Network fault early warning method and device, electronic equipment and storage medium

Abstract

The application discloses a network fault early warning method and device, electronic equipment and storage medium. The method comprises the following steps: acquiring network topology information of a target network; simulating a base station for accessing the target network and a terminal device for accessing the base station to obtain a simulated base station and a simulated terminal device, and connecting the simulated terminal device to the simulated base station; simulating a first network element device in the network topology information to obtain a simulated network element device, and connecting the simulated base station and the simulated network element device to the target network based on the network topology information; performing VoNR service through the simulated terminal device to obtain network state information of network nodes of the target network under the VoNR service, and generating a VoNR network state diagram based on the network state information; and predicting a network fault based on the VoNR network state diagram to obtain a target prediction result representing whether a network node of the target network has a fault.

Description

Technical Field

[0001] This application belongs to the field of communication technology, and specifically relates to a method, device, electronic device and storage medium for early warning of network faults. Background Technology

[0002] Voice over New Radio (VoNR) is a voice service solution based on the IP Multimedia Subsystem (IMS) network in 5G networks, which can significantly improve the security of voice calls and the quality of video calls. With the large-scale deployment of 5G networks, VoNR services have become an important business item and profit growth point for operators. As the core networks in the VoNR service bearer network, such as the 5G Core (5GC) network and the IMS network, they are of paramount importance for operation and maintenance.

[0003] For VoNR services, the early warning technologies for related network faults mainly employ traditional passive analysis methods such as network management, signaling monitoring, and deep packet inspection (DPI). Specifically, network management methods involve statistically analyzing the overall operating indicators of core network elements and triggering corresponding early warnings when device-level faults occur. Signaling monitoring methods involve collecting all signaling data from key core network signaling interfaces, performing decoding processing and statistical analysis of various indicators to detect interface degradation. DPI methods involve real-time monitoring of the source, destination, protocol, and port of core network traffic to promptly detect abnormal activity and abnormal traffic.

[0004] However, relevant network fault early warning technologies, based on passive perception fault detection methods, require the analysis and correlation of signaling or network traffic already generated in the network. Faced with large-scale data collection, the analysis and processing time will be very long. At the same time, they do not have the ability to actively monitor specific network elements or specific business scenarios, making it difficult to detect potential faults at the beginning.

[0005] In other words, the relevant network fault early warning technology has the problem of delayed fault detection. Summary of the Invention

[0006] This application provides a method, apparatus, electronic device, and storage medium for early warning of network faults, which can solve the problem of delayed fault detection in related network fault early warning technologies.

[0007] In a first aspect, embodiments of this application provide a method for early warning of network faults. The method includes: acquiring network topology information of a target network; wherein the target network includes a core network in a VoNR (Voice over Radio) service bearer network; simulating a base station used to access the target network and a terminal device used to access the base station to obtain a simulated base station and a simulated terminal device, and connecting the simulated terminal device to the simulated base station; simulating a first network element device in the network topology information to obtain a simulated network element device, and connecting the simulated base station and the simulated network element device to the target network based on the network topology information; simulating VoNR services using the simulated terminal device to obtain network status information of network nodes in the target network under the VoNR service, and generating a VoNR network status diagram based on the network status information; and performing network fault prediction based on the VoNR network status diagram to obtain a target prediction result characterizing whether the network nodes in the target network have faults.

[0008] Secondly, embodiments of this application provide a network fault early warning device, comprising: an acquisition module for acquiring network topology information of a target network; wherein the target network includes a core network in a VoNR service bearer network; a first simulation module for simulating a base station used to access the target network and a terminal device used to access the base station, obtaining a simulated base station and a simulated terminal device, and connecting the simulated terminal device to the simulated base station; a second simulation module for simulating a first network element device in the network topology information, obtaining a simulated network element device, and connecting the simulated base station and the simulated network element device to the target network based on the network topology information; a generation module for simulating VoNR service through the simulated terminal device, obtaining network status information of network nodes in the target network under the VoNR service, and generating a VoNR network status diagram based on the network status information; and a prediction module for predicting network faults based on the VoNR network status diagram, obtaining a target prediction result characterizing whether the network nodes in the target network have faults.

