CN116647442A - A Fault Location Method for IPTV Network - Google Patents

Abstract

The embodiment of the disclosure provides an IPTV network fault positioning method, which belongs to the technical field of electric communication and specifically comprises the following steps: defining a target network as an undirected connectivity graph, defining a detection range of a group of probes at the position of a terminal node, defining a load of each node in the target network, and defining a profit function of each node; iteratively selecting the position of the maximum benefit in each node as a starting probe to form a node position set; marking the node type on the detection path according to each start probe end-to-end measurement result in the node position set, and generating a detection result set; generating a new detection path set for the suspected node set and selecting a proper detection path until the detection path set is empty or all suspected nodes are judged, adding the newly detected fault nodes into the fault node set, and generating a fault positioning result. By the scheme, the detection coverage rate and the node fault positioning success rate are improved, and the fault positioning cost is reduced.

Description

A Fault Location Method for IPTV Network

technical field

The embodiments of the present disclosure relate to the technical field of electrical communication, and in particular, to a method for locating an IPTV network fault.

Background technique

At present, the complete IPTV fault detection, location, and repair process is actually a process of repeatedly collecting information, trying to locate faults, and troubleshooting until the fault is completely repaired. However, due to the complexity of the IPTV network, there are various reasons for faults, resulting in low network fault location accuracy and location efficiency. In addition, due to the clutter and redundancy of collected information, it is difficult for operation and maintenance personnel to locate faults in the cluttered information. Even if a fault diagnosis conclusion is obtained through some fault information, it can only be regarded as the probability information of the fault, and the information collected later needs to be gradually eliminated before making further diagnosis decisions and adjustments. Currently, IPTV network fault detection and location mainly have the following problems:

(1) The traditional manual fault location method has problems such as high cost. From the content provider's server to the IPTV set-top box, it needs to go through multiple devices and lines. If any device or line fails, it will affect the stability of the service. Manual fault location analysis relies on personal experience. Usually, operation and maintenance personnel come to the site to detect faults, coordinate and obtain data from different network nodes and departments, and then analyze the data and repeatedly test the cause of the fault. This will consume a lot of human resources and coordinate communication costs. Also very high.

(2) Traditional fault location methods based on artificial intelligence, graph theory, and large data sets have problems such as high location cost, low location success rate, and small detection coverage. There are tens of thousands of devices in the network, and too many probes are started at the same time for network performance detection, resulting in the system storing a large amount of status information, log data, alarm information, etc. of network devices; excessive data volume will lead to waste of storage space The network load may reduce the accuracy of fault location; a large amount of meaningless feature data causes data redundancy, and data processing is also very expensive, which increases the cost of fault location; in addition, network dynamics and scalability will also This leads to the uncertainty of model performance, and these models have problems such as inconsistent parameters and difficulty in obtaining and updating.

It can be seen that there is an urgent need for an IPTV network fault location method that can improve detection coverage, success rate, and reduce fault location costs.

Contents of the invention

In view of this, an embodiment of the present disclosure provides an IPTV network fault location method, which at least partially solves the problems of low location success rate and high location cost in the prior art.

An embodiment of the present disclosure provides an IPTV network fault location method, including:

Step 1, define the target network as an undirected connected graph, define the detection range of a group of probes where the terminal nodes are located, define the load of each node in the target network, and define the revenue function of each node;

Step 2, according to the undirected connected graph, the detection range, the load of each node and the income function, iteratively select the position of the maximum income in each node as the starting probe to form a node position set;

Step 3: According to the end-to-end measurement results of each starting probe in the node location set, mark the node types on the detection path and generate a detection result set, where the detection result set includes a normal node set, a faulty node set, a suspected node set and Unknown set of nodes;

Step 4: Generate a new detection path set for the suspected node set and select an appropriate detection path until the detection path set is empty or all suspected nodes are determined and the newly detected faulty node is added to the faulty node set to generate a fault location result.

