A communication network fault diagnosis method and diagnosis device

Abstract

The application provides a communication fault diagnosis method and a diagnosis device, relates to the technical field of communication networks, and is characterized in that the method is distributedly arranged with core layer, convergence layer and access layer diagnosis nodes in a communication network according to a topological hierarchy, the nodes exchange real-time data through a time-sensitive network to form a collaborative sensing network; the collaborative sensing network is used for dynamic collection of multi-modal data, a sound-light-protocol correlation matrix is constructed, and the correlation between signal distortion of a physical layer, equipment vibration and protocol layer abnormalities is mapped; features of the collected multi-modal data are extracted, a mapping relationship between data sending nodes and receiving nodes is generated, a minimum coverage detection path is generated based on current network load and a topological structure; a super-diffusion fault propagation model is constructed, a root cause node of the minimum coverage detection path is located through gradient reverse tracing, a fault severity parameter is calculated, a hierarchical repair strategy is triggered and a repair instruction is issued, and the stability and reliability of the communication network are improved.

Description

Technical Field

[0001] This application relates to the field of communication network technology, specifically to a communication network fault diagnosis method and diagnostic device. Background Technology

[0002] With the rapid development of modern communication technology, communication networks have become a crucial infrastructure in all sectors of society. However, the complexity of communication networks has also increased, leading to a greater diversity of fault types and causes. Traditional fault diagnosis methods, which rely primarily on human experience and simple rule matching, are no longer sufficient to meet the needs of current communication network fault diagnosis. Especially in large-scale networks, timely fault location and repair are critical. Existing technologies face the following problems in the fault diagnosis process:

[0003] Due to the large scale of networks and the variety of devices, traditional fault diagnosis methods require a lot of time to investigate and are difficult to quickly locate the fault point. Existing fault diagnosis methods mostly rely on preset rules, which are prone to missed diagnoses and misdiagnoses, and cannot fully cover various complex fault scenarios. Traditional methods lack self-learning and adaptive capabilities, making it difficult to cope with rapidly changing network environments and newly emerging fault types. Summary of the Invention

[0004] In view of this, the purpose of this invention is to propose a communication network fault diagnosis method and device, construct an acoustic-optical-protocol multimodal sensing network, dynamically simulate fault propagation through the super-diffusion equation, and realize "diagnosis-repair" closed-loop control by combining digital twins, which can significantly improve the stability and reliability of the communication network.

[0005] To achieve the above objectives, the embodiments of this application provide the following technical solutions:

[0006] In view of the above objectives, in a first aspect, the present invention provides a method for diagnosing communication network faults, comprising the following steps:

[0007] In the communication network, core layer, aggregation layer, and access layer diagnostic nodes are deployed in a distributed manner according to topology. The nodes exchange data in real time through time-sensitive networks to form a collaborative sensing network.

[0008] By using a collaborative sensing network to dynamically acquire multimodal data, an acoustic-optical-protocol correlation matrix is ​​constructed to map the correlation between physical layer signal distortion, equipment vibration, and protocol layer anomalies.

[0009] Feature extraction is performed on the collected multimodal data to generate a mapping relationship between data sending nodes and receiving nodes, and a minimum coverage detection path is generated based on the current network load and topology.

[0010] A super-diffusion fault propagation model is constructed. The root cause node of the minimum coverage detection path is located by gradient back-source tracing. The fault severity parameter is calculated. Based on the fault severity parameter, a graded repair strategy is triggered and a repair command is issued.

[0011] As a further embodiment of the present invention, in the cooperative sensing network, the core layer nodes are used to monitor the stability of high-bandwidth traffic and the status of key routing nodes; the aggregation layer nodes are used to monitor the communication quality and traffic balance between subnets; and the access layer nodes are used to monitor the stability of terminal connections and user experience indicators.

[0012] As a further aspect of the present invention, the construction of the cooperative sensing network further includes the following steps:

[0013] Based on the network topology diagram, select the core routers and switches in the network as core layer nodes;

[0014] Diagnostic equipment, including high-performance data acquisition modules and processing units, is installed on selected core nodes to monitor the stability of high-bandwidth traffic and the status of critical routing nodes.

[0015] Select an aggregation switch or router that connects the core network and the access network as an aggregation layer node, and install diagnostic equipment on the aggregation node to monitor the communication quality and traffic balance between subnets.

[0016] Select access switches and wireless access point devices as access layer nodes, and install diagnostic devices on the access nodes to monitor the stability of terminal connections and user experience indicators.

