WO2016023438A1 - 网络拓扑发现方法和设备 - Google Patents

网络拓扑发现方法和设备 Download PDF

Info

Publication number
WO2016023438A1
WO2016023438A1 PCT/CN2015/086155 CN2015086155W WO2016023438A1 WO 2016023438 A1 WO2016023438 A1 WO 2016023438A1 CN 2015086155 W CN2015086155 W CN 2015086155W WO 2016023438 A1 WO2016023438 A1 WO 2016023438A1
Authority
WO
WIPO (PCT)
Prior art keywords
link
same
network
links
port
Prior art date
Application number
PCT/CN2015/086155
Other languages
English (en)
French (fr)
Inventor
袁玉林
樊晓佶
叶智明
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 ES15831718T priority Critical patent/ES2745541T3/es
Priority to EP15831718.0A priority patent/EP3182656B1/en
Publication of WO2016023438A1 publication Critical patent/WO2016023438A1/zh
Priority to US15/429,946 priority patent/US10116516B2/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/16—Threshold monitoring
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks

Definitions

  • the embodiments of the present invention relate to the field of network connection detection, and in particular, to a network topology discovery method and device.
  • the operator needs to evaluate and optimize the network. It is necessary to analyze the network elements in the network and the service status, such as collecting network element configuration information, collecting port traffic information, performing port capacity evaluation, and discovering traffic overload. Port, expand it or adjust the traffic path.
  • the optimization analysis tool When using the optimization analysis tool to evaluate and optimize the network, it is necessary to rely on the optimization analysis tool to restore the network topology. Based on the network topology, network traffic assessment, service evaluation, simulation, and the results of the evaluation analysis of the network based on the network topology display can be performed.
  • the method for discovering the network topology in the prior art is: collecting network feature data of the network element of the network to be analyzed, calculating the link set corresponding to the algorithm according to the collected network feature data and the corresponding network topology discovery algorithm, and obtaining the network.
  • Topology For example, the network topology based on the port IP address is calculated according to the port Internet Protocol (IP) address and the IP address matching algorithm, or the network based on the port alias is calculated according to the port alias and the port alias matching algorithm.
  • IP Internet Protocol
  • CDP Cisco Discovery Protocol
  • the network topology discovery accuracy is low when the network to be analyzed does not support the network feature and the corresponding algorithm performs network topology discovery, or In the case of support, the link relationship in the network is uncertain due to incomplete data collection or data inconsistency, so the accuracy of network topology discovery is also low.
  • the results obtained by the network topology discovery are not comprehensively analyzed to improve the accuracy of network topology discovery.
  • the network topology discovery method and device provided by the embodiments of the present invention can comprehensively analyze the results obtained by using the network feature data to perform network topology discovery, and improve the accuracy of network topology discovery.
  • an embodiment of the present invention provides a network topology discovery method, where the method includes:
  • the credibility value of the link in the subset of the links is the same as the credibility value of the topology discovery algorithm corresponding to the subset of links, and the credibility values of the different topology discovery algorithms are different.
  • the link is a link composed of two ports of different network elements;
  • the merging the same link includes:
  • the operation further includes: retaining a credibility value in the at least two links After the largest link and deleting the remaining links,
  • the uncertainty inference algorithm includes:
  • CF i (H) is one of a plurality of credibility values of the same link
  • CF j (H) is among a plurality of credibility values of the same link.
  • Another confidence value, CF i,j (H) is the new confidence value for the same link calculated from CF i (H) and CF j (H).
  • the network feature data and the corresponding topology discovery algorithm comprise at least two of the following combinations: port internet protocol IP address and internet protocol IP address matching algorithm, port alias and port alias matching algorithm, port link layer discovery protocol LLDP neighbor information And port link layer discovery protocol LLDP link algorithm.
  • the embodiment of the present invention provides a network topology discovery device, where the device includes:
  • An acquiring unit configured to collect network feature data of all network elements of the network to be analyzed
  • Obtaining a link unit configured to acquire at least two link subsets corresponding to each type by using at least two topology discovery algorithms according to the network feature data, and collect all links in the at least two link subsets to Obtaining a first link set in a set; wherein, the credibility value of the link in each link subset is the same as the credibility value of the topology discovery algorithm corresponding to the link subset, and the different topology discovery algorithm The credibility values are different, and the link is a link composed of two ports of different network elements;
  • a link processing unit configured to obtain a second link set by performing operations on the first link set, where the operations include: merging the same link, and for at least two links having only one port being the same Retaining the link with the highest credibility value in the at least two links and deleting the remaining links, where the same link is at least two links including the same two ports;
  • Obtaining a topology unit configured to acquire a network topology of the network to be analyzed according to each link in the second link set.
  • the link processing unit is specifically configured to:
  • Obtaining the second link set by performing operations on the first link set comprising: merging the same link in the first link set and according to multiple of the same link
  • the credibility value and the uncertainty inference algorithm calculate a credibility value of the link retained after the merge, and for at least two links having only one port the same, retaining the at least two links
  • the link with the highest reliability value and the remaining links are deleted, wherein the same link is at least two links including the same two ports.
  • the link processing unit is specifically configured to:
  • the credibility value and the uncertainty inference algorithm calculate a credibility value of the link retained after the merge, and for at least two links having only one port the same, retaining the at least two links
  • the link with the highest reliability value deletes the remaining links, compares the reliability value of the link in the first link set with a preset threshold, and selects a link whose credibility value is greater than a preset threshold.
  • the same link is at least two links that include the same two ports.
  • the uncertainty inference algorithm includes:
  • CF i (H) is one of a plurality of credibility values of the same link
  • CF j (H) is among a plurality of credibility values of the same link.
  • Another confidence value, CF i,j (H) is a new confidence value for calculating the same link according to CF i (H) and CF j (H).
  • the network feature data and the corresponding topology discovery algorithm comprise at least two of the following combinations: port internet protocol IP address and internet protocol IP address matching algorithm, port alias and port alias matching algorithm, port link layer discovery protocol LLDP neighbor information And port link layer discovery Protocol LLDP link algorithm.
  • the network topology discovery method and device provided by the embodiment of the present invention firstly collect network feature data of all network elements of the network to be analyzed; and then, according to the network feature data, obtain at least two corresponding link pairs by using at least two topology discovery algorithms respectively. Collecting, concentrating all the links in the at least two link subsets into one set to obtain a first link set; and performing operations on the first link set to obtain a second link set, where the operations include: merging The same link, and for at least two links with only one port being the same, retain the link with the highest credibility value in at least two links and delete the remaining links, where the same link is included The two ports have the same at least two links; finally, the network topology of the network to be analyzed is obtained according to each link in the second link set. In this way, the results obtained by network topology discovery using multiple network feature data can be comprehensively analyzed to improve the accuracy of network topology discovery.
  • FIG. 1 is a schematic flowchart 1 of a network topology discovery method according to an embodiment of the present disclosure
  • FIG. 2 is a schematic diagram of an actual network topology assumed in an embodiment of the present invention and an effect of using a network topology extension obtained by using three topology discovery algorithms;
  • FIG. 3 is a schematic flowchart 2 of a network topology discovery method according to an embodiment of the present disclosure
  • FIG. 4 is a schematic structural diagram 1 of a network topology discovery device according to an embodiment of the present disclosure
