判别子单元,用于当所述第一类包长的数据包丟包率大于所述 第二类包长的数据包丟包率的次数大于预设的次数时,判定当前网 络的丟包类型为误码丟包; 当所述第一类包长的数据包丟包率小于 或等于所述第二类包长的数据包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型为拥塞丟包。
在第二种可能的实现方式中, 结合第三方面, 所述丟包识别单 元包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果 中根据所述第一分类方式,获取预设个不同包长类别的数据包的统 计结果,计算所述预设个不同包长类别中各个包长类别的数据包丟 包率,所述数据包丟包率为所述第一统计结果中和所述第二统计结 果中同一类包长的数据包的个数差值与所述第一统计结果中所述 同一类包长对应的数据包的个数的比值;
对比子单元,用于将所述预设个不同包长类别的数据包按照包 长类别进行升序排序,所述预设个不同包长类别的数据包丟包率中 丟包类型为误码丟包;若所述预设个不同包长类别的数据包丟包率 中按照升序排列的数据包丟包率的个数小于所述预设个数,则判定 当前网络的丟包类型为拥塞丟包。
在第三种可能的实现方式中, 结合第三方面, 第一种或第二种 可能的实现方式中的任意一种, 所述第一网元设备还包括:
拟合单元, 用于当判定当前网络的丟包类型为误码丟包时, 将 至少一个包长类别的数据包的平均包长与所述至少一个包长类别 的数据包丟包率进行线性拟合;
重判单元,用于当拟合得到的曲线的斜率为上升趋势或者所述 曲线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码 丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中所述拟合得到的曲线以所述包长类别为坐标空间 中的横坐标, 以所述数据包丟包率为纵坐标。
在第四种可能的实现方式中, 结合第三方面, 第一种或第二种 可能的实现方式中的任意一种, 所述第一网元设备还包括: 时延重判单元, 用于当判定当前网络的丟包类型为误码丟包 时,获取所述第二网元设备当前接收至少一个数据包的时延及所述 第二网元设备接收数据包的平均时延, 当所述当前接收至少一个数 据包的时延大于所述接收数据包的平均时延时,则判定当前网络同 时存在误码丟包及拥塞丟包; 当所述当前接收至少一个数据包的时 延小于或等于所述接收数据包的平均时延时,判定当前网络存在误 码丟包。
第四方面, 提供一种第二网元设备, 包括: 数据接收单元, 用于接收第一网元设备发送的数据包; 数据分类单元,用于将所述数据接收单元接收到的数据包按照 第一分类方式进行分类,所述第一分类方式为按所述数据包的包长 或包长的预设范围分类;
统计单元,用于统计所述数据分类单元分类后的所述接收到的 数据包中各类包长的数据包数量, 生成第二统计结果;
数据发送单元,用于将所述统计单元转发的所述第二统计结果 发送至所述第一网元设备,以便所述第一网元设备根据所述第一统 计结果及第二统计结果判断所述第一网元设备与所述第二网元设 备间的网络丟包的类型; 或者,
丟包识别单元,用于接收所述第一网元设备发送的所述第一统 计结果,根据所述统计单元转发的所述第一统计结果和所述第二统 计结果判断所述第一网元设备与所述第二网元设备间的网络丟包 的类型;
其中,所述第一统计结果为所述第一网元设备将发送至所述第 二网元设备的数据包按所述第一分类方式分类后,对所述发送至所 述第二网元设备的数据包中各类包长的数据包数量的统计结果。 在第一种可能的实现方式中, 结合第三方面, 所述丟包识别单 元包括: 釆样子单元,用于在所述第一统计结果中和所述第二统计结果 中至少一次获取第一类包长的数据包和第二类包长的数据包的统 计结果,所述第一类包长的数据包的平均包长大于所述第二类包长 的数据包的平均包长;
处理子单元,用于计算所述第一类包长的数据包丟包率及所述 第二类包长的数据包丟包率,所述数据包丟包率为所述第一统计结 果中和所述第二统计结果中同一类包长的数据包的个数差值与所 述第一统计结果中所述同一类包长的数据包的个数的比值; 判别子单元,用于当所述第一类包长的数据包丟包率大于所述 第二类包长的数据包丟包率的次数大于预设的次数时,判定当前网 络的丟包类型为误码丟包; 当所述第一类包长的数据包丟包率小于 或等于所述第二类包长的数据包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型为拥塞丟包。
在第二种可能的实现方式中, 结合第三方面, 所述丟包识别单 元包括: 釆样子单元,用于在所述第一统计结果中和所述第二统计结果 中根据所述第一分类方式,获取预设个不同包长类别的数据包的统 计结果,计算所述预设个不同包长类别中各个包长类别的数据包丟 包率,所述数据包丟包率为所述第一统计结果中和所述第二统计结 果中同一类包长的数据包的个数差值与所述第一统计结果中所述 同一类包长对应的数据包的个数的比值;
对比子单元,用于将所述预设个不同包长类别的数据包按照包 长类别进行升序排序,所述预设个不同包长类别的数据包丟包率中 丟包类型为误码丟包;若所述预设个不同包长类别的数据包丟包率 中按照升序排列的数据包丟包率的个数小于所述预设个数,则判定 当前网络的丟包类型为拥塞丟包。 在第三种可能的实现方式中, 结合第四方面, 第一种或第二种 可能的实现方式中的任意一种, 所述丟包识别单元包括:
拟合单元, 用于当判定当前网络的丟包类型为误码丟包时, 将 至少一个包长类别的数据包的平均包长与所述至少一个包长类别 的数据包丟包率进行线性拟合;
重判单元,用于当拟合得到的曲线的斜率为上升趋势或者所述 曲线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码 丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中所述拟合得到的曲线以所述包长类别为坐标空间 中的横坐标, 以所述数据包丟包率为纵坐标。
在第四种可能的实现方式中, 结合第四方面, 第一种或第二种 可能的实现方式中的任意一种, 所述第一网元设备还包括: 时延重判单元, 用于当判定当前网络的丟包类型为误码丟包 时, 获取当前接收至少一个数据包的时延及接收数据包的平均时 延, 当所述当前接收至少一个数据包的时延大于所述接收数据包的 平均时延时, 则判定当前网络同时存在误码丟包及拥塞丟包; 当所 述当前接收至少一个数据包的时延小于或等于所述接收数据包的 平均时延时, 判定当前网络存在误码丟包。 第五方面, 提供一种网络系统, 包括至少一个第一网元设备和 一个第二网元设备,其中所述第一网元设备和第二网元设备通过交 换网连接, 所述第一网元设备为上述的任一第一网元设备, 所述第 二网元设备为上述任一的第二网元设备。 本发明的实施例网络丟包类型识别方法、 设备及系统, 通过能 够通过网元设备间传输的数据包的包长及不同包长的数据包的丟 包率精确快速的判断出网络丟包类型。
附图说明 为了更清楚地说明本发明实施例或现有技术中的技术方案,下 面将对实施例或现有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 下面描述中的附图仅仅是本发明的一些实施例, 对于 本领域普通技术人员来讲, 在不付出创造性劳动的前提下, 还可以 根据这些附图获得其他的附图。
图 1 为本发明的实施例提供的一种网络丟包类型识别方法流 程示意图; 图 2 为本发明的实施例提供的另一种网络丟包类型识别方法 流程示意图;
图 3 为本发明的另一实施例提供的一种网络丟包类型识别方 法流程示意图; 图 4 为本发明的又一实施例提供的一种网络丟包类型识别方 法流程示意图; 图 5 为本发明的实施例提供的一种数据包丟包率与包长类别 的线性拟合示意图; 图 6为本发明的实施例提供的一种第一网元设备结构示意图; 图 7 为本发明的实施例提供的另一种第一网元设备结构示意 图;
图 8 为本发明的实施例提供的又一种第一网元设备结构示意 图;
图 9 为本发明的实施例提供的再一种第一网元设备结构示意 图;
图 10为本发明的另一实施例提供的一种第二网元设备结构示 意图;
图 1 1 为本发明的另一实施例提供的另一种第二网元设备结构 示意图;
图 12为本发明的另一实施例提供的又一种第一网元设备结构 示意图;
图 13 为本发明的另一实施例提供的再一种第一网元设备结构 示意图;
图 14为本发明的又一实施例提供的一种第一网元设备结构示 意图;
图 15为本发明的再一实施例提供的一种第二网元设备结构示 意图;
图 16为本发明的实施例提供的一种网络系统结构示意图。 具体实施方式 下面将结合本发明实施例中的附图,对本发明实施例中的技术 方案进行清楚、 完整地描述, 显然, 所描述的实施例仅仅是本发明 一部分实施例, 而不是全部的实施例。 基于本发明中的实施例, 本 领域普通技术人员在没有作出创造性劳动前提下所获得的所有其 他实施例, 都属于本发明保护的范围。 在现有的各种通讯系统网构中,两个网元设备通过中间网络传 输数据包是当前信息通信的基本需求, 中间网络的基本功能就是充 分利用 自 己的带宽将两个设备的数据包尽可能多的且准确可靠的 传送。 但是由于各种原因, 数据包在中间网络传输时会出现丟包, 从而影响两个设备之前的通信,为了准确识别网络丟包的类型以便 于对当前网络传输的实时控制和调节,本发明提供一种网络丟包类 型识别方法, 在作为数据包发送一侧的第一网元设备侧, 参照图 1 所示, 包括:
101、 第一网元设备向第二网元设备发送数据包。
102、 第一网元设备将发送至第二网元设备的数据包按第一分 类方式进行分类。
第一分类方式为按数据包的包长或包长的预设范围分类, 当然 这里以具体的包长值分类时第一分类方式中的包长类别即具体的 包长值, 而以包长的预设范围分类时, 可以以每个预设包长范围内 的数据包的平均包长值作为包长类别,具体的以包长的预设范围分 类的具体形式可以为按照等差数列将数据包按照包长排序,以预设 的比特数作为公差值, 以数据包的最大或最小包长为包长初始值, 如以数据包的最大包长为包长初始值, 公差值设定为 n比特, 则小 于该最大包长在 n 比特范围内的所有包长的数据包设定为同一类 数据包,据此按照数据包的包长减小方向依次将每一个公差 n比特 范围内的数据包分别设定为一类,此时将各个公差范围内数据包的 平均包长值作为包长类别。
103、 第一网元设备统计发送至第二网元设备的数据包中各类 包长的数据包数量, 生成第一统计结果。
104、 第一网元设备将第一统计结果发送至第二网元设备。 以便第二网元设备根据第一统计结果及第二统计结果判断第 一网元设备与第二网元设备间的网络丟包的类型。 或者,
105、 第一网元设备接收第二网元设备发送的第二统计结果, 第一网元设备根据第一统计结果和第二统计结果判断第一网元设 备与第二网元设备间的网络丟包的类型。 其中,第二统计结果为第二网元设备将接收到的数据包按第一 分类方式分类后,对接收到的数据包中各类包长的数据包数量的统 计结果。
本发明提供一种网络丟包类型识别方法,在作为数据包接收一 侧的第二网元设备侧, 参照图 2所示, 包括:
201、 第二网元设备接收第一网元设备发送的数据包。
202、 第二网元设备将接收到的数据包按照第一分类方式进行 分类。 其中, 第一分类方式为按数据包的包长或包长的预设范围分 类, 当然这里以具体的包长值分类时第一分类方式中的包长类别即 具体的包长值, 而以包长的预设范围分类时, 可以以每个预设包长 范围内的数据包的平均包长值作为包长类别,具体的以包长的预设 范围分类的具体形式可以为按照等差数列将数据包按照包长排序, 以预设的比特数作为公差值,以数据包的最大或最小包长为包长初 始值,如以数据包的最大包长为包长初始值,公差值设定为 n比特, 则小于该最大包长在 n 比特范围内的所有包长的数据包设定为同 一类数据包, 据此按照数据包的包长减小方向依次将每一个公差 n 比特范围内的数据包分别设定为一类,此时将各个公差范围内数据 包的平均包长值作为包长类别。
203、 第二网元设备统计接收到的数据包中各类包长的数据包 数量, 生成第二统计结果。
204、 第二网元设备将第二统计结果发送至第一网元设备。 以便第一网元设备根据第一统计结果及第二统计结果判断第 一网元设备与第二网元设备间的网络丟包的类型。 或者,
205、 第二网元设备接收第一网元设备发送的第一统计结果, 第二网元设备根据第一统计结果和第二统计结果判断第一网元设 备与第二网元设备间的网络丟包的类型。 其中,第一统计结果为第一网元设备将发送至第二网元设备的 数据包按第一分类方式分类后,对发送至第二网元设备的数据包中 各类包长的数据包数量的统计结果。 本发明的实施例网络丟包类型识别方法、 设备及系统, 通过能 够通过网元设备间传输的数据包的包长及不同包长的数据包的丟 包率精确快速的判断出网络丟包类型。
具体的, 以下结合具体的实施例进行说明。
参照图 3所示, 可在图 1和图 2所示的实施例的基础上, 本发 明的实施例提供一种网络丟包类型识别方法,提供了网元设备根据 第一统计结果和第二统计结果对网元设备间丟包类型判别的具体 方法, 具体包括以下步骤:
301、 第一网元设备向第二网元设备发送数据包;
302、 第一网元设备将发送至第二网元设备的数据包按第一分 类方式进行分类。
第一分类方式为按数据包的包长或包长的预设范围分类。
303、 第一网元设备统计发送至第二网元设备的数据包中各类 包长的数据包数量, 生成第一统计结果;
304、 第二网元设备接收第一网元设备发送的数据包。 305、 第二网元设备将接收到的数据包按照第一分类方式进行 分类。
第一分类方式为按数据包的包长或包长的预设范围分类, 当然 这里以具体的包长值分类时第一分类方式中的包长类别即具体的 包长值, 而以包长的预设范围分类时, 可以以每个预设包长范围内 的数据包的平均包长值作为包长类别,具体的以包长的预设范围分 类的具体形式可以为按照等差数列将数据包按照包长排序,以预设 的比特数作为公差值, 以数据包的最大或最小包长为包长初始值, 如以数据包的最大包长为包长初始值, 公差值设定为 n比特, 则小 于该最大包长在 n 比特范围内的所有包长的数据包设定为同一类 数据包,据此按照数据包的包长减小方向依次将每一个公差 n比特 范围内的数据包分别设定为一类,此时将各个公差范围内数据包的 平均包长值作为包长类别。
306、 第二网元设备统计接收到的数据包中各类包长的数据包 数量, 生成第二统计结果。
307、 第二网元设备将第二统计结果发送至第一网元设备, 以 便第一网元设备根据第一统计结果及第二统计结果判断第一网元 设备与第二网元设备间的网络丟包的类型。
308、 第一网元设备接收第二网元设备发送的第二统计结果, 在第一统计结果中和第二统计结果中至少一次获取第一类包长的 数据包和第二类包长的数据包的统计结果,第一类包长的数据包的 平均包长大于第二类包长的数据包的平均包长。
这里根据步骤 302 中所釆用的具体的分类方式,若以具体的包 长值分类则该平均包长即对应分类中包长的值,若分类以多个预设 的包长范围分类,则此处的平均包长为该预设的包长范围的包长平 均值。
假设在第一统计结果和第二统计结果中提取获取对第一类包 长的数据包 M和第二类包长 N的数据包的统计结果, 且 M所指示 的包长时大于 N所指示的包长。
309、 第一网元设备计算第一类包长的数据包丟包率及第二类 包长的数据包丟包率。 数据包丟包率为第一统计结果中和第二统计结果中同一类包 长的数据包的个数差值与第一统计结果中同一类包长的数据包的 个数的比值。 具体的, 根据第一统计结果中第一类包长的数据包 M 个数和 第二统计结果中第一类包长的数据包 M 个数获取第一类包长的数 据包 M 的丟包率 X; 根据第一统计结果中第二类包长的数据包 N 个数和第二统计结果中第二类包长的数据包 N 个数获取第二类包 长的数据包 N的丟包率 X。
310、 当第一类包长的数据包丟包率大于第二类包长的数据包 丟包率的次数大于预设的次数时,判定当前网络的丟包类型为误码 丟包; 当第一类包长的数据包丟包率小于或等于第二类包长的数据 包丟包率的次数大于预设的次数时,判定当前网络的丟包类型为拥 塞丟包。
在网元设备间的通讯过程中, 若为拥塞丟包时, 数据包丟包率 与数据包的包长无关, 各类包长的数据包的丟包率是均衡的, 而误 码引起的丟包的情况下, 包长越长数据包丟包率是越大的, 此外至 少一次获取了第一类包长的数据包和第二类包长的数据包的统计 结果, 当只获取一次第一类包长的数据包和第二类包长的数据包的 统计结果时, 只能表征一个比较短的时期内, 网络的丟包类型当 X 大于 Y时判定当前网络的丟包类型为误码丟包, 若 X 小于或等于 Y则判定当前网络的丟包类型为拥塞丟包; 然而由于在数据传输过 程中第一统计结果和第二统计结果在每个阶段是动态变化的, 因此 也可以 K 次获取第一类包长的数据包和第二类包长的数据包的统 计结果, 根据每次的釆样结果, 当第一类包长 M 的数据包丟包率 大于第二类包长 N 的数据包丟包率的次数大于预设的次数时, 判 定当前网络的丟包类型为误码丟包; 否则, 判定当前网络的丟包类 型为拥塞丟包。
当然, 步骤 307也可以替换为: 第一网元设备将第一统计结果 发送至第二网元设备,这样第二网元设备在接收到该第一统计结果 后亦可以根据第一统计结果和第二统计结果执行步骤 308~3 10 中 第一网元设备所执行的方法, 附图 3 中虚线部分示出, 具体不再赘 述。
本发明的实施例提供的网络丟包类型识别方法,能够根据发送 设备发送的数据包和接收设备接收到的数据包的包长及丟包率的 动态变化, 精确快速的判断出网络丟包类型。
参照图 4所示, 可在图 1和图 2所示的实施例的基础上, 本发 明的实施例提供一种网络丟包类型识别方法,提供了网元设备根据 第一统计结果和第二统计结果对网元设备间丟包类型判别的具体 方法, 具体包括以下步骤:
401、 第一网元设备向第二网元设备发送数据包;
402、 第一网元设备将发送至第二网元设备的数据包按第一分 类方式进行分类。 第一分类方式为按数据包的包长或包长的预设范围分类, 当然 这里以具体的包长值分类时第一分类方式中的包长类别即具体的 包长值, 而以包长的预设范围分类时, 可以以每个预设包长范围内 的数据包的平均包长值作为包长类别,具体的以包长的预设范围分 类的具体形式可以为按照等差数列将数据包按照包长排序,以预设 的比特数作为公差值, 以数据包的最大或最小包长为包长初始值, 如以数据包的最大包长为包长初始值, 公差值设定为 n比特, 则小 于该最大包长在 n 比特范围内的所有包长的数据包设定为同一类 数据包,据此按照数据包的包长减小方向依次将每一个公差 n比特 范围内的数据包分别设定为一类,此时将各个公差范围内数据包的 平均包长值作为包长类别。
403、 第一网元设备统计发送至第二网元设备的数据包中各类 包长的数据包数量, 生成第一统计结果;