[0017] As a further aspect of the present invention, multimodal data dynamic acquisition is performed using a collaborative sensing network, including: acquiring digital voltage signals reflected from optical signals, synchronously capturing TCP retransmission rate, routing oscillation counts, and network traffic fluctuation data, and acquiring the device vibration spectrum through a piezoelectric ceramic sensor array.

[0018] As a further aspect of the present invention, when collecting the reflected digital voltage signal of the optical signal, the reflected optical signal of the optical fiber is collected by a three-port circulator in the optical module and converted into a reflected digital voltage signal. When the signal exceeds a first threshold, a fault judgment is triggered.

[0019] As a further aspect of the present invention, the vibration spectrum of the device is collected by a piezoelectric ceramic sensor array to identify hidden hardware faults, including fan malfunctions and circuit board aging.

[0020] As a further aspect of the present invention, when nodes exchange data in real time through a time-sensitive network, the following is included:

[0021] Configure a high-precision clock source in the network and synchronize time using the IEEE 1588v2 protocol;

[0022] Deploy the TSN protocol stack on all diagnostic nodes to support priority queuing, time-aware shaping, and frame preprocessing. Deploy the TSN protocol stack, establish TSN streams for data transmission between diagnostic nodes, and configure traffic scheduling plans.

[0023] As a further aspect of the present invention, an acoustic-optical-protocol correlation matrix is ​​constructed, including:

[0024] The optical module collects reflected optical signals from the optical fiber using a three-port circulator, converts them into reflected digital voltage signals, and uses a piezoelectric ceramic sensor array to collect the device vibration spectrum, capturing TCP retransmission rate, routing oscillation counts, and network traffic fluctuation data.

[0025] The collected physical layer and protocol layer data are cleaned, denoised, and standardized. Feature extraction is performed on the protocol layer data to obtain statistical TCP retransmission counts, routing oscillation frequency, and traffic fluctuation amplitude.

[0026] The physical layer characteristics of optical signal reflection voltage and vibration spectrum are initially matched with protocol layer characteristics such as TCP retransmission rate, routing oscillation count, and traffic fluctuation data. Statistical analysis is then used to conduct in-depth correlation analysis to identify significant correlations.

[0027] Based on the correlation analysis results, an acoustic-optical-protocol correlation matrix is ​​constructed, where the rows of the matrix represent physical layer features, the columns represent protocol layer features, and the matrix elements represent the correlation strength between the two.

[0028] As a further aspect of the present invention, generating a minimum coverage detection path based on the current network load and topology includes:

[0029] Obtain the current network topology information, including nodes, links, and their corresponding weights;

[0030] Monitor network load in real time and use traffic analysis tools to determine the current network traffic distribution and bottleneck locations;

[0031] Based on the initial path search setting of search objectives and constraints, obtain the minimum path covering all nodes that need to be detected. The constraints of the path search include the current network load and link status.

[0032] Perform a path search on the network topology graph to find an initial set of paths that cover all nodes that need to be detected;

[0033] The initial path set is optimized to reduce redundant paths, and the generated detection paths are verified.

[0034] As a further aspect of the present invention, the construction of the super-diffusion fault propagation model includes:

[0035] Set the initial state parameters for each network node, including the node's health status, fault propagation probability, and fault propagation speed;

[0036] Based on the topology of the communication network, a fault propagation matrix is ​​constructed, which describes the connection relationship between nodes and the probability of fault propagation.

[0037] The super-diffusion factor is obtained through historical fault data and statistical analysis. The super-diffusion factor is used to characterize the rapid propagation characteristics of faults in complex networks.

[0038] By utilizing the minimum coverage detection path, multimodal data is sampled to obtain real-time status information of nodes;

[0039] Gradient calculations are performed on the sampled data to identify potential fault propagation paths. By tracing the fault propagation path in reverse, the process is gradually backtracked until the origin of the fault is found.

[0040] Set assessment indicators for the severity of the fault, perform a weighted comprehensive assessment of each indicator, and calculate the fault severity parameter.

[0041] Secondly, the present invention provides a communication network fault diagnosis device, comprising:

[0042] The node deployment module is used to deploy core layer, aggregation layer, and access layer diagnostic nodes in a distributed manner according to the topology of the communication network. The nodes exchange data in real time through a time-sensitive network to form a collaborative sensing network.

[0043] The multimodal data acquisition module is used to dynamically acquire multimodal data using a collaborative sensing network, construct an acoustic-optical-protocol correlation matrix, and map the correlation between physical layer signal distortion, equipment vibration, and protocol layer anomalies.