[0009] Thirdly, embodiments of this application provide an electronic device comprising: a processor; and a memory arranged to store computer-executable instructions configured to be executed by the processor, the executable instructions including a network fault early warning method as described in the first aspect.

[0010] Fourthly, embodiments of this application provide a storage medium for storing computer-executable instructions that cause a computer to execute the network fault early warning method as described in the first aspect.

[0011] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the network fault early warning method as described in the first aspect.

[0012] In a sixth aspect, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the network fault early warning method as described in the first aspect.

[0013] In this embodiment, network topology information of a target network is obtained; wherein the target network includes the core network in the VoNR service bearer network; a base station for accessing the target network and a terminal device for accessing the base station are simulated to obtain a simulated base station and a simulated terminal device, and the simulated terminal device is connected to the simulated base station; a first network element device in the network topology information is simulated to obtain a simulated network element device, and based on the network topology information, the simulated base station and the simulated network element device are connected to the target network; by simulating the simulated terminal device to perform VoNR services, network status information of the network nodes of the target network under the VoNR services is obtained, and a VoNR network status diagram is generated based on the network status information; based on the VoNR network status diagram, network fault prediction is performed to obtain a target prediction result characterizing whether the network nodes of the target network have faults. Compared to related network fault early warning technologies, this application uses simulation to actively probe the target network for VoNR services, obtaining network status information of network nodes in the target network under VoNR services. This results in the generation of a VoNR network status diagram, which, based on this diagram, can proactively identify potential problems in the target network, enabling rapid fault location and improving fault handling efficiency for network operations and maintenance departments. This solves the problem of delayed fault detection inherent in related network fault early warning technologies. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a network fault early warning method provided in an embodiment of this application; Figure 2(a) is a schematic diagram of a VoNR network state before node merging provided in an embodiment of this application; Figure 2(b) is a schematic diagram of the target VoNR network state after node merging provided in an embodiment of this application; Figure 3 This is a schematic diagram illustrating the construction of a network state database provided in an embodiment of this application; Figure 4 This is a flowchart illustrating another network fault early warning method provided in an embodiment of this application; Figure 5 This is a schematic diagram of the structure of a network fault early warning device provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0015] 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.

[0016] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0017] The network fault early warning method, device, electronic device, and storage medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0018] Figure 1 This illustration shows a network fault early warning method according to an embodiment of the present invention. The method can be executed by an electronic device, which may include a server and / or a terminal device, wherein the terminal device may be, for example, a vehicle-mounted terminal or a mobile phone terminal. In other words, the method can be executed by software or hardware installed on the electronic device, and the method includes the following steps: S102: Obtain network topology information of the target network.

[0019] The target network includes the core network within the VoNR service bearer network. The core network may include the 5GC network and the IMS network.

[0020] S104: Simulate the base station used to access the target network and the terminal equipment used to access the base station to obtain the simulated base station and the simulated terminal equipment, and connect the simulated terminal equipment to the simulated base station.

[0021] S106: Simulate the first network element device in the network topology information to obtain the simulated network element device, and connect the simulated base station and the simulated network element device to the target network based on the network topology information.

[0022] The first network element device may include: Authentication Management Function (AMF) device, Session Management Function (SMF) device, User Plane Function (UPF) device, Unified Data Management (UDM) device, Authentication Server Function (AUSF) device, and Proxy-Call Session Control Function (P-CSCF) device.

[0023] Specifically, the target network includes a second network element device, which includes the first network element device, and at least includes multiple of the following: an Interrogating-Call Session Control Function (I-CSCF) device, a Serving Call Session Control Function (S-CSCF) device, and a Telecom Application Server (TAS) device. The second network element device of the aforementioned 5GC network includes an AMF device, an SMF device, a UPF device, a UDM device, a PCF device, and an AUSF device; the second network element device of the aforementioned IMS network includes a P-CSCF device, an I-CSCF device, an S-CSCF device, and a TAS device.