404、 第二网元设备接收第一网元设备发送的数据包。
405、 第二网元设备将接收到的数据包按照第一分类方式进行 分类。 第一分类方式为按数据包的包长或包长的预设范围分类。
406、 第二网元设备统计接收到的数据包中各类包长的数据包 数量, 生成第二统计结果。
407、 第二网元设备将第二统计结果发送至第一网元设备, 以 便第一网元设备根据第一统计结果及第二统计结果判断第一网元 设备与第二网元设备间的网络丟包的类型。
408、 第一网元设备接收第二网元设备发送的第二统计结果, 在第一统计结果中和第二统计结果中根据所述第一分类方式,获取 预设个不同包长类别的数据包的统计结果,计算预设个不同包长类 别中各个包长类别的数据包丟包率。
数据包丟包率为第一统计结果中和第二统计结果中同一类包 长的数据包的个数差值与第一统计结果中同一类包长对应的数据 包的个数的比值; 这里可取 K 个不同包长类别的数据包的统计结 果, 并根据统计结果计算 K 个不同包长类别的数据包丟包率, 得 到对应的 K个数据包丟包率。
409、 第一网元设备将预设个不同包长类别按照包长进行升序 排序,预设个不同包长类别的数据包丟包率中至少包括预设个数的 数据包丟包率为升序排列, 则判定当前网络的丟包类型为误码丟 包;若预设个不同包长类别的数据包丟包率中按照升序排列的数据 包丟包率的个数小于预设个数,则判定当前网络的丟包类型为拥塞 丟包。
这里将统计结果中 K 个包长类别进行排序, 然后检测数据包 的包长类别对应的数据包丟包率的排序情况, 当 K 个丟包率中至 少有预设个数的数据包丟包率为升序排列时判定当前网络的丟包 类型为误码丟包, 否则判定当前网络的丟包类型为拥塞丟包。 同理 在网元设备间的通讯过程中, 若为拥塞丟包时, 丟包率与数据包的 包长无关, 各类包长的数据包丟包率是均衡的, 而误码引起的丟包 的情况下, 包长越长丟包率是越大的, 因此根据排序的结果按照升 序排列的数据包丟包率的个数超过预设个数( 当然这里的预设个数 是小于或等于 K 的) 后越多则造成当前网络丟包的原因为误码丟 包的概率越大。 当然, 步骤 407也可以替换为: 第一网元设备将第一统计结果 发送至第二网元设备,这样第二网元设备在接收到该第一统计结果 后亦可以根据第一统计结果和第二统计结果执行步骤 408~409 中 第一网元设备所执行的方法, 附图 4 中虚线部分示出, 具体不再赘 述。 本发明的实施例提供的网络丟包类型识别方法,能够根据发送 设备发送的数据包和接收设备接收到的数据包的包长及丟包率的 动态变化, 精确快速的判断出网络丟包类型。 进一步, 可选的, 根据上述的实施例判定当前网络的丟包类型 为误码丟包后, 该方案还包括: al、第一网元设备将至少一个包长类别的数据包的包长与该至 少一个包长类别的数据包的丟包率进行线性拟合;
b l、当拟合得到的曲线的斜率为上升趋势或者曲线的斜率为大 于预设斜率的固定值时, 则判定当前网络存在误码丟包; c l、 当拟合得到的曲线中所有包长类别对应的数据包丟包率均 大于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟 包及拥塞丟包。 其中拟合得到的曲线以包长类别为坐标空间中的横坐标,以数 据包丟包率为纵坐标。
具体的参照图 5所示,将数据包丟包率与包长类别的关系进行 线性拟合, 拟合方法不限, 其中在包长为 0时, 数据包丟包率为 J。 此时该通过拟合获得的曲线 1的斜率值为上升趋势时, 则判定当前 网络只存在误码丟包; 或者拟合得到的曲线为固定斜率 K1 , 即拟 合的到的为斜率为 K1 的直线, 当 K1 大于预设斜率 K2时, 则判定 当前网络只存在误码丟包; 如图 5 所示, 则在以 i , j 分别为横坐 标( 包长类别 )和纵坐标(数据包丟包率)的坐标系中, 坐标点( i , j ) = ( 0 , J ) 为该拟合曲线 1与纵坐标轴 i=0 的交点, 当拟合得到 的曲线 1中所有包长类别对应的数据包丟包率均大于预设的数据包 丟包率阔值时, 则判定当前网络同时存在误码丟包及拥塞丟包; 当 线性拟合得到的为曲线时, 由于斜率为上升趋势, 因此只要坐标点 ( 0 , J ) 处的 J大于预设的数据包丟包率阔值, 则判定当前网络同 时存在误码丟包及拥塞丟包。
当然根据上述的实施例当判定当前网络的丟包类型为误码丟 包为第二网元设备做出时, 步骤 a l、 b l、 c l 也可由第二网元设备 执行, 此处不再赘述。 可选的,根据上述的实施例判定当前网络的丟包类型为误码丟 包后, 该方案还包括:
a2、第二网元设备获取第二网元设备接收至少一个数据包的时 延及第二网元设备接收数据包的平均时延, 当接收至少一个数据包 的时延大于接收数据包的平均时延时,则判定当前网络同时存在误 码丟包及拥塞丟包; 当当前接收至少一个数据包的时延小于或等于 接收数据包的平均时延时, 判定当前网络存在误码丟包。 此时将一个时间段内第二网元设备接收至少一个数据包的时 延与全局的平均时延进行比较, 当该一个时间段内接收至少一个数 据包的时延大于全局的平均时延时,说明网络存在拥塞丟包的可能 性。
当然步骤 a2 为第一网元设备执行此时需要从第二网元设备侧 获取相关的时延, 当然根据上述的实施例当判定当前网络的丟包类 型为误码丟包为第二网元设备做出时, 该步骤 a2 也可由第二网元 设备直接执行。 由于拥塞丟包主要是网络状况因素引起,因此当前网络的丟包 因素若包括拥塞丟包时,则接收端接收数据包的时延会增加,因此, 当所述接收至少一个数据包的时延大于所述接收数据包的平均时 延时, 则判定当前网络同时存在误码丟包及拥塞丟包; 否则判定当 前网络存在误码丟包。 本发明的实施例提供一种第一网元设备,该第一网元设备具体 可以为网络构架中的任一节点网元设备, 如交换设备、 射频拉远单 元、 基站、 移动终端或者基站侧用于数据包收发的单元、 模块或实 体等在网络构架中任一能够实现数据包收发的设备,在本发明的实 施例中对网元设备的具体形式不做具体限制, 以可以实现本发明的 实施例所提供的上述任一网络丟包类型识别方法为准,具体的参照 图 6所示, 该网元设备包括: 数据发送单元 61 , 用于向第二网元设备发送数据包; 数据分类单元 62 , 用于将数据发送单元 61发送至第二网元设 备的数据包按第一分类方式进行分类,第一分类方式为按数据包的 包长或包长的预设范围分类; 统计单元 63 , 用于统计数据分类单元 62分类后的数据包中各 类包长的数据包数量, 生成第一统计结果;
所述数据发送单元 61 , 还用于将统计单元 63转发的第一统计 结果发送至第二网元设备,以便第二网元设备根据第一统计结果及 第二统计结果判断第一网元设备与第二网元设备间的网络丟包的 类型; 或者,
丟包识别单元 64 , 用于接收第二网元设备发送的第二统计结 果, 根据统计单元 63 转发的第一统计结果和第二统计结果判断第 一网元设备与第二网元设备间的网络丟包的类型; 其中,第二统计结果为第二网元设备将接收到的数据包按第一 分类方式分类后,对接收到的数据包中各类包长的数据包数量的统 计结果。
可选的, 参照图 7所示, 丟包识别单元 64 , 包括:
釆样子单元 641 , 用于用在第一统计结果中和第二统计结果中 至少一次获取第一类包长的数据包和第二类包长的数据包的统计 结果,第一类包长的数据包的平均包长大于第二类包长的数据包的 平均包长;
处理子单元 642 , 用于计算第一类包长的数据包丟包率及第二 类包长的数据包丟包率,数据包丟包率为所述第一统计结果中和所 述第二统计结果中同一类包长的数据包的个数差值与第一统计结 果中同一类包长的数据包的个数的比值; 判别子单元 643 , 用于当第一类包长的数据包丟包率大于第二 类包长的数据包丟包率的次数大于预设的次数时,判定当前网络的 丟包类型为误码丟包; 当第一类包长的数据包丟包率小于或等于第 二类包长的数据包丟包率的次数大于预设的次数时,判定当前网络 的丟包类型为拥塞丟包。
可选的, 参照图 8所示, 丟包识别单元 64 , 包括:
釆样子单元 644 , 用于在第一统计结果中和第二统计结果中根 据第一分类方式, 获取预设个不同包长类别的数据包的统计结果, 计算预设个不同包长类别中各个包长类别的数据包丟包率,数据包 丟包率为第一统计结果中和所述第二统计结果中同一类包长的数 据包的个数差值与第一统计结果中同一类包长对应的数据包的个 数的比值; 对比子单元 645 , 用于将预设个不同包长类别的数据包按照包 长类别进行升序排序,预设个不同包长类别的数据包丟包率中至少 包括预设个数的数据包丟包率为升序排列,则判定当前网络的丟包 类型为误码丟包;若预设个不同包长类别的数据包丟包率中按照升 序排列的数据包丟包率的个数小于所述预设个数,则判定当前网络 的丟包类型为拥塞丟包。
进一步, 可选的参照图 9所示, 当第一网元设备 6 包括丟包识 别单元 64时, 第一网元设备 6还包括: 拟合单元 65 , 用于当判定当前网络的丟包类型为误码丟包时, 将至少一个包长类别的数据包的平均包长与至少一个包长类别的 数据包丟包率进行线性拟合;
重判单元 66 , 用于当拟合得到的曲线的斜率为上升趋势或者 曲线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码 丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中拟合得到的曲线以包长类别为坐标空间中的横坐 标, 以数据包丟包率为纵坐标。