[0044] The path generation module is used to extract features from the collected multimodal data, generate a mapping relationship between data sending nodes and receiving nodes, and generate a minimum coverage detection path based on the current network load and topology.

[0045] The fault severity diagnosis module is used to construct an ultra-diffusion fault propagation model, locate the root cause node of the minimum coverage detection path through gradient back-source tracing, calculate the fault severity parameter, trigger a graded repair strategy based on the fault severity parameter, and issue repair instructions.

[0046] Compared with existing technologies, the communication network fault diagnosis method and device proposed in this invention have the following advantages:

[0047] This application embodiment deploys diagnostic nodes in the core layer, aggregation layer, and access layer in a distributed manner, and uses time-sensitive networking for real-time data exchange to form a collaborative sensing network. This network can accurately locate faults and reduce troubleshooting time. By using the collaborative sensing network to dynamically acquire multimodal data and construct an acoustic-optical-protocol correlation matrix, it can comprehensively analyze physical layer signals, equipment vibrations, and protocol layer data, thereby improving the accuracy and reliability of fault diagnosis.

[0048] This application embodiment constructs an over-diffusion fault propagation model and utilizes gradient reverse tracing technology to effectively identify and locate the root cause node of the minimum coverage detection path, thereby improving the efficiency and accuracy of fault diagnosis. It also extracts features from the collected multimodal data and generates a mapping relationship between data sending nodes and receiving nodes, which can dynamically adapt to changes in the network environment and improve the flexibility and adaptability of the diagnostic system.

[0049] This application's embodiments generate a minimum coverage detection path based on the current network load and topology, optimizing the fault detection path and reducing detection time and resource consumption. By calculating fault severity parameters and using evaluation indicators such as impact range, number of affected nodes, and fault duration, the impact of the fault can be comprehensively assessed, providing a scientific basis for repair strategies. Based on the fault severity parameters, a graded repair strategy is triggered, and repair instructions are issued, realizing the automation and intelligence of fault repair, improving fault handling efficiency, and reducing network downtime.

[0050] In summary, the communication network fault diagnosis method and device provided by this invention significantly improves the accuracy, efficiency, and reliability of communication network fault diagnosis through the comprehensive application of multi-level collaborative sensing networks, multi-modal data fusion, intelligent fault propagation models, and hierarchical repair strategies, and has important practical application value. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the accompanying drawings used in the description of the exemplary embodiments or related technologies will be briefly introduced below. The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation thereof. In the drawings:

[0052] Figure 1 This is a flowchart illustrating the implementation of a communication network fault diagnosis method according to an embodiment of the present invention.

[0053] Figure 2 This is a flowchart illustrating the construction of a collaborative sensing network in a communication network fault diagnosis method according to an embodiment of the present invention.

[0054] Figure 3This is a flowchart illustrating the construction of an acoustic-optical-protocol correlation matrix in a communication network fault diagnosis method according to an embodiment of the present invention.

[0055] Figure 4 This is a flowchart illustrating the generation of the minimum coverage detection path in a communication network fault diagnosis method according to an embodiment of the present invention.

[0056] Figure 5 This is a flowchart illustrating the construction of an over-diffusion fault propagation model in a communication network fault diagnosis method according to an embodiment of the present invention. Detailed Implementation

[0057] The present application will now be further described in conjunction with the accompanying drawings and specific embodiments. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0058] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to specific examples and the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the scope of this application.

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0061] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0062] This invention proposes a method and device for fault diagnosis in communication networks. It constructs an acoustic-optical-protocol multimodal sensing network, dynamically simulates fault propagation through the super-diffusion equation, and achieves closed-loop control of "diagnosis-repair" by combining digital twins, which can significantly improve the stability and reliability of communication networks.

[0063] See Figure 1 As shown, an embodiment of the present invention provides a method for diagnosing communication network faults, the method comprising the following steps:

[0064] Step S10: In the communication network, core layer, aggregation layer, and access layer diagnostic nodes are deployed in a distributed manner according to the topology hierarchy. The nodes exchange data in real time through a time-sensitive network to form a collaborative sensing network.

[0065] In this step, in the collaborative sensing network, core layer nodes are used to monitor the stability of high-bandwidth traffic and the status of key routing nodes. Specifically, core layer nodes collect traffic parameters of high-bandwidth links in real time through Time-Sensitive Networking (TSN). These traffic parameters include bandwidth utilization, transmission delay, packet loss rate, etc. By collecting traffic parameters, traffic stability is monitored, and the CPU load, memory usage, and routing table update frequency of key routing nodes are monitored. At the same time, regional network data uploaded by the aggregation layer is integrated to construct a global network status view.