  • FIG. 5 is a schematic structural diagram 2 of a network topology discovery device according to an embodiment of the present disclosure.
  • the network topology discovery method provided by the embodiment of the present invention is as shown in FIG. 1 , and the method includes:
  • Step 101 Collect network characteristic data of all network elements of the network to be analyzed.
  • Step 102 According to the network feature data, obtain at least two corresponding link subsets by using at least two topology discovery algorithms, and collect all the links in the at least two link subsets into one set to obtain a first link set. .
  • the credibility value of the link in each link sub-set is the same as the credibility value of the topology discovery algorithm corresponding to each link subset.
  • the credibility values of different topology discovery algorithms are different, and the links are different network elements.
  • Step 103 Obtain a second link set by performing operations on the first link set, where the operations include: merging the same link, and retaining at least two links for at least two links having only one port being the same The link with the highest credibility value and deletes the remaining links, where the same link is at least two links that contain the same two ports.
  • Step 104 Acquire a network topology of the network to be analyzed according to each link in the second link set.
  • the network topology discovery method provided by the embodiment of the present invention firstly collects network feature data of all network elements of the network to be analyzed; and then obtains at least two corresponding link subsets by using at least two topology discovery algorithms according to the network feature data. Concentrating all the links in the at least two link subsets into one set to obtain a first link set; and performing operations on the first link set to obtain a second link set, where the operations include: merging the same a link, and for at least two links having the same port, retaining the link with the highest credibility value in at least two links and deleting the remaining links, wherein the same link is the two included At least two links having the same port; and finally obtaining a network topology of the network to be analyzed according to each link in the second link set.
  • the results obtained by network topology discovery using multiple network feature data can be comprehensively analyzed to improve the accuracy of network topology discovery.
  • the network element is a network element or node in a network system
  • the unit is a device capable of independently performing one or several functions.
  • a base station is a network element; an entity that can perform a function alone can
  • a network element, a switch, a router, and the like also belong to a network element, and the link may be a physical link or a logical link.
  • the jth port is denoted as Pij, the range of i is [1, n], the range of j is [1, m], m is the number of ports corresponding to the network element N i , and the m values of different network elements may be different. .
  • the network tool feature data of all network elements of the network to be analyzed can be collected by using the collection tool, and the network feature data is used as the data input of the topology discovery algorithm. It can be clarified that the foregoing collection of network feature data of the network element using the collection tool can be implemented by those skilled in the art.
  • the network feature data includes at least two of the following: port IP address, port alias, port LLDP neighbor information, network element name, port traffic, media access control (MAC) forwarding table, and address resolution protocol (Address Resolution) Protocol, ARP) forwarding table, routing forwarding table, and virtual local area network (VLAN) configuration information.
  • the topology discovery algorithm uses: IP address matching algorithm, port alias algorithm and Port LLDP link algorithm, it should be noted that the selection method is only exemplary, and is only for the convenience of the technical solution in the embodiment. In practical applications, those skilled in the art can collect network feature data according to actual needs and Select the corresponding topology algorithm.
  • the network topology discovery method provided by the embodiment of the present invention based on the foregoing content includes:
  • Step 201 Collect network characteristic data of all network elements of the network to be analyzed.
  • the port IP address, the port alias, and the port LLDP neighbor information of the 20 ports of the five network elements of the network to be analyzed are respectively collected.
  • Step 202 Acquire, by using at least two topology discovery algorithms, respectively, according to network feature data. At least two link subsets are required, and all the links in the at least two link subsets are grouped into one set to obtain a first link set.
  • the credibility value of the link in each link sub-set is the same as the credibility value of the topology discovery algorithm corresponding to each link subset.
  • the credibility values of different topology discovery algorithms are different, and the links are different network elements.
  • the port IP address of the 20 ports is used as the input of the IP address matching algorithm, and the link subset L1 corresponding to the algorithm is obtained after calculation, and the corresponding link calculated according to the port alias and the port alias matching algorithm is obtained in turn.
  • the subset L2 and the corresponding link subset L3 calculated according to the port LLDP neighbor information and the port LLDP link algorithm, and the links in the link subsets L1, L2, and L3 are grouped into one set to obtain the first link set.
  • Case 1 two or more links exist in the same link, and the same two links mean that the two ports included in the same are the same; Case 2: There are only two or more links with the same link on one port.
  • different credibility values CF are set, and the value range of the credibility value CF can be set to [-1, 1], and is calculated according to a topology discovery algorithm.
  • the credibility value of all links in the obtained link set is the same as the credibility value of the topology discovery algorithm (or corresponding network feature data).
  • Step 203 Combine the same link in the first link set, and calculate a credibility value of the link retained after the merge according to respective credibility values of the same link and an uncertainty inference algorithm.
  • the same link is at least two links including the same two ports, for example, L12, L22, and L31 in Table 4 above. Taking the combined links L12, L22 and L31 as an example, the reliability value of L12 retained after the combination is calculated according to the uncertainty inference algorithm.
  • the uncertainty inference algorithm in the embodiment of the present invention is based on a credibility value, and the uncertainty inference algorithm includes:
  • CF i (H) is one of the plurality of credibility values of the same link
  • CF j (H) is the other of the plurality of credibility values of the same link.
  • the degree value, CF i,j (H) is the new confidence value of the same link calculated from CF i (H) and CF j (H).
  • CF i,j (H) CF i (H)+CF j (H)-CF i (H) ⁇ CF j (H);
  • CF i,j (H) CF i (H)+CF j (H)+CF i (H) ⁇ CF j (H);
  • the method for calculating the reliability value of the link retained after the combination is not limited to the above calculation formula, and those skilled in the art may also use other calculation methods to calculate, for example, a weighted summation algorithm may be used, to L12. (P22, P32), L22 (P22, P32), L31 (P22, P32), for example, the credibility value of the IP address matching algorithm, the credibility value of the port alias matching algorithm, and the port LLDP link algorithm can be set separately.
  • the weighting factor of the credibility value is 0.2, 0.3, 0.6, and the calculation
  • the method for calculating the integrated credibility value of the link according to the multiple credibility values of the same link is not limited in the embodiment of the present invention, and a computing method provided in the embodiment of the present invention may be used by a person skilled in the art. Other calculation methods can also be used.
  • Step 204 For at least two links with only one port being the same in the first link set, retain the link with the highest credibility value in at least two links and delete the remaining links.
  • L23 and L33 are only one port P44, L25 and L33 are only one port P54, first retain L33 in L23 and L33, delete L23, and then retain L33 in L25 and L33 to delete L25; Or you can first keep L33 in L25 and L33 and delete L25, then keep L33 in L23 and L33 and delete L23.
  • the order here is not limited, just follow the choice of multiple links with only one port. The principle of the link with the highest degree can be filtered.
  • Step 205 Compare the reliability value of the link in the first link set with a preset threshold, and select a link whose credibility value is greater than a preset threshold to obtain a second link set.
  • the reliability value of each link in the first link set obtained after step 203 and step 204 is compared with a preset threshold value of 0.7, and a link with a credibility value greater than 0.7 is selected to obtain a first Two link set G'.
  • a preset threshold value of 0.7 a link with a credibility value greater than 0.7 is selected to obtain a first Two link set G'.