可选的, 参照图 9所示, 当第一网元设备 6 包括丟包识别单元 64时, 第一网元设备 6还包括: 时延重判单元 67 , 用于当判定当前网络的丟包类型为误码丟 包时,获取第二网元设备当前接收至少一个数据包的时延及第二网 元设备接收数据包的平均时延, 当当前接收至少一个数据包的时延 大于接收数据包的平均时延时,则判定当前网络同时存在误码丟包 及拥塞丟包; 当当前接收至少一个数据包的时延小于或等于接收数 据包的平均时延时, 判定当前网络存在误码丟包。 本发明的实施例提供的第一网元设备,能够根据发送的数据包 和接收设备接收到的数据包的包长及丟包率的动态变化,精确快速 的判断出网络丟包类型。
本发明的实施例提供一种第二网元设备,该第二网元设备具体 可以为网络构架中的任一节点网元设备, 如交换设备、 射频拉远单 元、 基站、 移动终端或者基站侧用于数据包收发的单元、 模块或实 体等在网络构架中任一能够实现数据包收发的设备,在本发明的实 施例中对网元设备的具体形式不做具体限制,以可以实现本发明的 实施例所提供的上述任一网络丟包类型识别方法为准,具体的参照 图 10所示, 该网元设备包括: 数据接收单元 71 , 用于接收第一网元设备发送的数据包; 数据分类单元 72 , 用于将数据接收单元 71接收到的数据包按 照第一分类方式进行分类,第一分类方式为按数据包的包长或包长 的预设范围分类;
统计单元 73 , 用于统计数据分类单元 72分类后的接收到的数 据包中各类包长的数据包数量, 生成第二统计结果;
数据发送单元 74 , 用于将统计单元 73转发的第二统计结果发 送至第一网元设备,以便第一网元设备根据第一统计结果及第二统 计结果判断第一网元设备与第二网元设备间的网络丟包的类型; 或者,
丟包识别单元 75 , 用于接收第一网元设备发送的第一统计结 果, 根据统计单元 73 转发的第一统计结果和第二统计结果判断第 一网元设备与第二网元设备间的网络丟包的类型 其中,第一统计结果为第一网元设备将发送至第二网元设备的 数据包按第一分类方式分类后,对发送至第二网元设备的数据包中 各类包长的数据包数量的统计结果。
可选的, 参照图 1 1 所示, 丟包识别单元 75 包括:
釆样子单元 75 1 , 用于在第一统计结果中和第二统计结果中至 少一次获取第一类包长的数据包和第二类包长的数据包的统计结 果,第一类包长的数据包的平均包长大于第二类包长的数据包的平 均包长;
处理子单元 752 , 用于计算第一类包长的数据包丟包率及第二 类包长的数据包丟包率,数据包丟包率为第一统计结果中和第二统 计结果中同一类包长的数据包的个数差值与第一统计结果中所述 同一类包长的数据包的个数的比值;
判别子单元 753 , 用于当第一类包长的数据包丟包率大于第二 类包长的数据包丟包率的次数大于预设的次数时,判定当前网络的 丟包类型为误码丟包; 当第一类包长的数据包丟包率小于或等于第 二类包长的数据包丟包率的次数大于预设的次数时,判定当前网络 的丟包类型为拥塞丟包。
可选的, 参照图 1 2所示, 丟包识别单元 75 包括:
釆样子单元 754 , 用于在第一统计结果中和第二统计结果中根 据第一分类方式, 获取预设个不同包长类别的数据包的统计结果, 计算预设个不同包长类别中各个包长类别的数据包丟包率,数据包 丟包率为第一统计结果中和所述第二统计结果中同一类包长的数 据包的个数差值与第一统计结果中所述同一类包长对应的数据包 的个数的比值;
对比子单元 755 , 用于将预设个不同包长类别的数据包按照包 长类别进行升序排序,预设个不同包长类别的数据包丟包率中至少 包括预设个数的数据包丟包率为升序排列,则判定当前网络的丟包 类型为误码丟包;若预设个不同包长类别的数据包丟包率中按照升 序排列的数据包丟包率的个数小于所述预设个数,则判定当前网络 的丟包类型为拥塞丟包。
进一步可选的, 参照图 13所示, 第二网元设备还包括: 拟合单元 76 , 用于当判定当前网络的丟包类型为误码丟包时, 将至少一个包长类别的数据包的平均包长与至少一个包长类别的 数据包丟包率进行线性拟合;
重判单元 77 , 用于当拟合得到的曲线的斜率为上升趋势或者 曲线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码 丟包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大 于预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包 及拥塞丟包;其中拟合得到的曲线以包长类别为坐标空间中的横坐 标, 以数据包丟包率为纵坐标。
可选的, 参照图 13所示, 第二网元设备还包括: 还包括: 时延重判单元 78 , 用于当判定当前网络的丟包类型为误码丟 包时,获取第二网元设备当前接收至少一个数据包的时延及第二网 元设备接收数据包的平均时延, 当当前接收至少一个数据包的时延 大于接收数据包的平均时延时,则判定当前网络同时存在误码丟包 及拥塞丟包; 当当前接收至少一个数据包的时延小于或等于接收数 据包的平均时延时, 判定当前网络存在误码丟包。 本发明的实施例提供的第二网元设备,能够根据发送设备发送 的数据包和接收到的数据包的包长及丟包率的动态变化,精确快速 的判断出网络丟包类型。 本发明的实施例提供一种第一网元设备,该第一网元设备具体 可以为网络构架中的任一节点网元设备, 如交换设备、 射频拉远单 元、 基站、 移动终端或者基站侧用于数据包收发的单元、 模块或实 体等在网络构架中任一能够实现数据包收发的设备,在本发明的实 施例中对网元设备的具体形式不做具体限制, 以可以实现本发明的 实施例所提供的上述任一网络丟包类型识别方法为准,具体的参照 图 14所示, 该网元设备包括: 包括: 至少一个处理器 81、 存储器 82、 通信端口 83和总线 84 , 该至少一个处理器 81、 存储器 82和 通信接口 83通过总线 84连接并完成相互间的通信。 该总线 84 可以是工业标准体系 结构 ( Industry Standard Architecture , 简称为 ISA ) 总线、 外部设备互连 ( Peripheral Component , 简称为 PCI )总线或扩展工业标准体系结构( Extended Industry Standard Architecture , 简称为 EISA ) 总线等。 该总线 84 可以分为地址总线、 数据总线、 控制总线等。 为便于表示, 图 14 中仅用一条粗线表示, 但并不表示仅有一根总线或一种类型的总 线。 其中:
存储器 82用于存储可执行程序代码, 该程序代码包括计算机 操作指令。 存储器 82 可能包含高速 RAM存储器, 也可能还包括 非易失性存储器( non-volatile memory ) ,例如至少一个磁盘存储器。 所述存储设备中存储: 操作系统、 应用程序, 用于实现本发明实施 例的程序代码。所述操作系统用于控制和实现所述处理单元执行的 处理功能。 所述应用程序包含程序代码, 如字处理软件、 email软 件。
处理器 81可能是一个中央处理器 ( Central Processing Unit , 简称为 CPU ),或者是特定集成电路( Application Specific Integrated Circuit , 简称为 ASIC ) , 或者是被配置成实施本发明实施例的一个 或多个集成电路。
通信接口 83 , 主要用于实现本实施例中的装置之间的通信。 该处理器 81 , 用于通过通信端口 83向第二网元设备发送数据 包;
处理器 81通过通信端口 83发送至第二网元设备的数据包按第 一分类方式进行分类,第一分类方式为按数据包的包长或包长的预 设范围分类;
处理器 81 , 用于统计分类后的数据包中各类包长的数据包数 量, 生成第一统计结果;
处理器 81 , 用于通过通信端口 83将第一统计结果发送至第二 网元设备,以便第二网元设备根据第一统计结果及第二统计结果判 断第一网元设备与第二网元设备间的网络丟包的类型; 或者, 处理器 81 , 用于通过通信端口 83接收第二网元设备发送的第 二统计结果,根据第一统计结果和第二统计结果判断第一网元设备 与第二网元设备间的网络丟包的类型;
其中,第二统计结果为第二网元设备将接收到的数据包按第一 分类方式分类后,对接收到的数据包中各类包长的数据包数量的统 计结果。 可选的,
处理器 81 , 具体在第一统计结果中和第二统计结果中至少一 次获取第一类包长的数据包和第二类包长的数据包的统计结果,第 一类包长的数据包的平均包长大于第二类包长的数据包的平均包 长;
处理器 81 , 用于计算第一类包长的数据包丟包率及第二类包 长的数据包丟包率,数据包丟包率为第一统计结果中和第二统计结 果中同一类包长的数据包的个数差值与第一统计结果中所述同一 类包长的数据包的个数的比值;
处理器 81 , 用于当第一类包长的数据包丟包率大于第二类包 长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟包 类型为误码丟包; 当第一类包长的数据包丟包率小于或等于第二类 包长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟 包类型为拥塞丟包。