[0066] Furthermore, the aggregation layer nodes of the collaborative sensing network are used to monitor the communication quality and traffic balance between subnets. These aggregation layer nodes achieve real-time data interaction between subnets through the Time-Sensitive Networking (TSN) protocol, monitor key communication quality indicators and traffic balance status, optimize data transmission paths between subnets using dynamic routing adjustment and load-sharing algorithms, and encrypt and upload subnet status data to the core layer nodes. The key indicators include latency jitter, packet loss rate, and bandwidth utilization.

[0067] Furthermore, the access layer nodes of the collaborative sensing network are used to monitor terminal connection stability and user experience metrics. These access layer nodes monitor terminal device connection stability metrics and key user experience parameters in real time using Time-Sensitive Networking (TSN), and further employ adaptive frequency hopping and power adjustment algorithms to optimize terminal access quality. Terminal status data is then encrypted and uploaded to the aggregation layer nodes, providing raw terminal-side data support for subnet-level traffic balancing and global network status assessment. The connection stability metrics include signal strength, connection establishment success rate, handover latency, and disconnection frequency. The key user experience parameters include application layer response time, data transmission throughput, and service availability.

[0068] Among them, see Figure 2 As shown, the embodiments of this application further include the following steps when constructing a cooperative sensing network:

[0069] Step S101: Based on the network topology diagram, select the core routers and switches in the network as core layer nodes;

[0070] In step S101 of this embodiment, based on the critical path location and data throughput requirements of the core layer nodes in the network topology diagram, high-performance core routers and Layer 3 switches with Time Sensitive Networking (TSN) protocol support are selected as core layer nodes.

[0071] Step S102: Install diagnostic equipment, including a high-performance data acquisition module and processing unit, on the selected core nodes to monitor the stability of high-bandwidth traffic and the status of key routing nodes.

[0072] Specifically, in step S102 of this embodiment, a diagnostic device is integrated on the selected core node, which is configured to include a 10Gbps optical port data acquisition module, a multi-core parallel processing unit and a TSN time synchronization module. The device captures parameters such as bandwidth utilization, transmission delay and packet loss rate of high-bandwidth traffic in real time through the mirror port. At the same time, a routing node status monitoring unit is embedded to collect CPU load, memory usage and routing table update frequency. The data sampling frequency is not less than 1kHz to meet the real-time requirements.

[0073] Step S103: Select the aggregation switch or router that connects the core network and the access network as the aggregation layer node, and install diagnostic equipment on the aggregation node to monitor the communication quality and traffic balance between subnets.

[0074] In this embodiment, an aggregation switch or border router connecting the core network and the access network is selected as the aggregation layer node. Devices that support link aggregation and dynamic load balancing are preferred. A diagnostic module with subnet isolation monitoring function is deployed at the aggregation node. A dual-port data acquisition unit is configured to monitor the latency jitter, packet loss rate and bandwidth utilization of the uplink core layer link and the downlink access layer link respectively. A dynamic routing adjustment algorithm module is integrated to achieve real-time optimization of traffic balance status.

[0075] Step S104: Select the access switch and wireless access point device as the access layer node, and install diagnostic equipment on the access node to monitor the stability of terminal connection and user experience indicators.

[0076] Furthermore, in this embodiment, a collaborative sensing network is used to dynamically acquire multimodal data, including: acquiring digital voltage signals reflected from optical signals, synchronously capturing TCP retransmission rate, routing oscillation counts and network traffic fluctuation data, and acquiring the device vibration spectrum through a piezoelectric ceramic sensor array;

[0077] In this embodiment, when acquiring the reflected digital voltage signal of the optical signal, the reflected optical signal is acquired through a three-port circulator in the optical module and converted into a reflected digital voltage signal. When the signal exceeds a first threshold, a fault judgment is triggered. The vibration spectrum of the device is acquired through a piezoelectric ceramic sensor array to identify hidden hardware faults, including fan malfunctions and circuit board aging.

[0078] Furthermore, in this embodiment, when nodes exchange data in real time via a time-sensitive network, it includes:

[0079] Configure a high-precision clock source in the network and synchronize time using the IEEE 1588v2 protocol;

[0080] Deploy the TSN protocol stack on all diagnostic nodes to support priority queuing, time-aware shaping, and frame preprocessing. Deploy the TSN protocol stack, establish TSN streams for data transmission between diagnostic nodes, and configure traffic scheduling plans.