  • Step 206 Acquire a network topology of the network to be analyzed according to each link in the second link set.
  • the network of the network to be analyzed can be obtained according to each link L11 (P11, P21), L12 (P22, P32), L13 (P33, P43), L33 (P44, P54) of the second link set G'.
  • the port P33 of the network element N3 is connected to the port P43 of the network element N4, and the network element N4 is connected.
  • the port P44 is connected to the port P54 of the network element N5.
  • the network topology discovery method provided by the embodiment of the present invention firstly collects network feature data of all network elements of the network to be analyzed; and then obtains at least two corresponding link subsets by using at least two topology discovery algorithms according to the network feature data. Concentrating all the links in the at least two link subsets into one set to obtain a first link set; and performing operations on the first link set to obtain a second link set, where the operations include: merging the same a link, and for at least two links having the same port, retaining the link with the highest credibility value in at least two links and deleting the remaining links, wherein the same link is the two included At least two links having the same port; and finally obtaining a network topology of the network to be analyzed according to each link in the second link set.
  • the results obtained by network topology discovery using multiple network feature data can be comprehensively analyzed to improve the accuracy of network topology discovery.
  • the embodiment of the present invention provides a network topology discovery device 00.
  • the device 00 includes:
  • the collecting unit 10 is configured to collect network feature data of all network elements of the network to be analyzed.
  • the obtaining link unit 20 is configured to obtain at least two corresponding link subsets by using at least two topology discovery algorithms according to the network feature data, and collect all the links in the at least two link subsets into one set.
  • the first link set wherein the reliability value of the link in each link subset is the same as the reliability value of the topology discovery algorithm corresponding to each link subset, and different topology discovery algorithms
  • the credibility value is different, and the link is a link composed of two ports of different network elements.
  • the link processing unit 30 is configured to obtain a second link set by performing operations on the first link set, where the operations include: merging the same link, and for at least two links having only one port being the same, The link with the highest credibility value in at least two links is retained and the remaining links are deleted, wherein the same link is at least two links including the same two ports.
  • the acquiring topology unit 40 is configured to obtain a network topology of the network to be analyzed according to each link in the second link set.
  • the link processing unit 30 is specifically configured to:
  • a second set of links is obtained by performing operations on the first set of links, the operations comprising: merging the same links in the first set of links and based on multiple credibility values and uncertainties of the same link
  • the inference algorithm calculates the credibility value of the link retained after the merge, and for at least two links having the same port, retains the link with the highest credibility value in at least two links and deletes the remaining chains.
  • a road in which the same link is at least two links including the same two ports.
  • the link processing unit 30 is further configured to:
  • a second set of links is obtained by performing operations on the first set of links, the operations comprising: merging the same links in the first set of links and based on multiple credibility values and uncertainties of the same link
  • the inference algorithm calculates the credibility value of the link retained after the merge, and for at least two links having the same port, retains the link with the highest credibility value in at least two links and deletes the remaining chains.
  • the route compares the reliability value of the link in the first link set with a preset threshold, and selects a link whose credibility value is greater than a preset threshold, where the same link is the same as the two ports included. At least two links.
  • the uncertainty reasoning algorithm includes:
  • CF i (H) is one of the plurality of credibility values of the same link
  • CF j (H) is the other of the plurality of credibility values of the same link.
  • the degree value, CF i,j (H) is the new confidence value of the same link calculated from CF i (H) and CF j (H).
  • the network feature data and the corresponding topology discovery algorithm comprise at least two of the following: a port internet protocol IP address and an internet protocol IP address matching algorithm, a port alias and a port alias matching algorithm, and a port link layer discovery protocol LLDP. Neighbor information and port link layer discovery protocol LLDP link algorithm.
  • the network topology discovery device provided by the embodiment of the present invention first collects network feature data of all network elements of the network to be analyzed, and then obtains at least two corresponding link subsets by using at least two topology discovery algorithms according to the network feature data. Concentrating all the links in the at least two link subsets into one set to obtain a first link set; and performing operations on the first link set to obtain a second link set, where the operations include: merging the same a link, and for at least two links having the same port, retaining the link with the highest credibility value in at least two links and deleting the remaining links, wherein the same link is the two included At least two links having the same port; and finally obtaining a network topology of the network to be analyzed according to each link in the second link set. In this way, the results obtained by network topology discovery using multiple network feature data can be comprehensively analyzed to improve the accuracy of network topology discovery.
  • the embodiment of the present invention further provides a network topology discovery device 90.
  • the device 90 includes: a bus 94; and a processor 91, a memory 92 and an interface 93 connected to the bus 94, wherein the interface 93 is used.
  • the memory 92 is for storing instructions
  • the processor 91 is configured to execute the instructions for:
  • the reliability value of the link in the link sub-set is the same as the credibility value of the topology discovery algorithm corresponding to the link subset.
  • the credibility values of different topology discovery algorithms are different, and the link is two different network elements.
  • the network topology of the network to be analyzed is obtained according to each link in the second link set.
  • the processor 91 is configured to use the instruction to merge the same link, and specifically includes:
  • the same link in the first link set is merged, and the credibility values of the merged link are calculated according to multiple credibility values of the same link and an uncertainty inference algorithm.
  • the processor 91 executes the instruction, by performing operations on the first link set, to obtain a second link set, where the operation further comprises: retaining a maximum credibility value in at least two links. After the link and delete the remaining links,
  • the uncertainty reasoning algorithm includes:
  • CF i (H) is one of the plurality of credibility values of the same link
  • CF j (H) is the other of the plurality of credibility values of the same link.
  • the degree value, CF i,j (H) is the new confidence value of the same link calculated from CF i (H) and CF j (H).
  • the network feature data and the corresponding topology discovery algorithm comprise at least two of the following: a port internet protocol IP address and an internet protocol IP address matching algorithm, a port alias and a port alias matching algorithm, and a port link layer discovery protocol LLDP. Neighbor information and port link layer discovery protocol LLDP link algorithm.
  • the network topology discovery device provided by the embodiment of the present invention first collects network feature data of all network elements of the network to be analyzed, and then obtains at least two corresponding link subsets by using at least two topology discovery algorithms according to the network feature data. Concentrating all the links in the at least two link subsets into one set to obtain a first link set; and performing operations on the first link set to obtain a second link set, the operations include: merging the same link And for at least two links having only one port the same, retaining the link with the highest credibility value in at least two links and deleting the remaining links, wherein the same link is the two ports included At least two links that are identical; last The network topology of the network to be analyzed is obtained according to each link in the second link set. In this way, the results obtained by network topology discovery using multiple network feature data can be comprehensively analyzed to improve the accuracy of network topology discovery.
  • the aforementioned program can be stored in a computer readable storage medium.
  • the program when executed, performs the steps including the foregoing method embodiments; and the foregoing storage medium includes various media that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