According to a specific implementation manner of an embodiment of the present disclosure, the step 1 specifically includes:

Step 1.1, define an undirected connected graph G(L, V, E) according to the topological structure of the target network, where L represents the set of the lowest terminal leaf nodes, V represents the set of intermediate nodes except the terminal nodes in the graph, and E represents A collection of adjacent two-level node relationship paths;

Step 1.2, define the detection path DP _i of the probe at the location of the terminal node i, and accordingly define the sum of the nodes that the probe’s detection path passes and its associated isolated nodes, and then combine all The union of the sets of detection ranges serves as the detection range DR of a group of probes;

Step 1.3, take the node i as the root node, and take the ratio of the number of leaf nodes it contains to the total number of nodes N in the network as its load Wi;

Step 1.4, define the revenue function of terminal node i as

Among them, DP _i represents the detection path of the probe at the location of node i, IN _j represents the isolated node on the detection path, DR represents the detection range of a group of probes, and the denominator represents the load sum of the detection path.

According to a specific implementation manner of an embodiment of the present disclosure, the step 2 specifically includes:

Step 2.1, according to the network topology G, traverse all intermediate nodes, and calculate the intermediate node load W _i .

Step 2.2, traverse all terminal leaf nodes, calculate the detection path DP _i of each candidate probe position, calculate the isolated node IN _j on the detection path and the load sum of the detection path according to these detection paths, and obtain the detection range DR of the probe , if there are multiple paths from the probe to the root node, take the path with the smallest node load sum on the path.

Step 2.3, the algorithm loops the preset number of times, each time the position with the highest current income is selected to start the probe, that is, the algorithm iteratively calculates the income value V _i of each leaf node and selects the node with the highest income, and every time the position of the probe is selected, it is added to After starting the probe position set PS, delete it from the silent probe set, and then recalculate the revenue value of the remaining terminal nodes until the number of selected probes is the same as the preset number of times to obtain the node position set.

According to a specific implementation manner of an embodiment of the present disclosure, the step 3 specifically includes:

According to the collection of node locations, the server side activates the start-up probes at the positions and sends distributed active detection tasks at the same time. The activated probes actively connect to the server for network performance measurement and media quality monitoring, and feed back the data to the server side. The server analyzes and processes the information fed back by the probe, and generates fault detection results based on the network topology;

Determine the type of each node according to the fault detection results;

If the fault detection result is normal, add the nodes on the detection path to the normal node set;

If the fault detection result is abnormal, add the nodes on the detection path to the suspected node set;

If there is only one suspected node, it is determined to be a faulty node and added to the faulty node set;

If the suspected node cannot be reached from any normal node, it is determined to be an unknown node and added to the unknown node set;

If a node is not passed by any detection path, it is determined to be an unknown node and added to the unknown node set.

According to a specific implementation manner of an embodiment of the present disclosure, the step 4 specifically includes:

When the current suspected node set is not empty, traverse the suspected node set, calculate the detection path from the root node to each suspected node, and generate a candidate detection path set, and calculate the weight w(p) of each candidate path;

Select the candidate path with the highest weight to detect, if the detection is successful, add the suspected node to the normal node;

If the detection fails and the suspected node of the path is 1, the suspected node is judged as a faulty node. If the suspected node of the path is not 1, the end node of the path is changed to an unknown node and marked;

If the upper node of the marked node is a normal node, change the marked node to a faulty node;

After each predetermined number of faulty nodes are determined, they are added to the faulty node set, and candidate detection paths are regenerated and determined for the remaining suspected nodes, and the updated states of each node are output to form fault location results.

According to a specific implementation of an embodiment of the present disclosure, the calculation formula of the weight value is

Among them, s(p) represents the number of suspected nodes on the detection path p, E(p) represents the set of all links on the path p, and n(l) is the number of times the selected detection path passes through the path.

The IPTV network fault location solution in the embodiment of the present disclosure includes: step 1, defining the target network as an undirected connected graph, defining the detection range of a group of probes where the terminal nodes are located, and defining the load of each node in the target network , and define the revenue function of each node; step 2, according to the undirected connected graph, the detection range, the load and revenue function of each node, iteratively select the position of the maximum revenue in each node as the starting probe, and form the node position Set; step 3, according to the end-to-end measurement results of each startup probe in the node position set, mark the node type on the detection path, and generate a detection result set, where the detection result set includes a normal node set, a faulty node set, a suspected node set set and unknown node set; step 4, generate a new detection path set for the suspected node set and select an appropriate detection path until the detection path set is empty or all suspected nodes are determined and the newly detected faulty node is added to the faulty node set, Generate fault location results.