[0081] Please continue to refer to Figure 1 The present invention provides a method for diagnosing communication network faults, which further includes the following steps:

[0082] Step S20: Use a collaborative sensing network to dynamically acquire multimodal data, construct an acoustic-optical-protocol correlation matrix, and map the correlation between physical layer signal distortion, equipment vibration and protocol layer anomalies.

[0083] In this step, see Figure 3 As shown, the acoustic-optical-protocol correlation matrix is ​​constructed, including:

[0084] Step S201: Collect the reflected optical signal from the optical fiber through the three-port circulator in the optical module and convert it into a reflected digital voltage signal. Use a piezoelectric ceramic sensor array to collect the vibration spectrum of the device and capture TCP retransmission rate, routing oscillation count and network traffic fluctuation data.

[0085] In step S201 of this embodiment, the reflected optical signal of the optical fiber is collected by the three-port circulator in the optical module, which can directly capture the abnormal state of the optical fiber physical link (such as signal attenuation and distortion), and provide the original signal basis for physical layer faults. When the reflected digital voltage signal exceeds the preset threshold, it can be preliminarily determined that there is a risk of fault in the optical fiber link.

[0086] In this embodiment, the TCP retransmission rate refers to the fact that the TCP protocol triggers retransmission when data packets are lost to ensure reliable data transmission. An increased retransmission rate usually indicates packet loss, network congestion, or abnormal terminal reception. The number of routing oscillations refers to the instability of routing status caused by frequent updates to the routing table of routing nodes. An increase in the number of oscillations will directly affect the stability of network transmission. In addition, the network traffic fluctuation data refers to the real-time collection of traffic transmission rates between nodes. Sudden peaks or sustained troughs in traffic may correspond to abnormal network load or insufficient link bandwidth.

[0087] Step S202: Clean, denoise and standardize the collected physical layer and protocol layer data, extract features from the protocol layer data to obtain the statistical TCP retransmission count, routing oscillation frequency and traffic fluctuation amplitude.

[0088] Step S203: Perform preliminary matching between the physical layer characteristics of optical signal reflection voltage and vibration spectrum characteristics and the protocol layer characteristics such as TCP retransmission rate, routing oscillation count, and traffic fluctuation data, and use statistical analysis to perform in-depth correlation analysis to identify significant correlation relationships.

[0089] In this embodiment, the statistical analysis algorithms used in the step of performing in-depth correlation analysis using statistical analysis include, but are not limited to, Pearson correlation analysis, chi-square test and regression analysis.

[0090] Step S204: Based on the correlation analysis results, construct the acoustic-optical-protocol correlation matrix, where the rows of the matrix represent physical layer features, the columns represent protocol layer features, and the matrix elements represent the correlation strength between the two.

[0091] Traditional fault diagnosis techniques require separate investigation of the physical layer and protocol layer. However, the acousto-optic-protocol correlation matrix in this embodiment can directly establish the correspondence between protocol layer anomalies and physical layer fault sources, directly locate fiber optic link faults, and significantly shorten fault investigation time.

[0092] Please continue to refer to Figure 1 The communication network fault diagnosis method provided in this embodiment of the invention further includes the following steps:

[0093] Step S30: Extract features from the collected multimodal data, generate a mapping relationship between data sending nodes and receiving nodes, and generate a minimum coverage detection path based on the current network load and topology.

[0094] In this step, see Figure 4 As shown, a minimum coverage detection path is generated based on the current network load and topology, including:

[0095] Step S301: Obtain the current network topology information, including nodes, links, and corresponding weights;

[0096] In step S301 of this embodiment, the node information is used to distinguish between core layer, aggregation layer, and access layer nodes, including recording the ID, geographical location, hardware performance, and detection priority of each node, such as core node priority > aggregation node > access node; the link information includes recording the physical / logical connections between nodes and collecting key link parameters, including link bandwidth, link latency, link packet loss rate, and currently available link bandwidth. The link bandwidth characterizes the maximum transmission capacity of the link, the link latency characterizes the transmission time of data packets from the start point to the end point of the link, the link packet loss rate characterizes the probability of data packet loss during link transmission, and the currently available link bandwidth characterizes the remaining bandwidth resources available for detection in real time.

[0097] Furthermore, to quantify the detection cost of the link, this embodiment of the application fuses multi-dimensional link parameters into a single weight, represented as: A higher weight value indicates a higher detection cost for that link; weight The calculation formula is expressed as:

[0098] ;

[0099] In the formula, , and Represents the weighting coefficient, and ; This represents the maximum latency across all links in the network. The normalized value representing the link delay; Indicates link bandwidth utilization; Indicates the packet loss rate of the link;

[0100] Step S302: Monitor network load in real time and use traffic analysis tools to determine the current network traffic distribution and bottleneck location;

[0101] Specifically, in step S302, when monitoring network load in real time, the monitoring indicators include node CPU utilization, node memory utilization, actual link traffic, and port queue length.