Abstract

本发明实施例提供的网络拓扑发现方法和设备，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。具体方案为：首先采集待分析网络的网元的网络特征数据，然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个的链路子集中的所有链路集中到一个集合中得到第一链路集，在第一链路集中合并相同链路并对仅有一个端口相同的多个链路只保留其中可信度值最大的链路同时删除其余的链路，得到第二链路集，最后根据第二链路集得到待分析网络的网络拓扑。本发明实施例用于网络拓扑发现。

Description

网络拓扑发现方法和设备
本申请要求于2014年8月12日提交中国专利局、申请号为201410395692.4、发明名称为“网络拓扑发现方法和设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明实施例涉及网络连接检测领域，尤其涉及一种网络拓扑发现方法和设备。
背景技术
在网络评估优化阶段，运营商需要对网络进行评估和优化，需要分析网络中的网元，以及业务状态，比如：采集网元配置信息、采集端口流量信息，进行端口容量评估，发现流量过载的端口，对其进行扩容或者调整流量路径。使用优化分析工具对网络进行评估优化时，需要依赖优化分析工具还原网络拓扑，基于网络拓扑才能进行网络流量评估、业务评估、仿真，以及基于网络拓扑显示对网络的评估分析的结果。
现有技术中的网络拓扑发现的方法是：采集待分析网络的网元的网络特征数据，根据采集到的网络特征数据以及对应的网络拓扑发现算法计算得到对应该算法的链路集合进而得到网络拓扑。例如：根据端口互联网协议(Internet Protocol，IP)地址以及IP地址匹配算法计算得到基于端口IP地址这一特征的网络拓扑，或根据端口别名以及端口别名匹配算法计算得到基于端口别名这一特征的网络拓扑，或根据思科发现协议(Cisco Discovery Protocol，CDP)可以得到基于思科设备组建的网络的网络拓扑(仅支持由思科设备组建的网络)等等。
在现有技术中，基于单一的一种网络特征数据进行网络拓扑发现时，在待分析网络不支持该种网络特征和对应算法进行网络拓扑发现的情况下网络拓扑发现的准确率低，或者在支持的情况下由于数据收集不完整或数据不一致导致网络中链路关系存在不确定性，从而网络拓扑发现的准确率也较低，另外在使用多种网络特征数据进行网络拓扑发现时，对多种网络拓扑发现得到的结果也不做综合分析以提高网络拓扑发现的准确率。
发明内容
本发明实施例提供的网络拓扑发现方法和设备，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
第一方面，本发明实施例提供一种网络拓扑发现方法，所述方法包括：
采集待分析网络的所有网元的网络特征数据；
根据所述网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将所述至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与所述链路子集对应的拓扑发现算法的可信度值相同，所述不同拓扑发现算法的可信度值不同，所述链路为不同网元的两个端口组成的链路；
通过对所述第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路；
根据所述第二链路集中的每个链路获取所述待分析网络的网络拓扑。
结合第一方面，在第一种可能的实现方式中，所述合并相同的链路包括：
合并所述第一链路集中的相同的链路，并根据所述相同的链路的多个可信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值。
结合第一方面或第一方面的第一种可能的实现方式，在第二种可能的实现方式中，所述操作还包括：在所述保留所述至少两个链路中的可信度值最大的链路并删除其余的链路之后，
将所述第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路。
结合第一方面的第一种可能的实现方式，在第三种可能的实现方式中，所述不确定性推理算法包括：
Figure PCTCN2015086155-appb-000001
其中，CFi(H)为所述相同的链路的多个可信度值中的一个可信度值，CFj(H)为所述相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算的所述相同的链路的新的可信度值。
结合第一方面至第一方面的第三种可能的实现方式中任一可能的实现方式，在第四种可能的实现方式中，
所述网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现协议LLDP链路算法。
第二方面，本发明实施例提供一种网络拓扑发现设备，所述设备包括：
采集单元，用于采集待分析网络的所有网元的网络特征数据；
获取链路单元，用于根据所述网络特征数据，利用至少两种拓扑发现算法分别获取种对应的至少两个链路子集，将所述至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与所述链路子集对应的拓扑发现算法的可信度值相同，所述不同拓扑发现算法的可信度值不同，所述链路为不同网元的两个端口组成的链路；
链路处理单元，用于通过对所述第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路,其中，所述相同的链路为包含的两个端口均相同的至少两个链路；
获取拓扑单元，用于根据所述第二链路集中的每个链路获取所述待分析网络的网络拓扑。
结合第二方面，在第一种可能的实现方式中，所述链路处理单元具体用于：
通过对所述第一链路集执行操作，得到所述第二链路集，所述操作包括：合并所述第一链路集中的相同的链路并根据所述相同的链路的多个可信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路。
结合第二方面，在第二种可能的实现方式中，所述链路处理单元具体用于：
通过对所述第一链路集执行操作，得到所述第二链路集，所述操作包括：合并所述第一链路集中的相同的链路并根据所述相同的链路的多个可信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，将所述第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路。
结合第二方面的第一种可能的实现方式，在第三种可能的实现方式中，所述不确定性推理算法包括：
Figure PCTCN2015086155-appb-000002
其中，CFi(H)为所述相同的链路的多个可信度值中的一个可信度值，CFj(H)为所述相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算所述相同的链路的新的可信度值。
结合第二方面至第二方面的第三种可能的实现方式中任一可能的实现方式，在第四种可能的实现方式中，
所述网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现 协议LLDP链路算法。
本发明实施例提供的网络拓扑发现方法和设备，首先采集待分析网络的所有网元的网络特征数据；然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；再通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；最后根据第二链路集中的每个链路获取待分析网络的网络拓扑。这样，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
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一简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本发明实施例提供的网络拓扑发现方法的流程示意图一；
图2为本发明实施例中假设的实际的网络拓扑以及使用三种拓扑发现算法得到的网络物理拓的效果示意图；
图3为本发明实施例提供的网络拓扑发现方法的流程示意图二；
图4为本发明实施例提供的网络拓扑发现设备的结构示意图一；
图5为本发明实施例提供的网络拓扑发现设备的结构示意图二。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例提供的网络拓扑发现方法,如图1所示，该方法包括：
步骤101、采集待分析网络的所有网元的网络特征数据。
步骤102、根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集。
其中，各链路子集中的链路的可信度值与各链路子集对应的拓扑发现算法的可信度值相同，不同拓扑发现算法的可信度值不同，链路为不同网元的两个端口组成的链路。
步骤103、通过对第一链路集执行操作，得到第二链路集，该操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路。
步骤104、根据第二链路集中的每个链路获取待分析网络的网络拓扑。
本发明实施例提供的网络拓扑发现方法，首先采集待分析网络的所有网元的网络特征数据；然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；再通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；最后根据第二链路集中的每个链路获取待分析网络的网络拓扑。这样，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
为了使本领域技术人员能够更清楚地理解本发明实施例提供的技术方案，下面通过具体的实施例，对本发明的实施例提供的网络拓扑发现方法进行详细说明：
在介绍本实施例提供的技术方案前，对技术方案中的一些基本内容做简单介绍如下：
在本实施例提供的技术方案中，网元就是一个网络系统中的某个网络单元或者节点，该单元是能独立完成一种或几种功能的设备。例如：在GSM网络系统中，一个基站就是一个网元；能单独完成一项功能的实体就可以 成为一个网元，交换机、路由器等也属于网元，链路可以是物理链路或逻辑链路。
定义待分析网络的所有网元组成的集合为N＝{N1，N2，...，Nn}；所有网元的所有端口组成的集合为P；其中，第i个网元Ni的第j个端口记为Pij，i的范围为[1，n]，j的范围为[1，m]，m为网元Ni对应的端口个数，不同网元的m取值可以不同。
使用采集工具可以采集待分析网络的所有网元的网络特征数据，网络特征数据作为拓扑发现算法的数据输入。可以明确的是前述使用采集工具采集网元的网络特征数据对于本领域内的普通技术人员都可以实现。网络特征数据包括以下中的至少两种：端口IP地址、端口别名、端口LLDP邻居信息、网元名称、端口流量、媒体接入控制(Media Access Control，MAC)转发表、地址解析协议(Address Resolution Protocol，ARP)转发表、路由转发表以及虚拟局域网(Virtual Local Area Network，VLAN)配置信息。
为了方便阐述本发明实施例提供的技术方案，在下面的实施例中，假设待分析网络的网元个数为5(即n＝5)，所有网元组成的集合为N＝{N1，N2，N3，N4，N5}，每个网元的端口个数为4个(即m＝4)，所有网元的所有物理端口组成的集合为P＝{P11，P12，P13，P14，P21，P22，...，P51，P52，P53，P54}，采集时网络特征数据选择采集：端口IP地址、端口别名和端口LLDP邻居信息，拓扑发现算法对应采用：IP地址匹配算法、端口别名算法以及端口LLDP链路算法，需要说明的是，这样的选取方式只是示例性的，仅仅是为了在实施例中方便阐述技术方案，在实际应用中，本领域技术人员可以根据实际需要采集网络特征数据以及选择对应的拓扑算法。
假设实际的网络拓扑和使用IP地址匹配算法、端口别名算法以及端口LLDP链路算法分别计算得到的网络拓扑如图2所示。
如图3所示，为基于上述内容的本发明实施例提供的网络拓扑发现方法，该方法包括：
步骤201、采集待分析网络的所有网元的网络特征数据。
示例性的，分别采集待分析网络的5个网元的20个端口的端口IP地址、端口别名和端口LLDP邻居信息。
步骤202、根据网络特征数据，利用至少两种拓扑发现算法分别获取对 应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集。
其中，各链路子集中的链路的可信度值与各链路子集对应的拓扑发现算法的可信度值相同，不同拓扑发现算法的可信度值不同，链路为不同网元的两个端口组成的链路。