The beneficial effects of the embodiments of the present disclosure are: through the scheme of the present disclosure, the startup probe selection method uses the probe detection range, combined with the network topology node weights, calculates the income of each probe node, and iteratively selects the terminal node with the largest income to start all Deploy the probe software and issue measurement tasks. According to the fault detection results of the probe, the node types on the detection path are firstly classified, and then the optimal detection path is iteratively selected by an adaptive faulty node determination method, and the fault determination of suspected nodes is carried out. The fault location method can improve the success rate of node fault location and reduce the cost of fault location while ensuring the detection coverage.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present disclosure. Those of ordinary skill in the art can also obtain other drawings based on these drawings without any creative effort.

FIG. 1 is a schematic flow diagram of an IPTV network fault location method provided by an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of a probe detection range at the location of a node 12 provided by an embodiment of the present disclosure;

FIG. 3 is a diagram of the detection range after starting the soft probes at the positions of nodes 7 and 12 provided by an embodiment of the present disclosure;

FIG. 4 is a diagram of marking node types according to fault detection results provided by an embodiment of the present disclosure;

FIG. 5 is a suspected node determination diagram provided by an embodiment of the present disclosure.

Detailed ways

Embodiments of the present disclosure will be described in detail below in conjunction with the accompanying drawings.

Embodiments of the present disclosure are described below through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Apparently, the described embodiments are only some of the embodiments of the present disclosure, not all of them. The present disclosure can also be implemented or applied through different specific implementation modes, and various modifications or changes can be made to the details in this specification based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, in the case of no conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in the present disclosure, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present disclosure.

It is noted that the following describes various aspects of the embodiments that are within the scope of the appended claims. It should be apparent that the aspects described herein may be embodied in a wide variety of forms and that any specific structure and/or function described herein is illustrative only. Based on the present disclosure one skilled in the art should appreciate that an aspect described herein may be implemented independently of any other aspects and that two or more of these aspects may be combined in various ways. For example, any number of the aspects set forth herein can be used to implement an apparatus and/or practice a method. In addition, such an apparatus may be implemented and/or such a method practiced using other structure and/or functionality than one or more of the aspects set forth herein.

It should also be noted that the diagrams provided in the following embodiments are only schematically illustrating the basic ideas of the present disclosure, and only the components related to the present disclosure are shown in the drawings rather than the number, shape and shape of the components in actual implementation. Dimensional drawing, the type, quantity and proportion of each component can be changed arbitrarily during actual implementation, and the component layout type may also be more complicated.

Additionally, in the following description, specific details are provided to facilitate a thorough understanding of examples. However, it will be understood by those skilled in the art that the described aspects may be practiced without these specific details.

An embodiment of the present disclosure provides a method for locating a fault in an IPTV network, and the method may be applied in a fault diagnosis process of an IPTV network in a telecommunication scene.

Referring to FIG. 1 , it is a schematic flowchart of an IPTV network fault location method provided by an embodiment of the present disclosure. As shown in Figure 1, the method mainly includes the following steps:

Step 1, define the target network as an undirected connected graph, define the detection range of a group of probes where the terminal nodes are located, define the load of each node in the target network, and define the revenue function of each node;

Further, the step 1 specifically includes:

Step 1.1, define an undirected connected graph G(L, V, E) according to the topological structure of the target network, where L represents the set of the lowest terminal leaf nodes, V represents the set of intermediate nodes except the terminal nodes in the graph, and E represents A collection of adjacent two-level node relationship paths;

Step 1.2, define the detection path DP _i of the probe at the location of the terminal node i, and accordingly define the sum of the nodes that the probe’s detection path passes and its associated isolated nodes, and then combine all The union of the sets of detection ranges serves as the detection range DR of a group of probes;

Step 1.3, take the node i as the root node, and take the ratio of the number of leaf nodes it contains to the total number of nodes N in the network as its load Wi;

Step 1.4, define the revenue function of terminal node i as

Among them, DP _i represents the detection path of the probe at the location of node i, IN _j represents the isolated node on the detection path, DR represents the detection range of a group of probes, and the denominator represents the load sum of the detection path.