[0102] Traffic analysis tools such as NetFlow Analyzer and Wireshark can be used to collect traffic data in real time; and the snmpwalk command deployed on the diagnostic node can be used to obtain the device's CPU / memory utilization; the actual link traffic can be counted through the traffic scheduling module of TSN;

[0103] Step S303: Based on the initial path search, set the search target and constraints, and obtain the minimum path that covers all nodes that need to be detected. The constraints of the path search include the current network load and link status.

[0104] Specifically, in step S303, the search objective is set as minimizing the total path weight, expressed as:

[0105] ;

[0106] In the formula, P represents the generated detection path, which consists of multiple links; e represents the link index. Indicates weight;

[0107] Step S304: Perform a path search on the network topology graph to find an initial set of paths that cover all nodes that need to be detected;

[0108] In one implementation of this application, a greedy algorithm or a genetic algorithm is used to perform path search based on topological weights, optimization objectives and constraints to generate an initial path set covering all nodes to be detected;

[0109] For small-to-medium-sized networks with fewer than 50 nodes, this embodiment uses a greedy algorithm to perform path search.

[0110] For large-scale networks with more than 50 nodes, this embodiment uses a genetic algorithm to perform path search;

[0111] Step S305: Optimize the initial path set, reduce redundant paths, and verify the generated detection paths.

[0112] Please continue to refer to Figure 1 The present invention provides a method for diagnosing communication network faults, which further includes the following steps:

[0113] Step S40: Construct an over-diffusion fault propagation model, locate the root cause node of the minimum coverage detection path through gradient reverse tracing, calculate the fault severity parameter, trigger a graded repair strategy based on the fault severity parameter, and issue a repair command.

[0114] In this step, see Figure 5 As shown, constructing the super-diffusion fault propagation model includes:

[0115] Step S401: Set the initial state parameters for each network node, including the node's health status, fault propagation probability, and fault propagation speed;

[0116] Step S403: Obtain the super-diffusion factor through historical fault data and statistical analysis. The super-diffusion factor is used to characterize the rapid propagation characteristics of faults in complex networks.

[0117] In this embodiment, the rapid propagation characteristics of faults are characterized by the superdiffusion factor;

[0118] Specifically, the fault state of node i at time t is set as follows: ,when This indicates good health. Time indicates a fault;

[0119] The superdiffusion equation for fault propagation is then expressed as:

[0120] ;

[0121] In the formula, D represents the conventional diffusion coefficient, which is used to describe the basic propagation speed of a fault and is set based on the inter-node link bandwidth. Indicates the superdiffusion factor. This was obtained through statistical analysis of historical fault data. When the value is greater than 0, it is used to characterize the jump propagation characteristics of faults, such as a core node fault rapidly affecting multiple access nodes; This represents the Laplace operator, used to describe the normal propagation of faults among neighboring nodes. This represents the fourth-order Laplace operator, used to describe the long-range super-diffusion of faults; Let i represent the set of neighboring nodes of node i. This represents the probability of fault propagation from node j to node i;

[0122] in addition, This represents the instantaneous rate of change of the fault state of node i over time;

[0123] When the rate of change is greater than 0, it indicates that the fault status value of node i is rising, which means that the fault is spreading or aggravating to this node, or it may be caused by the propagation of the fault by adjacent faulty nodes through the link.

[0124] When the rate of change is equal to 0, it means that the fault state of node i is stable and unchanged, either the node is completely healthy and there is no external fault propagation, or the node is completely faulty and there is no state change.

[0125] When the rate of change is less than 0, it indicates that the fault state value of node i is decreasing, which means that the fault is being repaired or the fault propagation is blocked.

[0126] Step S402: Based on the topology of the communication network, construct a fault propagation matrix, where the matrix describes the connection relationship between nodes and the probability of fault propagation;

[0127] In step S402, the meaning of the matrix elements of the propagation matrix M is expressed as follows: , This represents the probability of fault propagation from node j to node i. If node i and node j have no direct link connection, then... This embodiment is based on the physical topology of the communication network, and transforms the existence of links into non-zero elements of a matrix; in addition, when the network topology changes, the matrix is ​​updated in real time to ensure that the matrix is ​​consistent with the actual network state.