可选的, 处理器 81 , 用于在第一统计结果中和第二统计结果 中根据第一分类方式,获取预设个不同包长类别的数据包的统计结 果, 计算预设个不同包长类别中各个包长类别的数据包丟包率, 数 据包丟包率为第一统计结果中和第二统计结果中同一类包长的数 据包的个数差值与第一统计结果中同一类包长对应的数据包的个 数的比值; 处理器 81 , 用于将预设个不同包长类别的数据包按照包长类 别进行升序排序,预设个不同包长类别的数据包丟包率中至少包括 预设个数的数据包丟包率为升序排列,则判定当前网络的丟包类型 为误码丟包;若预设个不同包长类别的数据包丟包率中按照升序排 列的数据包丟包率的个数小于预设个数,则判定当前网络的丟包类 型为拥塞丟包。
进一步, 可选的, 处理器 81 , 还用于当判定当前网络的丟包 类型为误码丟包时,将至少一个包长类别的数据包的平均包长与至 少一个包长类别的数据包丟包率进行线性拟合;
处理器 81 , 用于当拟合得到的曲线的斜率为上升趋势或者曲 线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码丟 包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于 预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包及 拥塞丟包;其中拟合得到的曲线以所述包长类别为坐标空间中的横 坐标, 以所述数据包丟包率为纵坐标 可选的,
处理器 81 , 用于当判定当前网络的丟包类型为误码丟包时, 通过通信接口 83 获取第二网元设备当前接收至少一个数据包的时 延及第二网元设备接收数据包的平均时延, 当当前接收至少一个数 据包的时延大于接收数据包的平均时延时,则判定当前网络同时存 在误码丟包及拥塞丟包; 当当前接收至少一个数据包的时延小于或 等于接收数据包的平均时延时, 判定当前网络存在误码丟包。 本发明的实施例提供的第一网元设备,能够根据发送的数据包 和接收设备接收到的数据包的包长及丟包率的动态变化,精确快速 的判断出网络丟包类型。 本发明的实施例提供一种第二网元设备,该网元设备具体可以 为网络构架中的任一节点网元设备, 如交换设备、 射频拉远单元、 基站、 移动终端或者基站侧用于数据包收发的单元、 模块或实体等 在网络构架中任一能够实现数据包收发的设备,在本发明的实施例 中对网元设备的具体形式不做具体限制, 以可以实现本发明的实施 例所提供的上述任一网络丟包类型识别方法为准, 具体的参照图
15所示, 该网元设备包括: 包括: 至少一个处理器 91、 存储器 92、 通信端口 93和总线 94 , 该至少一个处理器 91、 存储器 92和通信 接口 93通过总线 94连接并完成相互间的通信。
该总线 94 可以是工业标准体系 结构 ( Industry Standard Architecture , 简称为 ISA ) 总线、 外部设备互连 ( Peripheral Component , 简称为 PCI )总线或扩展工业标准体系结构( Extended Industry Standard Architecture , 简称为 EISA ) 总线等。 该总线 94 可以分为地址总线、 数据总线、 控制总线等。 为便于表示, 图 1 5 中仅用一条粗线表示, 但并不表示仅有一根总线或一种类型的总 线。 其中:
存储器 92用于存储可执行程序代码, 该程序代码包括计算机 操作指令。 存储器 92 可能包含高速 RAM存储器, 也可能还包括 非易失性存储器( non-volatile memory ) ,例如至少一个磁盘存储器。 所述存储设备中存储: 操作系统、 应用程序, 用于实现本发明实施 例的程序代码。所述操作系统用于控制和实现所述处理单元执行的 处理功能。 所述应用程序包含程序代码, 如字处理软件、 email软 件。
处理器 91可能是一个中央处理器 ( Central Processing Unit , 简称为 CPU ),或者是特定集成电路( Application Specific Integrated Circuit , 简称为 ASIC ) , 或者是被配置成实施本发明实施例的一个 或多个集成电路。
通信接口 93 , 主要用于实现本实施例中的装置之间的通信。 处理器 91 , 用于通过通信端口 93接收第一网元设备发送的数 据包;
处理器 91 , 用于将通过通信端口 93接收到的数据包按照第一 分类方式进行分类,第一分类方式为按数据包的包长或包长的预设 范围分类;
处理器 91 , 用于分类后的接收到的数据包中各类包长的数据 包数量, 生成第二统计结果;
处理器 91 , 用于将第二统计结果通过通信端口 93发送至第一 网元设备,以便第一网元设备根据第一统计结果及第二统计结果判 断第一网元设备与第二网元设备间的网络丟包的类型; 或者,
处理器 91 , 用于通过通信端口 93接收第一网元设备发送的第 一统计结果,根据第一统计结果和第二统计结果判断第一网元设备 与第二网元设备间的网络丟包的类型; 其中,第一统计结果为第一网元设备将发送至第二网元设备的 数据包按第一分类方式分类后,对发送至第二网元设备的数据包中 各类包长的数据包数量的统计结果。
可选的, 处理器 91 , 具体用于在第一统计结果中和第二统计 结果中至少一次获取第一类包长的数据包和第二类包长的数据包 的统计结果,第一类包长的数据包的平均包长大于所述第二类包长 的数据包的平均包长;
处理器 91 , 用于计算第一类包长的数据包丟包率及第二类包 长的数据包丟包率,数据包丟包率为第一统计结果中和第二统计结 果中同一类包长的数据包的个数差值与第一统计结果中同一类包 长的数据包的个数的比值; 处理器 91 , 用于当第一类包长的数据包丟包率大于第二类包 长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟包 类型为误码丟包; 当第一类包长的数据包丟包率小于或等于第二类 包长的数据包丟包率的次数大于预设的次数时,判定当前网络的丟 包类型为拥塞丟包。
可选的, 处理器 91 , 具体用于在第一统计结果中和第二统计 结果中根据第一分类方式,获取预设个不同包长类别的数据包的统 计结果, 计算预设个不同包长类别中各个包长类别的数据包丟包 率,数据包丟包率为第一统计结果中和第二统计结果中同一类包长 的数据包的个数差值与第一统计结果中同一类包长对应的数据包 的个数的比值;
处理器 91 , 用于将预设个不同包长类别的数据包按照包长类 别进行升序排序,预设个不同包长类别的数据包丟包率中至少包括 预设个数的数据包丟包率为升序排列,则判定当前网络的丟包类型 为误码丟包;若预设个不同包长类别的数据包丟包率中按照升序排 列的数据包丟包率的个数小于预设个数,则判定当前网络的丟包类 型为拥塞丟包。
进一步可选的, 处理器 91 , 用于当判定当前网络的丟包类型 为误码丟包时,将至少一个包长类别的数据包的平均包长与所述至 少一个包长类别的数据包丟包率进行线性拟合;
处理器 91 , 用于当拟合得到的曲线的斜率为上升趋势或者曲 线的斜率为大于预设斜率的固定值时,则判定当前网络存在误码丟 包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于 预设的数据包丟包率阔值时,则判定当前网络同时存在误码丟包及 拥塞丟包;其中拟合得到的曲线以所述包长类别为坐标空间中的横 坐标, 以数据包丟包率为纵坐标。
可选的, 处理器 91 , 用于当判定当前网络的丟包类型为误码 丟包时,获取当前接收至少一个数据包的时延及接收数据包的平均 时延, 当当前接收至少一个数据包的时延大于接收数据包的平均时 延时, 则判定当前网络同时存在误码丟包及拥塞丟包; 当当前接收 至少一个数据包的时延小于或等于接收数据包的平均时延时,判定 当前网络存在误码丟包。 本发明的实施例提供的第二网元设备,能够根据发送设备发送 的数据包和接收到的数据包的包长及丟包率的动态变化,精确快速 的判断出网络丟包类型。
参照图 16所示, 本发明的实施例提供的一种网络系统, 包括 至少一个第一网元设备 101和一个第二网元设备 102 , 其中第一网 元设备 101 和第二网元设备 102 通过交换网连接, 第一网元设备 101 为上述设备实施例提供的任一第一网元设备, 第二网元设备 102为上述设备实施例提供的任一第二网元设备。 当然本发明的实 施例所提供的网络系统可以应用本发明的实施例所提供的网元设 备实现本发明的实施例所提供的网络丟包类型识别方法, 其中, 该 系统所包含的各个网元设备的结构,及通过各个网元设备所实现的 网络丟包类型识别方法的具体过程参照以上关于方法及设备的实 施例的描述, 这里不再赘述。 本发明的实施例提供的网络系统,能够根据发送设备发送的数 据包和接收设备接收到的数据包的包长及丟包率的动态变化,精确 快速的判断出网络丟包类型。 本领域普通技术人员可以理解: 实现上述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 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。
以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围 并不局限于此,任何熟悉本技术领域的技术人员在本发明揭露的技 术范围内, 可轻易想到变化或替换, 都应涵盖在本发明的保护范围 之内。因此,本发明的保护范围应以所述权利要求的保护范围为准。