示例性的，将20个端口的端口IP地址作为IP地址匹配算法的输入，经过计算后得到该算法对应的链路子集L1，依次得到根据端口别名和端口别名匹配算法计算的对应的链路子集L2以及根据端口LLDP邻居信息和端口LLDP链路算法计算的对应的链路子集L3，将链路子集L1、L2和L3中的链路集中到一个集合中得到第一链路集G，即G＝L1+L2+L3。对于第一链路集G，其中的链路可能存在以下两种情况，情况一：同一条链路存在两个或两个以上，两条链路相同是指其中包含的两个端口均相同；情况二：仅有一个端口相同的链路存在两个或两个以上。
其中，链路子集由若干个链路组成，例如Li＝Li1∪Li2∪...∪Lik，其中Lik为对应算法i计算得到的第k条链路，k∈[1，K]，K为对应算法计算得到的链路的数量；链路Lik由两个端口组成，Lik＝(端口1，端口2)，端口1与端口2是不同的端口，端口1和端口2均属于集合P，对于一条链路Lik其中的端口1和端口2没有顺序之分。
对应不同的拓扑发现算法(或对应的网络特征数据)设置不同的可信度值CF，可信度值CF的取值范围可以设置为[-1,1]，同时根据一种拓扑发现算法计算得到的链路集合中的所有链路的可信度值与该拓扑发现算法(或对应的网络特征数据)的可信度值相同。
为了使本领域技术人员更加清楚上述的内容，可以参考图2以及以下的表格：
表1
算法 算法名称 对应的网络特征数据 算法可信度值
1 IP地址匹配算法 端口IP地址 0.8
2 端口别名匹配算法 端口别名 0.5
3 LLDP链路算法 端口LLDP邻居信息 1.0
表2
Figure PCTCN2015086155-appb-000003
表3
Figure PCTCN2015086155-appb-000004
步骤203、合并第一链路集中的相同的链路，并根据多个相同的链路各自的可信度值以及不确定性推理算法计算合并后保留的链路的可信度值。
示例性的，对于第一链路集G中存在的至少两个相同的链路，合并多个相同的链路，即在第一链路集G中只保留一个，并根据多个相同的链路各自的可信度值以及不确定性推理算法计算合并后保留的链路的可信度值。参考下表4：
表4
Figure PCTCN2015086155-appb-000005
其中，相同的链路为包含的两个端口均相同的至少两个链路，例如上表4中的L12、L22以及L31。现以合并链路L12、L22以及L31为例，对根据不确定性推理算法计算合并后保留的L12的可信度值进行说明。
本发明实施例中的不确定性推理算法是基于可信度值的，不确定性推理算法包括：
Figure PCTCN2015086155-appb-000006
其中，CFi(H)为相同的链路的多个可信度值中的一个可信度值，CFj(H)为相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算的相同的链路的新的可信度值。
对上述公式的具体含义进行详细说明：
当CFi(H)≥0且CFj(H)≥0时，使用公式
CFi,j(H)＝CFi(H)+CFj(H)－CFi(H)×CFj(H)；
当CFi(H)＜0且CFj(H)＜0时，使用公式
CFi,j(H)＝CFi(H)+CFj(H)+CFi(H)×CFj(H)；
当CFi(H)与CFj(H)异号时，也即CFi(H)≥0且CFj(H)＜0，或者CFj(H)≥0且CFi(H)＜0时，使用公式
Figure PCTCN2015086155-appb-000007
对于相同的链路有两个以上可信度值的，计算合并后保留的链路的可信度值时，先选取两个可信度值作为上述公式的输入得到一个新的可信度值，然后将新的可信度值与一个未参加过计算的可信度值作为上述公式的输入得到一个更新的可信度值，重复前述步骤直至该链路的所有可信度值均在公式中参加过一次计算，最后得到合并后保留的链路的可信度值。
例如，计算合并后保留的链路L12的可信度值：
首先选取L12以及L22的可信度值0.8和0.5作为上述公式的输入即0.8+0.5-0.8×0.5＝0.9,然后把0.9和L31的可信度值1.0作为上述公式的输入即0.9+1.0-0.9×1.0＝1.0，最后将1.0作为合并后保留的链路L12的可信度值。
需要说明的是，计算合并后保留的链路的可信度值的方法并不限于上述的计算公式，本领域技术人员还可以利用其他的计算方法计算，例如可以采用加权求和算法，以L12(P22，P32)、L22(P22，P32)、L31(P22，P32)为例，可以分别设置IP地址匹配算法的可信度值、端口别名匹配算法的可信度值以及端口LLDP链路算法的可信度值的加权系数为0.2、0.3、0.6，计算 合并后保留的链路L12的可信度值：0.8×0.2+0.5×0.3+1.0×0.6＝0.91。本发明实施例对根据相同的链路的多个可信度值计算该链路综合可信度值的方法并不做限定，本领域普通技术人员可以采用本发明实施例中提供的计算方法，也可以采用别的计算方法。
步骤204、对于第一链路集中仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路。
示例性的，参考下表5：
表5
Figure PCTCN2015086155-appb-000008
参照上述表格5，其中L23与L33中仅有一个端口P44相同，L25与L33中仅有一个端口P54相同，先在L23与L33中保留L33删除L23，然后在L25与L33中保留L33删除L25；或者也可以先在L25与L33中保留L33删除L25，然后在L23与L33中保留L33删除L23，这里的顺序不作限定，只需按照对存在仅有一个端口相同的多个链路选择保留可信度值最高的链路这样的原则进行筛选即可。
步骤205、将第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路，得到第二链路集。
示例性的，将根据步骤203以及步骤204后得到的第一链路集中的各个链路的可信度值与预设阈值0.7进行比较，挑选出可信度值大于0.7的链路，得到第二链路集G’。参考下表6：
表6
Figure PCTCN2015086155-appb-000009
步骤206、根据第二链路集中的每个链路获取待分析网络的网络拓扑。
示例性的，根据第二链路集G’的各个链路L11(P11，P21)、L12(P22，P32)、L13(P33，P43)、L33(P44，P54)可以得到待分析网络的网络拓扑：网元N1的端口P11与网元N2的端口P21相连，网元N3的端口P22与网元N2的端口P32相连，网元N3的端口P33与网元N4的端口P43相连，网元N4的端口P44与网元N5的端口P54相连。可以看出，根据上述实施例提供的技术方案得到的待分析网络的网络拓扑与图2所示的待分析网络的网络拓扑是一致的。
本发明实施例提供的网络拓扑发现方法，首先采集待分析网络的所有网元的网络特征数据；然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；再通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；最后根据第二链路集中的每个链路获取待分析网络的网络拓扑。这样，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
本发明实施例提供一种网络拓扑发现设备00，如图4所示，该设备00包括：
采集单元10，用于采集待分析网络的所有网元的网络特征数据。
获取链路单元20，用于根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与各链路子集对应的拓扑发现算法的可信度值相同，不同拓扑发现算法 的可信度值不同，链路为不同网元的两个端口组成的链路。
链路处理单元30，用于通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路,其中，相同的链路为包含的两个端口均相同的至少两个链路。
获取拓扑单元40，用于根据第二链路集中的每个链路获取待分析网络的网络拓扑。
可选的，链路处理单元30具体用于：
通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并第一链路集中的相同的链路并根据相同的链路的多个可信度值以及不确定性推理算法计算合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路。
可选的，链路处理单元30还可以具体用于：
通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并第一链路集中的相同的链路并根据相同的链路的多个可信度值以及不确定性推理算法计算合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，将第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路。
可选的，不确定性推理算法包括：
Figure PCTCN2015086155-appb-000010
其中，CFi(H)为相同的链路的多个可信度值中的一个可信度值，CFj(H)为相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算的相同的链路的新的可信度值。
可选的，网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现协议LLDP链路算法。
本实施例用于实现上述各方法实施例，本实施例中各个单元的工作流程和工作原理参见上述各方法实施例中的描述，在此不再赘述。
本发明实施例提供的网络拓扑发现设备，首先采集待分析网络的所有网元的网络特征数据；然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；再通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；最后根据第二链路集中的每个链路获取待分析网络的网络拓扑。这样，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
本发明实施例还提供了一种网络拓扑发现设备90，如图5所示，该设备90包括：总线94；以及连接到总线94的处理器91、存储器92和接口93，其中该接口93用于通信；该存储器92用于存储指令，处理器91用于执行该指令用于：
采集待分析网络的所有网元的网络特征数据；
根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与链路子集对应的拓扑发现算法的可信度值相同，不同拓扑发现算法的可信度值不同，链路为不同网元的两个端口组成的链路；
通过对第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；
根据第二链路集中的每个链路获取待分析网络的网络拓扑。
可选地，处理器91执行该指令用于合并相同的链路，具体可以包括：
合并第一链路集中的相同的链路，并根据相同的链路的多个可信度值以及不确定性推理算法计算合并后保留的链路的可信度值。
可选地，处理器91执行该指令用于通过对第一链路集执行操作，得到第二链路集，所述操作还包括：在保留至少两个链路中的可信度值最大的链路并删除其余的链路之后，
将第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路。
可选的，不确定性推理算法包括：
Figure PCTCN2015086155-appb-000011
其中，CFi(H)为相同的链路的多个可信度值中的一个可信度值，CFj(H)为相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算的相同的链路的新的可信度值。
可选的，网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现协议LLDP链路算法。
本实施例用于实现上述各方法实施例，本实施例中各个单元的工作流程和工作原理参见上述各方法实施例中的描述，在此不再赘述。
本发明实施例提供的网络拓扑发现设备，首先采集待分析网络的所有网元的网络特征数据；然后根据网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；再通过对第一链路集执行操作，得到第二链路集，操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留至少两个链路中的可信度值最大的链路并删除其余的链路，其中，相同的链路为包含的两个端口均相同的至少两个链路；最后 根据第二链路集中的每个链路获取待分析网络的网络拓扑。这样，可以对使用多种网络特征数据进行网络拓扑发现得到的结果进行综合分析，提高网络拓扑发现的准确率。
本领域普通技术人员可以理解：实现上述各方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成。前述的程序可以存储于一计算机可读取存储介质中。该程序在执行时，执行包括上述各方法实施例的步骤；而前述的存储介质包括：ROM、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (10)