[0128] Step S404: Using the minimum coverage detection path, sample the multimodal data to obtain the real-time status information of the nodes;

[0129] In the minimum coverage detection path of step S404 in this embodiment, by covering all nodes to be detected and minimizing the total link weight, complete and real-time node status data is provided for dynamic fault tracing of the hyperdiffusion model under the premise of efficient resource utilization.

[0130] Step S405: Perform gradient calculation on the sampled data to identify potential fault propagation paths. By tracing the fault propagation path in reverse, backtrack step by step until the starting point of the fault is found.

[0131] In step S405 of this embodiment, by analyzing the spatial gradient change of the fault state, the fault propagation path is traced in reverse to finally locate the initial fault node, thus solving the problem that traditional technologies can only detect the fault symptoms and cannot trace the source.

[0132] Specifically, the fault state gradient in this embodiment is represented as follows: This is used to describe the difference in health status between node i and its neighboring nodes, i.e., the spatial rate of change of fault status; this embodiment is based on sampled real-time status data to describe the health status differences between node i and its neighboring nodes j. and The gradient is calculated and expressed as:

[0133] ;

[0134] In the formula, k represents the number of adjacent nodes. Represents the set of adjacent nodes. Indicates link weight; This represents the fault state of node i at time t. This indicates the fault state of node j at time t;

[0135] In path identification, the direction of positive gradient indicates the direction of fault propagation. Starting from the currently detected high-risk node, the path searches for the previous level node along the opposite direction of the gradient until it backtracks to the node with a gradient of 0 or negative, thus obtaining the root cause node.

[0136] Step S406: Set the assessment index for the severity of the fault, perform a weighted comprehensive assessment of each index, and calculate the fault severity parameter.

[0137] As can be seen, this invention, by distributing diagnostic nodes in the core layer, aggregation layer, and access layer, and utilizing time-sensitive networking for real-time data exchange to form a collaborative sensing network, can accurately locate fault locations and reduce fault investigation time. By using the collaborative sensing network for dynamic acquisition of multimodal data and constructing an acoustic-optical-protocol correlation matrix, it can comprehensively analyze physical layer signals, equipment vibration, and protocol layer data, improving the accuracy and reliability of fault diagnosis. Furthermore, by constructing an over-diffusion fault propagation model and utilizing gradient reverse tracing technology, it can effectively identify and locate the root cause node with the smallest coverage detection path, improving the efficiency and accuracy of fault diagnosis.

[0138] In addition, the present invention extracts features from the collected multimodal data and generates a mapping relationship between data sending nodes and receiving nodes, which can dynamically adapt to changes in the network environment and improve the flexibility and adaptability of the diagnostic system.

[0139] This invention generates a minimum coverage detection path based on the current network load and topology, optimizing the fault detection path and reducing detection time and resource consumption. By calculating fault severity parameters and using evaluation indicators such as impact range, number of affected nodes, and fault duration, it can comprehensively assess the impact of faults and provide a scientific basis for repair strategies. Based on the fault severity parameters, it triggers a graded repair strategy and issues repair instructions, realizing the automation and intelligence of fault repair, improving fault handling efficiency, and reducing network downtime.

[0140] It should be noted that the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may, for example, be executed synchronously or asynchronously in multiple modules.

[0141] It should be understood that although the above description follows a certain order, these steps are not necessarily executed in that order. Unless otherwise expressly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, some steps in this embodiment may include multiple steps or multiple stages, which are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least a portion of steps or stages in other steps.

[0142] In a second aspect, the present invention also provides a communication network fault diagnosis device, comprising the following modules:

[0143] The node deployment module is used to deploy core layer, aggregation layer, and access layer diagnostic nodes in a distributed manner according to the topology of the communication network. The nodes exchange data in real time through a time-sensitive network to form a collaborative sensing network.

[0144] The multimodal data acquisition module is used to dynamically acquire multimodal data using a collaborative sensing network, construct an acoustic-optical-protocol correlation matrix, and map the correlation between physical layer signal distortion, equipment vibration, and protocol layer anomalies.

[0145] The path generation module is used to extract features from the collected multimodal data, generate a mapping relationship between data sending nodes and receiving nodes, and generate a minimum coverage detection path based on the current network load and topology.

[0146] The fault severity diagnosis module is used to construct an ultra-diffusion fault propagation model, locate the root cause node of the minimum coverage detection path through gradient back-source tracing, calculate the fault severity parameter, trigger a graded repair strategy based on the fault severity parameter, and issue repair instructions.