The fault location of this method mainly includes the steps of probe selection, fault detection, node marking and suspected node judgment. Given an IPTV layered network topology, that is, known network node locations and connection modes between nodes. The method aims at reducing the detection load and increasing the detection revenue, and dynamically selects and deploys soft probes on the end user's set-top box and issues detection tasks. The soft probe uploads the measurement results to the server. The server uses the collected probe data and associated information for analysis, processing and fault detection, and marks different types of nodes in combination with the network topology, then regenerates candidate detection paths for suspected nodes generated by detection results, and selects appropriate paths to detect suspected nodes Judgment, and finally achieve fault location.

In specific implementation, define a hierarchical network represented by an undirected connected graph G(L, V, E), where L represents the set of bottom-level terminal leaf nodes, terminal leaf nodes are deployed with soft probes, and V represents the removal of terminal nodes in the graph The set of intermediate nodes generally represents network intermediate node devices, E represents the set of adjacent two-layer node relationship paths, and the root node generally represents the server location where the probe management platform is located. When the probe is deployed at the bottom end node in the network, the probe can transmit data along the network to detect the status of nodes and links on the path, and can also perform fault reasoning on the isolated nodes connected to the nodes on the detection path.

Define the detection path DPi (Detetion Path) of the probe at the location of the terminal node i, that is, the transmission path from the node i deployed by the probe to the root node, which is a collection of intermediate nodes passing through the hierarchical network, and includes the probe itself The node where it is deployed.

Define the associated isolated node set IN (Isolated Node) of terminal node i, that is, the detection path DP of node i passes through the set of isolated nodes associated with the node. An associated isolated node means that the only parent node is in the detection path DPi and has no child nodes. The detection range of a single probe is defined as the sum of the nodes that the probe's detection path passes and its associated isolated nodes. A group of probes PS (Probe Set) have their detection range DR (Detetion Range), that is, the union of the collections of all detection ranges of the probes.

In order to quantify the load of each node in the network, the load Wi of node i is defined as the ratio of the number of leaf nodes it contains to the total number of nodes N in the network, taking the node as the root node. Among them, the numerator of the formula means that node j takes node i as the root of the tree, and the sum of the number of leaf nodes under all subtrees. That is, when node j is a leaf node of the subtree of node i, Cij=1, otherwise Cij=0.

Define the income function Vi of the terminal node i, which represents the income of the terminal node in the network to start the probe. According to the current number of detectable nodes and the load sum of the detection path, the higher the Vi, the greater the income of the node to start the probe. The calculation method is as follows: Among them, the numerator is the node within the detection range of the current node i, except the nodes already within the detection range of the probe set, the larger the detection range of the current node, the higher the income. The denominator is the node load sum on the detection path from node i to the root node. The higher the detection path load sum, the lower the benefit of starting the probe at this node.

Step 2, according to the undirected connected graph, the detection range, the load of each node and the income function, iteratively select the position of the maximum income in each node as the starting probe to form a node position set;

On the basis of the foregoing embodiments, the step 2 specifically includes:

Step 2.1, according to the network topology G, traverse all intermediate nodes, and calculate the intermediate node load W _i .

Step 2.2, traverse all terminal leaf nodes, calculate the detection path DP _i of each candidate probe position, calculate the isolated node IN _j on the detection path and the load sum of the detection path according to these detection paths, and obtain the detection range DR of the probe , if there are multiple paths from the probe to the root node, take the path with the smallest node load sum on the path.

Step 2.3, the algorithm loops the preset number of times, each time the position with the highest current income is selected to start the probe, that is, the algorithm iteratively calculates the income value V _i of each leaf node and selects the node with the highest income, and every time the position of the probe is selected, it is added to After starting the probe position set PS, delete it from the silent probe set, and then recalculate the revenue value of the remaining terminal nodes until the number of selected probes is the same as the preset number of times to obtain the node position set.

For example, the input of the algorithm is the hierarchical network topology G, the number of probes to be activated is M, and the output of the algorithm is the set PS of node positions deployed by the soft probes to be activated.