  1. 一种网络拓扑发现方法，其特征在于，包括：
    采集待分析网络的所有网元的网络特征数据；
    根据所述网络特征数据，利用至少两种拓扑发现算法分别获取对应的至少两个链路子集，将所述至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与所述链路子集对应的拓扑发现算法的可信度值相同，所述不同拓扑发现算法的可信度值不同，所述链路为不同网元的两个端口组成的链路；
    通过对所述第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路；
    根据所述第二链路集中的每个链路获取所述待分析网络的网络拓扑。
  2. 根据权利要求1所述的方法，其特征在于，所述合并相同的链路包括：
    合并所述第一链路集中的相同的链路，并根据所述相同的链路的多个可信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值。
  3. 根据权利要求1或2所述的方法，所述操作还包括：在所述保留所述至少两个链路中的可信度值最大的链路并删除其余的链路之后，
    将所述第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路。
  4. 根据权利要求2所述的方法，其特征在于，所述不确定性推理算法包括：
    Figure PCTCN2015086155-appb-100001
    其中，CFi(H)为所述相同的链路的多个可信度值中的一个可信度值，CFj(H)为所述相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H) 为根据CFi(H)与CFj(H)计算的所述相同的链路的新的可信度值。
  5. 根据权利要求1至4任一所述的方法，其特征在于，所述网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现协议LLDP链路算法。
  6. 一种网络拓扑发现设备，其特征在于，包括：
    采集单元，用于采集待分析网络的所有网元的网络特征数据；
    获取链路单元，用于根据所述网络特征数据，利用至少两种拓扑发现算法分别获取种对应的至少两个链路子集，将所述至少两个链路子集中的所有链路集中到一个集合中得到第一链路集；其中，各链路子集中的链路的可信度值与所述链路子集对应的拓扑发现算法的可信度值相同，所述不同拓扑发现算法的可信度值不同，所述链路为不同网元的两个端口组成的链路；
    链路处理单元，用于通过对所述第一链路集执行操作，得到第二链路集，所述操作包括：合并相同的链路，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路,其中，所述相同的链路为包含的两个端口均相同的至少两个链路；
    获取拓扑单元，用于根据所述第二链路集中的每个链路获取所述待分析网络的网络拓扑。
  7. 根据权利要求6所述的设备，其特征在于，所述链路处理单元具体用于：
    通过对所述第一链路集执行操作，得到所述第二链路集，所述操作包括：合并所述第一链路集中的相同的链路并根据所述相同的链路的多个可信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路。
  8. 根据权利要求6所述的设备，其特征在于，所述链路处理单元具体用于：
    通过对所述第一链路集执行操作，得到所述第二链路集，所述操作包括：合并所述第一链路集中的相同的链路并根据所述相同的链路的多个可 信度值以及不确定性推理算法计算所述合并后保留的链路的可信度值，并且对于仅有一个端口相同的至少两个链路，保留所述至少两个链路中的可信度值最大的链路并删除其余的链路，将所述第一链路集中的链路的可信度值与预设阈值比较，挑选出可信度值大于预设阈值的链路，其中，所述相同的链路为包含的两个端口均相同的至少两个链路。
  9. 根据权利要求7所述的设备，其特征在于，所述不确定性推理算法包括：
    Figure PCTCN2015086155-appb-100002
    其中，CFi(H)为所述相同的链路的多个可信度值中的一个可信度值，CFj(H)为所述相同的链路的多个可信度值中的另外一个可信度值，CFi,j(H)为根据CFi(H)与CFj(H)计算所述相同的链路的新的可信度值。
  10. 根据权利要求6至9任一所述的设备，其特征在于，所述网络特征数据和对应的拓扑发现算法包括以下组合中的至少两种：端口互联网协议IP地址和互联网协议IP地址匹配算法、端口别名和端口别名匹配算法、端口链路层发现协议LLDP邻居信息和端口链路层发现协议LLDP链路算法。
PCT/CN2015/086155 2014-08-12 2015-08-05 网络拓扑发现方法和设备 WO2016023438A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
ES15831718T ES2745541T3 (es) 2014-08-12 2015-08-05 Método y dispositivo para descubrir la topología de una red
EP15831718.0A EP3182656B1 (en) 2014-08-12 2015-08-05 Method and device for discovering a network topology
US15/429,946 US10116516B2 (en) 2014-08-12 2017-02-10 Network topology discovery method and device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410395692.4A CN104125154B (zh) 2014-08-12 2014-08-12 网络拓扑发现方法和设备
CN201410395692.4 2014-08-12