[0024] In practical applications, based on network topology information, simulated base stations can be connected to all AMF devices in the target network; based on network topology information, the corresponding AMF devices can be simulated to obtain simulated AMF devices, which are then connected to all SMF devices, UDM devices, PCF devices, and AUSF devices in the target network; based on network topology information, the corresponding UPF devices can be simulated to obtain simulated UPF devices, which are then connected to all P-CSCF devices in the target network; based on network topology information, the corresponding P-CSCF devices can be simulated to obtain simulated P-CSCF devices, which are then connected to all I-CSCF devices and S-CSCF devices in the target network.

[0025] S108: By simulating the terminal equipment to perform VoNR service, obtain the network status information of the network nodes of the target network under the VoNR service, and generate a VoNR network status diagram based on the network status information.

[0026] The network status information of the network node may include one or more key performance indicators (KPIs) of the second network element device corresponding to the network node. Specifically, the key performance indicators of the second network element device may include, but are not limited to, one or more of the following: IMS Protocol Data Unit (PDU) session establishment success rate or latency information, VoNR registration success rate or latency information, VoNR network connection rate or latency information, VoNR audio user connection rate or latency information, voice dedicated bearer establishment success rate or latency information, video dedicated bearer establishment success rate or latency information, voice media stream data transmission packet loss rate or jitter information, video media stream data transmission packet loss rate or jitter information, voice call one-way pass rate, and video call one-way pass rate.

[0027] S110: Based on the VoNR network state diagram, perform network fault prediction to obtain the target prediction result that characterizes whether the network nodes of the target network have faults.

[0028] In practical applications, the prediction result of the target can be obtained based on the VoNR network state diagram and the trained network fault prediction model.

[0029] Specifically, the target prediction result may include the node labels of network nodes in the target network; for example, the node label of a faulty target network node may be "1", and the node label of network nodes other than the target network node may be "0". It should be noted that the above "1" and "0" are only for ease of understanding, and the node labels can be set according to actual needs and do not constitute a specific limitation.

[0030] The network fault early warning method provided in this invention involves: acquiring network topology information of a target network, wherein the target network includes the core network in the VoNR service bearer network; simulating base stations and terminal devices used to access the target network to obtain simulated base stations and simulated terminal devices, and connecting the simulated terminal devices to the simulated base stations; simulating the first network element device in the network topology information to obtain simulated network element devices, and connecting the simulated base stations and simulated network element devices to the target network based on the network topology information; simulating VoNR services through the simulated terminal devices to obtain network status information of network nodes in the target network under the VoNR service, and generating a VoNR network status diagram based on the network status information; and performing network fault prediction based on the VoNR network status diagram to obtain a target prediction result characterizing whether network nodes in the target network have faults. Compared to related network fault early warning technologies, this application uses simulation to actively probe the target network for VoNR services, obtaining network status information of network nodes in the target network under VoNR services. This results in the generation of a VoNR network status diagram, which, based on this diagram, can proactively identify potential problems in the target network, enabling rapid fault location and improving fault handling efficiency for network operations and maintenance departments. This solves the problem of delayed fault detection inherent in related network fault early warning technologies.

[0031] In one implementation, based on the VoNR network state diagram, network fault prediction is performed to obtain a target prediction result (i.e., S110) characterizing whether network nodes in the target network have faults. The following steps A1 to A3 can be executed: Step A1: Based on the network topology information, merge nodes in the VoNR network state diagram to generate the target VoNR network state diagram.

[0032] Specifically, based on network topology information, node merging can be performed on the VoNR network state diagram to generate a coarse-grained target VoNR network state diagram. For example, the VoNR network state diagram before node merging can be shown in Figure 2(a), and the target VoNR network state diagram generated after node merging can be shown in Figure 2(b). As can be seen from the figures, merging local states can effectively reduce the dimensionality of the VoNR network state diagram. It should be noted that the above VoNR network state diagram and target VoNR network state diagram only show some network nodes and the connections between them; other information is not shown, and the network nodes shown are only for ease of understanding and do not constitute a specific limitation on the VoNR network state diagram and target VoNR network state diagram.