Claims

权 利 要 求 书
1、 一种网络丟包类型识别方法, 其特征在于, 包括:
第一网元设备向第二网元设备发送数据包;
所述第一网元设备将发送至所述第二网元设备的数据包按第一 分类方式进行分类, 所述第一分类方式为按所述数据包的包长或包长 的预设范围分类;
所述第一网元设备统计所述发送至所述第二网元设备的数据包 中各类包长的数据包数量, 生成第一统计结果;
所述第一网元设备将所述第一统计结果发送至所述第二网元设 备, 以便所述第二网元设备根据所述第一统计结果及第二统计结果判 断所述第一网元设备与所述第二网元设备间的网络丟包的类型; 或者,
所述第一网元设备接收所述第二网元设备发送的所述第二统计 结果, 所述第一网元设备根据所述第一统计结果和所述第二统计结果 判断所述第一网元设备与所述第二网元设备间的网络丟包的类型; 其中,所述第二统计结果为所述第二网元设备将接收到的数据包 按所述第一分类方式分类后, 对所述接收到的数据包中各类包长的数 据包数量的统计结果。
2、 根据权利要求 1 所述的方法, 其特征在于, 所述第一网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元设 备与所述第二网元设备间的网络丟包的类型包括:
在所述第一统计结果中和所述第二统计结果中至少一次获取第 一类包长的数据包和第二类包长的数据包的统计结果, 所述第一类包 长的数据包的平均包长大于所述第二类包长的数据包的平均包长; 计算所述第一类包长的数据包丟包率及所述第二类包长的数据 包丟包率, 所述数据包丟包率为所述第一统计结果中和所述第二统计 结果中同一类包长的数据包的个数差值与所述第一统计结果中所述 同一类包长的数据包的个数的比值;
当所述第一类包长的数据包丟包率大于所述第二类包长的数据 包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型为误码 丟包;
当所述第一类包长的数据包丟包率小于或等于所述第二类包长 的数据包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型 为拥塞丟包。
3、 根据权利要求 1 所述的方法, 其特征在于, 所述第一网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元设 备与所述第二网元设备间的网络丟包的类型包括:
在所述第一统计结果中和所述第二统计结果中根据所述第一分 类方式, 获取预设个不同包长类别的数据包的统计结果, 计算所述预 设个不同包长类别中各个包长类别的数据包丟包率, 所述数据包丟包 率为所述第一统计结果中和所述第二统计结果中同一类包长的数据 包的个数差值与所述第一统计结果中所述同一类包长对应的数据包 的个数的比值;
将所述预设个不同包长类别的数据包按照包长类别进行升序排 序, 所述预设个不同包长类别的数据包丟包率中至少包括预设个数的 数据包丟包率为升序排列, 则判定当前网络的丟包类型为误码丟包; 若所述预设个不同包长类别的数据包丟包率中按照升序排列的 数据包丟包率的个数小于所述预设个数, 则判定当前网络的丟包类型 为拥塞丟包。
4、 根据权利要求 1 ~3任一项所述的方法, 其特征在于, 所述方 法还包括:
当判定当前网络的丟包类型为误码丟包时,将至少一个包长类别 的数据包的平均包长与所述至少一个包长类别的数据包丟包率进行 线性拟合;
当拟合得到的曲线的斜率为上升趋势或者所述曲线的斜率为大 于预设斜率的固定值时, 则判定当前网络存在误码丟包;
当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于 预设的数据包丟包率阔值时, 则判定当前网络同时存在误码丟包及拥 塞丟包; 其中所述拟合得到的曲线以所述包长类别为坐标空间中的横 坐标, 以所述数据包丟包率为纵坐标。
5、 根据权利要求 1 ~3任一项所述的方法, 其特征在于, 所述方 法还包括:
当判定当前网络的丟包类型为误码丟包时,获取所述第二网元设 备当前接收至少一个数据包的时延及所述第二网元设备接收数据包 的平均时延, 当所述当前接收至少一个数据包的时延大于所述接收数 据包的平均时延时, 则判定当前网络同时存在误码丟包及拥塞丟包; 当所述当前接收至少一个数据包的时延小于或等于所述接收数 据包的平均时延时, 判定当前网络存在误码丟包。
6、 一种网络丟包类型识别方法, 其特征在于, 包括
第二网元设备接收第一网元设备发送的数据包;
将接收到的数据包按照第一分类方式进行分类,所述第一分类方 式为按所述数据包的包长或包长的预设范围分类;
所述第二网元设备统计所述接收到的数据包中各类包长的数据 包数量, 生成第二统计结果;
所述第二网元设备将所述第二统计结果发送至所述第一网元设 备, 以便所述第一网元设备根据所述第一统计结果及第二统计结果判 断所述第一网元设备与所述第二网元设备间的网络丟包的类型;
或者,
所述第二网元设备接收所述第一网元设备发送的所述第一统计 结果, 所述第二网元设备根据所述第一统计结果和所述第二统计结果 判断所述第一网元设备与所述第二网元设备间的网络丟包的类型; 其中,所述第一统计结果为所述第一网元设备将发送至所述第二 网元设备的数据包按所述第一分类方式分类后, 对所述发送至所述第 二网元设备的数据包中各类包长的数据包数量的统计结果。
7、 根据权利要求 6所述的方法, 其特征在于, 所述第二网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元设 备与所述第二网元设备间的网络丟包的类型包括: 在所述第一统计结果中和所述第二统计结果中至少一次获取第 一类包长的数据包和第二类包长的数据包的统计结果, 所述第一类包 长的数据包的平均包长大于所述第二类包长的数据包的平均包长; 计算所述第一类包长的数据包丟包率及所述第二类包长的数据 包丟包率, 所述数据包丟包率为所述第一统计结果中和所述第二统计 结果中同一类包长的数据包的个数差值与所述第一统计结果中所述 同一类包长的数据包的个数的比值;
当所述第一类包长的数据包丟包率大于所述第二类包长的数据 包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型为误码 丟包;
当所述第一类包长的数据包丟包率小于或等于所述第二类包长 的数据包丟包率的次数大于预设的次数时, 判定当前网络的丟包类型 为拥塞丟包。
8、 根据权利要求 6所述的方法, 其特征在于, 所述第二网元设 备根据所述第一统计结果和所述第二统计结果判断所述第一网元设 备与所述第二网元设备间的网络丟包的类型包括:
在所述第一统计结果中和所述第二统计结果中根据所述第一分 类方式, 获取预设个不同包长类别的数据包的统计结果, 计算所述预 设个不同包长类别中各个包长类别的数据包丟包率, 所述数据包丟包 率为所述第一统计结果中和所述第二统计结果中同一类包长的数据 包的个数差值与所述第一统计结果中所述同一类包长对应的数据包 的个数的比值;
将所述预设个不同包长类别的数据包按照包长类别进行升序排 序, 所述预设个不同包长类别的数据包丟包率中至少包括预设个数的 数据包丟包率为升序排列, 则判定当前网络的丟包类型为误码丟包; 若所述预设个不同包长类别的数据包丟包率中按照升序排列的 数据包丟包率的个数小于所述预设个数, 则判定当前网络的丟包类型 为拥塞丟包。
9、 根据权利要求 6~8任一项所述的方法, 其特征在于, 所述方 法还包括:
当判定当前网络的丟包类型为误码丟包时,将至少一个包长类别 的数据包的平均包长与所述至少一个包长类别的数据包丟包率进行 线性拟合;
当拟合得到的曲线的斜率为上升趋势或者所述曲线的斜率为大 于预设斜率的固定值时, 则判定当前网络存在误码丟包;
当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于 预设的数据包丟包率阔值时, 则判定当前网络同时存在误码丟包及拥 塞丟包; 其中所述拟合得到的曲线以所述包长类别为坐标空间中的横 坐标, 以所述数据包丟包率为纵坐标。
10、 根据权利要求 6~8任一项所述的方法, 其特征在于, 所述方 法还包括:
当判定当前网络的丟包类型为误码丟包时,获取当前接收至少一 个数据包的时延及接收数据包的平均时延, 当所述当前接收至少一个 数据包的时延大于所述接收数据包的平均时延时, 则判定当前网络同 时存在误码丟包及拥塞丟包;
当所述当前接收至少一个数据包的时延小于或等于所述接收数 据包的平均时延时, 判定当前网络存在误码丟包。
1 1、 一种第一网元设备, 其特征在于, 包括:
数据发送单元, 用于向第二网元设备发送数据包;
数据分类单元,用于将所述数据发送单元发送至所述第二网元设 备的数据包按第一分类方式进行分类, 所述第一分类方式为按所述数 据包的包长或包长的预设范围分类;
统计单元,用于统计所述数据分类单元分类后的所述数据包中各 类包长的数据包数量, 生成第一统计结果;