The algorithm first traverses all intermediate nodes according to the network topology G, and calculates the intermediate node load Wi.

Secondly, traverse all terminal leaf nodes, calculate the detection path DPi of each candidate probe position, calculate the isolated node INj on the detection path according to these detection paths, and calculate the detection path load sum (the denominator part of the income Vi), and get the probe The detection range DR. If there are multiple paths from the probe to the root node, take the path with the smallest node load sum on the path.

Finally, the algorithm loops N times, each time the position with the highest current income is selected to start the probe, that is, the algorithm iteratively calculates the income value Vi of each leaf node and selects the node with the highest income, and the position of the probe is added to the start probe every time it is selected. After entering the position set, it needs to be deleted from the silent probe set, and then recalculate the revenue value of the remaining terminal nodes until the number of probes is selected as M, and the node position set is obtained. For example, the detection range of the probe at the position of node 12 is as follows Figure 2 shows.

Step 3: According to the end-to-end measurement results of each starting probe in the node location set, mark the node types on the detection path and generate a detection result set, where the detection result set includes a normal node set, a faulty node set, a suspected node set and Unknown set of nodes;

Further, the step 3 specifically includes:

According to the collection of node locations, the server side activates the start-up probes at the positions and sends distributed active detection tasks at the same time. The activated probes actively connect to the server for network performance measurement and media quality monitoring, and feed back the data to the server side. The server analyzes and processes the information fed back by the probe, and generates fault detection results based on the network topology;

Determine the type of each node according to the fault detection results;

If the fault detection result is normal, add the nodes on the detection path to the normal node set;

If the fault detection result is abnormal, add the nodes on the detection path to the suspected node set;

If there is only one suspected node, it is determined to be a faulty node and added to the faulty node set;

If the suspected node cannot be reached from any normal node, it is determined to be an unknown node and added to the unknown node set;

If a node is not passed by any detection path, it is determined to be an unknown node and added to the unknown node set.

During the specific implementation, the detection range diagram after starting the soft probes at the positions of nodes 7 and 12 is shown in Figure 3. The server can activate the soft probes at the positions according to the constructed node position set PS, and at the same time issue distributed In the active detection task, the activated probe actively connects to the server for network performance measurement and media quality monitoring, and then feeds the data back to the server. The server analyzes and processes the information fed back by the probes, and realizes rapid fault detection in combination with the network topology.

All nodes in the network are divided into normal node set NNS, faulty node set FNS, suspected node set SNS, and unknown node set UNS.

According to the fault detection results of the startup probe, if the soft probe fault detection is normal, the nodes on the detection path are added to the normal node set; if the soft probe fault detection is abnormal, the nodes on the detection path are added to the suspected node set. If there is only one suspected node, it can be judged as a faulty node. If a suspected node cannot be reached from any normal node, it is defined as an unknown node. If a node is not passed by any detection path, it is also defined as an unknown node. The process of marking the node type according to the fault detection result is shown in Figure 4, where Figure 4(a) shows that the probe detection result at the position of node 7 is normal, When the detection result of the probe at the position of node 12 is abnormal, the labeling of the network topology node type; Figure 4(b) shows that the detection result of the probe at the position of node 12 is normal, and when the detection result of the probe at the position of node 7 is abnormal, the network topology node type of the markup.

Step 4: Generate a new detection path set for the suspected node set and select an appropriate detection path until the detection path set is empty or all suspected nodes are determined and the newly detected faulty node is added to the faulty node set to generate a fault location result.

On the basis of the foregoing embodiments, the step 4 specifically includes:

When the current suspected node set is not empty, traverse the suspected node set, calculate the detection path from the root node to each suspected node, and generate a candidate detection path set, and calculate the weight w(p) of each candidate path;

Select the candidate path with the highest weight to detect, if the detection is successful, add the suspected node to the normal node;

If the detection fails and the suspected node of the path is 1, the suspected node is judged as a faulty node. If the suspected node of the path is not 1, the end node of the path is changed to an unknown node and marked;

If the upper node of the marked node is a normal node, change the marked node to a faulty node;

After each predetermined number of faulty nodes are determined, they are added to the faulty node set, and candidate detection paths are regenerated and determined for the remaining suspected nodes, and the updated states of each node are output to form fault location results.