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/429,946 Continuation US10116516B2 (en) 2014-08-12 2017-02-10 Network topology discovery method and device

Publications (1)

Publication Number Publication Date
WO2016023438A1 true WO2016023438A1 (zh) 2016-02-18

Family

ID=51770422

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/086155 WO2016023438A1 (zh) 2014-08-12 2015-08-05 网络拓扑发现方法和设备

Country Status (5)

Country Link
US (1) US10116516B2 (zh)
EP (1) EP3182656B1 (zh)
CN (1) CN104125154B (zh)
ES (1) ES2745541T3 (zh)
WO (1) WO2016023438A1 (zh)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104125154B (zh) 2014-08-12 2017-09-26 华为技术有限公司 网络拓扑发现方法和设备
JP6700674B2 (ja) * 2015-05-21 2020-05-27 三菱重工サーマルシステムズ株式会社 電動圧縮機用モータハウジングおよびそれを用いた車載用電動圧縮機
CN105072039A (zh) * 2015-07-31 2015-11-18 山东蚁巡网络科技有限公司 一种链路层网络拓扑发现方法
EP3552353B1 (en) 2016-12-09 2024-04-10 Robert Bosch GmbH Computer network and method for running a computer network
US10447553B2 (en) * 2017-05-05 2019-10-15 Servicenow, Inc. Systems and methods for service-aware mapping of a system infrastructure
US10425281B2 (en) * 2017-11-10 2019-09-24 Cisco Technology, Inc. Automated network entity replacement based on historical topology consciousness
CN109150640B (zh) * 2018-11-08 2021-07-09 中电莱斯信息系统有限公司 一种基于二层网络协议的网络拓扑发现方法及系统
CN109842520B (zh) * 2018-12-27 2021-06-22 华为技术有限公司 网络拓扑的确定方法、装置及系统
CN110149225B (zh) * 2019-05-13 2022-01-21 中国联合网络通信集团有限公司 网络拓扑结构评估方法及装置
US11262990B2 (en) 2020-05-26 2022-03-01 International Business Machines Corporation Application topology discovery
CN112671567B (zh) * 2020-12-16 2023-02-21 中盈优创资讯科技有限公司 一种基于服务化接口的5g核心网拓扑发现方法及装置
CN113364628A (zh) * 2021-06-11 2021-09-07 上海中通吉网络技术有限公司 服务器与交换机拓扑关系建立方法及设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020165957A1 (en) * 2001-05-02 2002-11-07 Devoe Jiva Gandhara Intelligent dynamic route selection based on active probing of network operational characteristics
US20080170513A1 (en) * 2007-01-12 2008-07-17 Samsung Electronics Co., Ltd Group communication in a mobile AD-HOC network
WO2009118050A1 (en) * 2008-03-28 2009-10-01 Telefonaktiebolaget Lm Ericsson (Publ) End-to-end inter-domain routing
CN101874388A (zh) * 2007-10-31 2010-10-27 爱立信电话股份有限公司 在节点之间具有多个路径的网络和用于这样的网络的节点
CN103828310A (zh) * 2013-11-29 2014-05-28 华为技术有限公司 一种链路发现的方法、系统及设备
CN104125154A (zh) * 2014-08-12 2014-10-29 华为技术有限公司 网络拓扑发现方法和设备