所述数据发送单元,还用于将所述统计单元转发的所述第一统计 结果发送至所述第二网元设备, 以便所述第二网元设备根据所述第一 统计结果及第二统计结果判断所述第一网元设备与所述第二网元设 备间的网络丟包的类型; 或者,
丟包识别单元,用于接收所述第二网元设备发送的所述第二统计 结果, 根据所述统计单元转发的所述第一统计结果和所述第二统计结 果判断所述第一网元设备与所述第二网元设备间的网络丟包的类型; 其中,所述第二统计结果为所述第二网元设备将接收到的数据包 按所述第一分类方式分类后, 对所述接收到的数据包中各类包长的数 据包数量的统计结果。
12、 根据权利要求 11 所述的设备, 其特征在于, 所述丟包识别 单元, 包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果中 至少一次获取第一类包长的数据包和第二类包长的数据包的统计结 果, 所述第一类包长的数据包的平均包长大于所述第二类包长的数据 包的平均包长;
处理子单元,用于计算所述第一类包长的数据包丟包率及所述第 二类包长的数据包丟包率, 所述数据包丟包率为所述第一统计结果中 和所述第二统计结果中同一类包长的数据包的个数差值与所述第一 统计结果中所述同一类包长的数据包的个数的比值;
判别子单元,用于当所述第一类包长的数据包丟包率大于所述第 二类包长的数据包丟包率的次数大于预设的次数时, 判定当前网络的 丟包类型为误码丟包; 当所述第一类包长的数据包丟包率小于或等于 所述第二类包长的数据包丟包率的次数大于预设的次数时, 判定当前 网络的丟包类型为拥塞丟包。
13、 根据权利要求 11 所述的设备, 其特征在于, 所述丟包识别 单元包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果中 根据所述第一分类方式, 获取预设个不同包长类别的数据包的统计结 果, 计算所述预设个不同包长类别中各个包长类别的数据包丟包率, 所述数据包丟包率为所述第一统计结果中和所述第二统计结果中同 一类包长的数据包的个数差值与所述第一统计结果中所述同一类包 长对应的数据包的个数的比值;
对比子单元,用于将所述预设个不同包长类别的数据包按照包长 类别进行升序排序, 所述预设个不同包长类别的数据包丟包率中至少 包括预设个数的数据包丟包率为升序排列, 则判定当前网络的丟包类 型为误码丟包; 若所述预设个不同包长类别的数据包丟包率中按照升 序排列的数据包丟包率的个数小于所述预设个数, 则判定当前网络的 丟包类型为拥塞丟包。
14、 根据权利要求 1 1 ~ 13任一项所述的设备, 其特征在于, 所述 第一网元设备还包括:
拟合单元, 用于当判定当前网络的丟包类型为误码丟包时, 将至 少一个包长类别的数据包的平均包长与所述至少一个包长类别的数 据包丟包率进行线性拟合;
重判单元,用于当拟合得到的曲线的斜率为上升趋势或者所述曲 线的斜率为大于预设斜率的固定值时, 则判定当前网络存在误码丟 包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于预 设的数据包丟包率阔值时, 则判定当前网络同时存在误码丟包及拥塞 丟包; 其中所述拟合得到的曲线以所述包长类别为坐标空间中的横坐 标, 以所述数据包丟包率为纵坐标。
15、 根据权利要求 1 1 ~ 13任一项所述的设备, 其特征在于, 所述 第一网元设备还包括:
时延重判单元, 用于当判定当前网络的丟包类型为误码丟包时, 获取所述第二网元设备当前接收至少一个数据包的时延及所述第二 网元设备接收数据包的平均时延, 当所述当前接收至少一个数据包的 时延大于所述接收数据包的平均时延时, 则判定当前网络同时存在误 码丟包及拥塞丟包; 当所述当前接收至少一个数据包的时延小于或等 于所述接收数据包的平均时延时, 判定当前网络存在误码丟包。
16、 一种第二网元设备, 其特征在于, 包括
数据接收单元, 用于接收第一网元设备发送的数据包;
数据分类单元,用于将所述数据接收单元接收到的数据包按照第 一分类方式进行分类, 所述第一分类方式为按所述数据包的包长或包 长的预设范围分类;
统计单元,用于统计所述数据分类单元分类后的所述接收到的数 据包中各类包长的数据包数量, 生成第二统计结果;
数据发送单元,用于将所述统计单元转发的所述第二统计结果发 送至所述第一网元设备, 以便所述第一网元设备根据所述第一统计结 果及第二统计结果判断所述第一网元设备与所述第二网元设备间的 网络丟包的类型;
或者,
丟包识别单元,用于接收所述第一网元设备发送的所述第一统计 结果, 根据所述统计单元转发的所述第一统计结果和所述第二统计结 果判断所述第一网元设备与所述第二网元设备间的网络丟包的类型; 其中,所述第一统计结果为所述第一网元设备将发送至所述第二 网元设备的数据包按所述第一分类方式分类后, 对所述发送至所述第 二网元设备的数据包中各类包长的数据包数量的统计结果。
17、 根据权利要求 16所述的设备, 其特征在于, 所述丟包识别 单元包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果中 至少一次获取第一类包长的数据包和第二类包长的数据包的统计结 果, 所述第一类包长的数据包的平均包长大于所述第二类包长的数据 包的平均包长;
处理子单元,用于计算所述第一类包长的数据包丟包率及所述第 二类包长的数据包丟包率, 所述数据包丟包率为所述第一统计结果中 和所述第二统计结果中同一类包长的数据包的个数差值与所述第一 统计结果中所述同一类包长的数据包的个数的比值;
判别子单元,用于当所述第一类包长的数据包丟包率大于所述第 二类包长的数据包丟包率的次数大于预设的次数时, 判定当前网络的 丟包类型为误码丟包; 当所述第一类包长的数据包丟包率小于或等于 所述第二类包长的数据包丟包率的次数大于预设的次数时, 判定当前 网络的丟包类型为拥塞丟包。
18、 根据权利要求 16所述的设备, 其特征在于, 所述丟包识别 单元包括:
釆样子单元,用于在所述第一统计结果中和所述第二统计结果中 根据所述第一分类方式, 获取预设个不同包长类别的数据包的统计结 果, 计算所述预设个不同包长类别中各个包长类别的数据包丟包率, 所述数据包丟包率为所述第一统计结果中和所述第二统计结果中同 一类包长的数据包的个数差值与所述第一统计结果中所述同一类包 长对应的数据包的个数的比值;
对比子单元,用于将所述预设个不同包长类别的数据包按照包长 类别进行升序排序, 所述预设个不同包长类别的数据包丟包率中至少 包括预设个数的数据包丟包率为升序排列, 则判定当前网络的丟包类 型为误码丟包; 若所述预设个不同包长类别的数据包丟包率中按照升 序排列的数据包丟包率的个数小于所述预设个数, 则判定当前网络的 丟包类型为拥塞丟包。
19、 根据权利要求 16~ 18任一项所述的设备, 其特征在于, 所述 第二网元设备包括:
拟合单元, 用于当判定当前网络的丟包类型为误码丟包时, 将至 少一个包长类别的数据包的平均包长与所述至少一个包长类别的数 据包丟包率进行线性拟合;
重判单元,用于当拟合得到的曲线的斜率为上升趋势或者所述曲 线的斜率为大于预设斜率的固定值时, 则判定当前网络存在误码丟 包; 当拟合得到的曲线中所有包长类别对应的数据包丟包率均大于预 设的数据包丟包率阔值时, 则判定当前网络同时存在误码丟包及拥塞 丟包; 其中所述拟合得到的曲线以所述包长类别为坐标空间中的横坐 标, 以所述数据包丟包率为纵坐标。
20、 根据权利要求 16~ 18任一项所述的设备, 其特征在于, 所述 第二网元设备还包括:
时延重判单元, 用于当判定当前网络的丟包类型为误码丟包时, 获取当前接收至少一个数据包的时延及接收数据包的平均时延, 当所 述当前接收至少一个数据包的时延大于所述接收数据包的平均时延 时, 则判定当前网络同时存在误码丟包及拥塞丟包; 当所述当前接收 至少一个数据包的时延小于或等于所述接收数据包的平均时延时, 判 定当前网络存在误码丟包。
21、 一种网络系统, 其特征在于, 包括至少一个第一网元设备和 一个第二网元设备, 其中所述第一网元设备和第二网元设备通过交换 网连接, 所述第一网元设备为权利要求 1 1 ~ 15 任一项所述的网元设 备, 所述第二网元设备为权利要求 16~20任一项所述的网元设备。
PCT/CN2012/087673 2012-12-27 2012-12-27 一种网络丢包类型识别方法、设备及系统 WO2014101047A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201280002149.2A CN104322017B (zh) 2012-12-27 2012-12-27 一种网络丢包类型识别方法、设备及系统
PCT/CN2012/087673 WO2014101047A1 (zh) 2012-12-27 2012-12-27 一种网络丢包类型识别方法、设备及系统