Further, the calculation formula of the weight value is

Among them, s(p) represents the number of suspected nodes on the detection path p, E(p) represents the set of all links on the path p, and n(l) is the number of times the selected detection path passes through the path.

During specific implementation, as shown in Figure 5, wherein, Figure 5 (a) shows that after the second failure judgment is carried out to the suspected fault node 5, when the detection result is normal, the network topology node marking situation; After the suspected faulty node 5 makes a secondary fault judgment, when the detection result is abnormal, the network topology node marking situation. This step performs secondary fault judgment on the suspected node as much as possible. There are many paths from the server to the suspected node in the network topology, and there are multiple suspected nodes. Therefore, it is necessary to select the optimal path from the candidate paths for fault judgment. Secondary fault judgment and optimal path selection are used to improve the success rate of fault location and reduce the cost of location.

Define the weight value w(p) of the candidate detection path p. For the selected detection path, if the path is passed more times, the weight of selecting the path is lower. The more suspected nodes the detection path passes through, the higher the weight. The calculation is shown in the following formula. Among them, s(p) represents the number of suspected nodes on the detection path p. E(p) represents the set of all links on the path p, and n(l) is the number of times the selected detection path passes through the path.

The faulty node determination method needs to generate a new detection path set DPS for the remaining suspected nodes and select an appropriate detection path until the detection path set is empty or all suspected nodes are determined. The input of the algorithm is the result of node marking in step 3, that is, the set of normal nodes NNS, the set of faulty nodes FNS, the set of suspected nodes SNS, the set of unknown nodes UNS, and the output of the algorithm is the updated sets.

At the beginning of the algorithm, it is necessary to determine whether the size of the current suspected node set SNS is empty. If not empty, the algorithm first traverses the suspected node set SNS, calculates the detection path from the root node to each suspected node SNS, and generates a candidate detection path set DPS, and calculates the weight w(p) of each candidate path.

If the current set of candidate detection paths is empty, determine the SNS node in the suspected node set as an unknown node, add it to the UNS, and end the loop ahead of time, and return the result.

Otherwise, the algorithm iteratively selects the highest weight candidate path to explore. If the detection is successful, the suspected node will be added to the normal node; if the detection fails and the suspected node of the path is 1, the suspected node will be judged as a faulty node; if the suspected node of the path is not 1, the end node of the path will be changed to an unknown node and mark. If the upper node of the marked node is a normal node, change the marked node to a faulty node.

Every time a part of faulty nodes are identified, they are added to the set of faulty nodes, and candidate detection paths are regenerated for the remaining suspected nodes. Realize the judgment of the suspected node set, and output the updated status of each node. Finally, the fault location is realized and the fault location result is generated.

The IPTV network fault location method provided in this embodiment uses the probe detection range by starting the probe selection method, and combines the network topology node weights to calculate the income of each probe node, and iteratively selects the terminal node with the largest income to start the deployment. Probe software and issue measurement tasks. According to the fault detection results of the probe, the node types on the detection path are firstly classified, and then the optimal detection path is iteratively selected by an adaptive faulty node determination method, and the fault determination of suspected nodes is carried out. The fault location method can improve the success rate of node fault location and reduce the cost of fault location while ensuring the detection coverage.

The units involved in the embodiments described in the present disclosure may be implemented by software or by hardware.

It should be understood that various parts of the present disclosure may be implemented in hardware, software, firmware or a combination thereof.

The above is only a specific implementation of the present disclosure, but the scope of protection of the present disclosure is not limited thereto, any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present disclosure, should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be determined by the protection scope of the claims.

Claims (6)

1. An IPTV network failure location method, comprising:

step 1, defining a target network as an undirected communication graph, defining a detection range of a group of probes at the position of a terminal node, defining the load of each node in the target network, and defining a benefit function of each node;

step 2, iteratively selecting the position of the maximum benefit in each node as a starting probe according to the undirected communication graph, the detection range, the load of each node and the benefit function, and forming a node position set;

marking node types on a detection path according to each start probe end-to-end measurement result in the node position set to generate a detection result set, wherein the detection result set comprises a normal node set, a fault node set, a suspected node set and an unknown node set;

and 4, generating a new detection path set for the suspected node set and selecting a proper detection path until the detection path set is empty or all suspected nodes are judged, adding the newly detected fault nodes into the fault node set, and generating a fault positioning result.