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6646989B1 (en) * 1998-06-29 2003-11-11 Lucent Technologies Inc. Hop-by-hop routing with node-dependent topology information
GB2374756B (en) * 2001-04-20 2004-07-28 Discreet Logic Inc Image processing
EP1387527A1 (en) * 2002-07-30 2004-02-04 Agilent Technologies Inc. Identifying network routers and paths
US7535911B2 (en) * 2003-05-27 2009-05-19 Alcatel-Lucent Usa Inc. System and method for determining the physical topology of a network having multiple subnets
US20060256733A1 (en) * 2005-05-12 2006-11-16 Yigal Bejerano Methods and devices for discovering the topology of large multi-subnet LANs
CN1921419A (zh) 2005-08-25 2007-02-28 英业达股份有限公司 网络物理结构的拓扑方法
US20090116404A1 (en) * 2007-11-01 2009-05-07 Telefonaktiebolaget Lm Ericsson (Publ) Topology discovery in heterogeneous networks
US8493849B2 (en) * 2009-02-13 2013-07-23 Miraveo, Inc. Systems and methods for creating, managing and communicating users and applications on spontaneous area networks
CN101640612B (zh) 2009-09-07 2011-12-07 杭州华三通信技术有限公司 一种流量路径发现与故障快速定位的方法及装置
FR2954644A1 (fr) * 2009-12-21 2011-06-24 Thales Sa Protocole de routage fiabilise
CN101917310B (zh) 2010-08-27 2012-05-23 电子科技大学 基于流量分析的网络拓扑推测方法
US8443065B1 (en) * 2010-11-08 2013-05-14 Adtran, Inc. System and method for locating, identifying and provisioning newly deployed network devices
US8774050B2 (en) * 2010-11-09 2014-07-08 Cisco Technology, Inc. Dynamic wake-up time adjustment based on designated paths through a computer network
US8578034B2 (en) * 2010-11-24 2013-11-05 Verizon Patent And Licensing Inc. Optimized network device discovery
US9705747B1 (en) * 2012-12-04 2017-07-11 Qualcomm Incorporated Distributed path selection in hybrid networks
US9749214B2 (en) * 2014-02-26 2017-08-29 Futurewei Technologies, Inc. Software defined networking (SDN) specific topology information discovery
CN104125153B (zh) * 2014-08-12 2018-04-27 华为技术有限公司 网络拓扑发现方法和设备

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020165957A1 (en) * 2001-05-02 2002-11-07 Devoe Jiva Gandhara Intelligent dynamic route selection based on active probing of network operational characteristics
US20080170513A1 (en) * 2007-01-12 2008-07-17 Samsung Electronics Co., Ltd Group communication in a mobile AD-HOC network
CN101874388A (zh) * 2007-10-31 2010-10-27 爱立信电话股份有限公司 在节点之间具有多个路径的网络和用于这样的网络的节点
WO2009118050A1 (en) * 2008-03-28 2009-10-01 Telefonaktiebolaget Lm Ericsson (Publ) End-to-end inter-domain routing
CN103828310A (zh) * 2013-11-29 2014-05-28 华为技术有限公司 一种链路发现的方法、系统及设备
CN104125154A (zh) * 2014-08-12 2014-10-29 华为技术有限公司 网络拓扑发现方法和设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3182656A4 *

Also Published As

Publication number Publication date
US10116516B2 (en) 2018-10-30
EP3182656A1 (en) 2017-06-21
CN104125154A (zh) 2014-10-29
ES2745541T3 (es) 2020-03-02
EP3182656B1 (en) 2019-06-26
EP3182656A4 (en) 2017-08-23
US20170155556A1 (en) 2017-06-01
CN104125154B (zh) 2017-09-26

Similar Documents

Publication Publication Date Title
WO2016023438A1 (zh) 网络拓扑发现方法和设备
Chaudhary et al. Optimized big data management across multi-cloud data centers: Software-defined-network-based analysis
US9203743B2 (en) Packet forwarding system, control device, forwarding device and method and program for preparing processing rules
WO2016023437A1 (zh) 网络拓扑发现方法和设备
US20160294632A1 (en) Using spanning tree protocol to determine a layer 2 topology of an ethernet type network
CN101510841B (zh) 端到端流量识别方法和系统
US10880178B2 (en) Automatic assignment of coordinates to network elements interconnected in a cartesian topology
CN106576065B (zh) 将互联网协议无线接入网自动拆分为内部网关协议区域
CN107786449B (zh) 基于fsr协议的路径选择方法、装置、服务器和存储介质
US10560367B2 (en) Bidirectional constrained path search
CN108632235A (zh) 一种网包分类决策树建立方法及装置
CN105429879A (zh) 流表项查询方法、设备及系统
CN106982164A (zh) 一种网络拓扑发现方法及设备
CN109088756B (zh) 一种基于网络设备识别的网络拓扑补全方法
CN107809346A (zh) 路由追踪节点和相关联的设备的标识
CN114401516A (zh) 一种基于虚拟网络流量分析的5g切片网络异常检测方法
Kardes et al. Graph based induction of unresponsive routers in internet topologies
CN113411766A (zh) 智慧物联网综合感知系统及方法
CN109218184B (zh) 基于端口和结构信息的路由器归属as识别方法
CN114866301B (zh) 基于直推图的加密流量识别与分类方法及系统
CN106161255A (zh) 一种改进型路由器别名过滤方法
Meena et al. Instant detection of host in SDN (IDH-SDN)
Baltra et al. Ingress point spreading: A new primitive for adaptive active network mapping
CN106850336A (zh) 一种监控系统的数据流合并方法及服务端
Grailet et al. Virtual insanity: linear subnet discovery

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: 15831718

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

REEP Request for entry into the european phase

Ref document number: 2015831718

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2015831718

Country of ref document: EP