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2012/087673 WO2014101047A1 (zh) 2012-12-27 2012-12-27 一种网络丢包类型识别方法、设备及系统

Publications (1)

Publication Number Publication Date
WO2014101047A1 true WO2014101047A1 (zh) 2014-07-03

Family

ID=51019699

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/087673 WO2014101047A1 (zh) 2012-12-27 2012-12-27 一种网络丢包类型识别方法、设备及系统

Country Status (2)

Country Link
CN (1) CN104322017B (zh)
WO (1) WO2014101047A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10523549B1 (en) * 2019-06-02 2019-12-31 Cybertoka Ltd Method and system for detecting and classifying networked devices

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110913421B (zh) * 2018-09-18 2021-10-29 大唐移动通信设备有限公司 一种语音包数量的确定方法及装置
CN112953847B (zh) * 2021-01-27 2022-11-11 北京字跳网络技术有限公司 网络的拥塞控制方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101146033A (zh) * 2007-10-31 2008-03-19 重庆邮电大学 一种基于跨层协作机制提高无线自组织网络性能的方法
CN101686100A (zh) * 2008-09-25 2010-03-31 华为技术有限公司 处理丢包的方法、传输质量控制方法、装置及系统
CN102104468A (zh) * 2011-02-18 2011-06-22 中兴通讯股份有限公司 一种基于路由代理的媒体感知arq控制方法及系统
CN102204164A (zh) * 2011-05-24 2011-09-28 华为技术有限公司 网络丢包信息报告方法及装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070115814A1 (en) * 2003-03-29 2007-05-24 Regents Of The University Of California, The Method and apparatus for improved data transmission

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101146033A (zh) * 2007-10-31 2008-03-19 重庆邮电大学 一种基于跨层协作机制提高无线自组织网络性能的方法
CN101686100A (zh) * 2008-09-25 2010-03-31 华为技术有限公司 处理丢包的方法、传输质量控制方法、装置及系统
CN102104468A (zh) * 2011-02-18 2011-06-22 中兴通讯股份有限公司 一种基于路由代理的媒体感知arq控制方法及系统
CN102204164A (zh) * 2011-05-24 2011-09-28 华为技术有限公司 网络丢包信息报告方法及装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10523549B1 (en) * 2019-06-02 2019-12-31 Cybertoka Ltd Method and system for detecting and classifying networked devices

Also Published As

Publication number Publication date
CN104322017A (zh) 2015-01-28
CN104322017B (zh) 2017-04-26

Similar Documents

Publication Publication Date Title
US9391907B2 (en) Packet aggregation
JP3720345B2 (ja) 伝送装置
US8976669B2 (en) Switch fabric end-to-end congestion avoidance mechanism
KR100757872B1 (ko) 네트워크에서의 혼잡 발생 예고 시스템 및 방법
US6801530B1 (en) Communication system and method in a communication system
CN106612284B (zh) 一种流数据的传输方法和装置
CN108833293B (zh) 一种基于软件定义网络sdn的数据中心拥塞控制方法及装置
US20130311643A1 (en) System and method for latency reduction in a network environment
CN106878192B (zh) 一种自适应mptcp的数据调度方法
US8072886B2 (en) Method and system for transmission control protocol (TCP) traffic smoothing
CN103348647A (zh) 在无线网络中改进传输控制协议性能的机制
US9515940B2 (en) Method for transmitting data in a packet-oriented communications network and correspondingly configured user terminal in said communications network
US11165716B2 (en) Data flow processing method and device
US20110292800A1 (en) Systems and Methods For Controlling Data Transmission Rates
CN110138610A (zh) 一种业务报文的发送方法及装置
CN102724123A (zh) 网络流量控制方法及控制装置
CN106658644B (zh) 通信网络选路方法和装置
CN111756641A (zh) 一种发送设备的调整方法和通信装置
CN115037700A (zh) 一种复杂网络数据包传送方法、系统、终端及存储介质
WO2014101047A1 (zh) 一种网络丢包类型识别方法、设备及系统
US20200136944A1 (en) Data Transmission Performance Detection
CN111404872A (zh) 一种报文处理方法、装置及系统
US7535834B2 (en) Node device configuring ring network and data frame control method
JP3974027B2 (ja) 基地局制御装置、データ伝送方法及びプログラム
US8542618B2 (en) Systems and methods of improving protocol performance

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: 12890868

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12890868

Country of ref document: EP

Kind code of ref document: A1