2. The method according to claim 1, wherein the step 1 specifically comprises:

step 1.1, defining an undirected communication graph G (L, V, E) according to a topological structure of a target network, wherein L represents a terminal leaf node set at the bottommost layer, V represents an intermediate node set except terminal nodes in the graph, and E represents a node relation path set at two adjacent layers;

step 1.2, defining a probe path DP of the probe at the position of the terminal node i _i Defining the sum of nodes passed by the detection path of the probe and the associated isolated nodes, and taking the union of all detection ranges of a group of probes as the detection range DR of the group of probes;

step 1.3, taking the node i as a root node, and taking the ratio of the number of leaf nodes contained in the node i to the number N of total nodes in the network as a load Wi;

step 1.4, defining the benefit function of the terminal node i as

Wherein DP _i A detection path of the probe indicating the position of the node i, IN _j Representing isolated nodes on the probe path, DR representing the probe range of a set of probes, and denominator representing the sum of the loads of the probe paths.

3. The method according to claim 2, wherein the step 2 specifically comprises:

step 2.1, traversing all intermediate nodes according to the network topology structure G, and calculating the load W of the intermediate nodes _i ；

Step 2.2, traversing all terminal leaf nodes, and calculating the detection path DP of each candidate probe position _i From these probe paths, an isolated node IN on the probe path is calculated _j The path load sum is detected, the detection range DR of the probe is obtained, and if a plurality of paths exist from the probe to the root node, the node load sum of the paths is taken;

step 2.3, the algorithm loops for preset times, and the position with the highest current benefit is selected each time to start the probe, namely the algorithm iteratively calculates the benefit value V of each leaf node _i And selecting the node with the biggest profit, deleting the probe from the silent probe set after the probe position set PS is added to the start probe position set once, and then recalculating the profit value of the rest terminal nodes until the number of the selected probes is the same as the preset times, so as to obtain the node position set.

4. A method according to claim 3, wherein said step 3 comprises:

the server activates a starting probe at a position according to the node position set, simultaneously issues a distributed active detection task, the activated and started probe is actively connected with the server to perform network performance measurement and media quality monitoring, data are fed back to the server, the server analyzes and processes information fed back by the probe, and a fault detection result is generated by combining a network topology structure;

judging the type of each node according to the fault detection result;

if the fault detection result is normal, adding the nodes on the detection path into a normal node set;

if the fault detection result is abnormal, adding the nodes on the detection path into a suspected node set;

if only one suspected node exists, judging that the suspected node is a fault node and adding the fault node into a fault node set;

if the suspected node which can not be reached from any normal node is judged to be an unknown node and is added into the unknown node set;

if a node is not passed by any detection path, the node is judged to be an unknown node and added into the unknown node set.

5. The method according to claim 4, wherein the step 4 specifically includes:

when the current suspected node set is not empty, traversing the suspected node set, calculating the detection path from the root node to each suspected node, generating a candidate detection path set, and calculating the weight w (p) of each candidate path;

selecting the candidate path with the highest weight for detection, and if the detection is successful, adding the suspected node into the normal node;

if the detection fails and the suspected node of the path is 1, judging the suspected node as a fault node, and if the suspected node of the path is not 1, changing the end node of the path into an unknown node and marking;

if the upper node of the marked node is a normal node, changing the marked node into a fault node;

after each time the preset number of fault nodes are determined, the fault nodes are added into the fault node set, candidate detection paths are regenerated for the rest suspected nodes, the candidate detection paths are judged, and the updated states of all the nodes are output to form a fault positioning result.

6. The method according to claim 5, wherein the weight value is calculated by the formula

Where s (p) represents the number of suspected nodes on the probe path p, E (p) represents all link sets on the path p, and n (l) is the number of times the selected probe path passes through the path.