[0147] Through the detailed steps described above, the communication network fault diagnosis device of the present invention is used to execute the steps of the communication network fault diagnosis method in the above embodiments, which will not be repeated here. The communication network fault diagnosis method and device provided by the present invention significantly improves the accuracy, efficiency and reliability of communication network fault diagnosis through the comprehensive application of multi-level collaborative sensing networks, multi-modal data fusion, intelligent fault propagation models and hierarchical repair strategies, and has important practical application value.

[0148] A third aspect of the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program that, when executed by the processor, implements the method of any of the above embodiments.

[0149] This computer device includes a processor and a memory, and may also include an input system and an output system. The processor, memory, input system, and output system can be connected via a bus or other means. The input system can receive input digital or character information and generate signal inputs related to communication network fault diagnosis. The output system may include display devices such as a screen.

[0150] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples. Within the framework of the invention, technical features of the above embodiments or different embodiments can be combined, and many other variations of different aspects of the invention exist, which are not provided in the details for the sake of brevity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A method for diagnosing faults in a communication network, characterized in that, The method includes the following steps: In the communication network, core layer, aggregation layer, and access layer diagnostic nodes are deployed in a distributed manner according to topology. The nodes exchange data in real time through time-sensitive networks to form a collaborative sensing network. By using a collaborative sensing network to dynamically acquire multimodal data, an acoustic-optical-protocol correlation matrix is ​​constructed to map the correlation between physical layer signal distortion, equipment vibration and protocol layer TCP retransmission rate, routing oscillation count, and abnormal traffic fluctuation data. Feature extraction is performed on the collected multimodal data to generate a mapping relationship between data sending nodes and receiving nodes. Based on the current network load and topology, a minimum path covering all nodes requiring detection (i.e., the minimum coverage path) is generated. The topology information includes nodes, links, and corresponding weights, where the weights are represented as... , ; In the formula, This represents the link weight between node i and its neighboring node j; , and Represents the weighting coefficient, and ; This represents the maximum latency across all links in the network. The normalized value representing the link delay; Indicates link bandwidth utilization; Indicates the packet loss rate of the link; A super-diffusion fault propagation model is constructed. The root cause node of the minimum coverage path is located by gradient back-source tracing. The fault severity parameter is calculated. Based on the fault severity parameter, a graded repair strategy is triggered and a repair command is issued. The super-diffusion equation for fault propagation in the super-diffusion fault propagation model is: In the formula, D represents the conventional diffusion coefficient, which is used to describe the basic propagation speed of the fault; Indicates the superdiffusion factor. When the value is greater than 0, it is used to characterize the jump propagation characteristics of faults; The second-order Laplace operator represents the normal propagation of a fault in adjacent nodes. A fourth-order Laplace operator representing long-range super-diffusion of faults; Represents the set of all neighboring nodes of node i. This represents the probability of fault propagation from node j to node i; Among them, for node i and its neighboring node j and The gradient is calculated and expressed as: In the formula, k represents the number of neighboring nodes of node i. , This represents the fault state of node i and its neighboring node j at time t.

2. The communication network fault diagnosis method as described in claim 1, characterized in that, The construction of a collaborative sensing network also includes the following steps: Based on the network topology diagram, select the core routers and switches in the network as core layer nodes; Diagnostic devices, including high-performance data acquisition modules and processing units, are installed on selected core layer nodes to monitor the stability of high-bandwidth traffic and the status of critical routing nodes. Select an aggregation switch or router that connects the core network and the access network as an aggregation layer node, and install diagnostic equipment on the aggregation layer node to monitor the communication quality and traffic balance between subnets. Select access switches and wireless access point devices as access layer nodes, and install diagnostic devices on the access layer nodes to monitor the stability of terminal connections and user experience indicators.

3. The communication network fault diagnosis method as described in claim 2, characterized in that, Multimodal data dynamic acquisition is performed using a collaborative sensing network, including: acquiring digital voltage signals reflected from optical signals, synchronously capturing TCP retransmission rate, routing oscillation counts, and network traffic fluctuation data, and acquiring the device vibration spectrum through a piezoelectric ceramic sensor array.

4. The communication network fault diagnosis method as described in claim 3, characterized in that, When collecting the reflected digital voltage signal of the optical signal, the reflected optical signal is collected through the three-port circulator in the optical module and converted into a reflected digital voltage signal. When the signal exceeds the first threshold, a fault judgment is triggered.

5. The communication network fault diagnosis method as described in claim 4, characterized in that, The vibration spectrum of the device is collected by a piezoelectric ceramic sensor array to identify hidden hardware faults, including fan malfunctions and circuit board aging.

Images