[0033] Step A2: Based on the target VoNR network state graph, perform coarse-grained fault prediction using the trained network fault prediction model to obtain initial prediction results that characterize whether there are abnormal network nodes in the target VoNR network state graph.

[0034] Specifically, the initial prediction results may include node labels of network nodes in the target network and location information of abnormal network nodes. The location information of network nodes can be the location information of abnormal network nodes in the target VoNR network state graph; for example, the node label of an abnormal network node can be "1", and the node label of network nodes other than the abnormal network node can be "0". It should be noted that the "1" and "0" mentioned above are for ease of understanding only, and the node labels can be set according to actual needs and do not constitute a specific limitation.

[0035] Step A3: If the initial prediction results indicate the presence of abnormal network nodes, based on the initial prediction results, simulate the forward network element device of the target network element device corresponding to the abnormal network node, and perform fine-grained fault prediction by connecting the forward network element device to the target network element device to obtain the target prediction result.

[0036] The forward network element device is used to interface with the target network element device. The target prediction result may also include the location information of the target network node; the target network node is an abnormal network node with a fault.

[0037] For example, if the target network element device corresponding to the abnormal network node includes an AMF device, then the forward network element device is a simulated base station. By connecting the simulated base station to the AMF device, VoNR service dialing is performed, i.e., fine-grained fault prediction is performed, thereby further verifying whether there is a fault. If there is a fault, the network node corresponding to the AMF device is the target network node, and the location information of the AMF device in the target VoNR network status diagram is given.

[0038] In this embodiment, a coarse-grained target VoNR network state graph is generated by merging nodes, thereby performing coarse-grained early warning first and then fine-grained verification. This can effectively reduce the computational load of the trained network fault prediction model when performing node anomaly early warning tasks when the VoNR network is large, and improve the overall efficiency of fault prediction.

[0039] Compared to existing network fault early warning technologies, which often require clear indications of abnormal node states before providing feedback, this application considers that earlier detection of abnormal network node states allows for earlier warnings of potential faults, thus informing operators to take preventative measures. Therefore, it specifically constructs a correspondence between early fault network states—that is, historical network state information before a true fault occurs (i.e., historical network state information at the time of historical detection prior to the target historical detection time below)—and the node state labels described below. Furthermore, it considers the significant differences in the development speed of different faults; some faults progress rapidly, while others... The fault progresses slowly. To address this, a network state database is constructed based on the cumulative characteristics of the fault occurrence process. This database, using time-slice data and node state labels, is used to train the network fault prediction model. This enhances the generalization ability of the trained model for faults of varying degrees, enabling it to accurately determine initial predictions under different levels of anomalies. Furthermore, it provides reliable data support for further research into the fault progression mechanism in VoNR service networks and related technologies such as more advanced fault early warning models. Specifically, the following embodiment illustrates this.

[0040] In one implementation, before performing coarse-grained fault prediction based on the target VoNR network state graph and using the trained network fault prediction model (i.e., step A2), steps B1 to B3 can also be performed to obtain the trained network fault prediction model: Step B1: Obtain the historical network state information and corresponding node state labels of each network node in the target network at the target historical detection time, as well as the historical network state information at historical detection times before the target historical detection time.

[0041] Among them, the target historical detection time is the historical detection time when the target network has a fault; the node status label indicates whether the network node has a fault.

[0042] In practical applications, the aforementioned historical network state information and node state labels can be obtained through active probing. That is, steps b1 to b4 below can be executed to obtain the historical network state information and corresponding node state labels at the target historical probing time, as well as the historical network state information at historical probing times prior to the target historical probing time. Step b1: Obtain the historical network topology information of the target network mentioned above; Step b2: Simulate the base station used to access the target network and the terminal equipment used to access the base station to obtain the simulated base station and the simulated terminal equipment, and then connect the simulated terminal equipment to the simulated base station; Step b3: Simulate the first network element device in the historical network topology information to obtain the simulated network element device, and connect the simulated base station and the simulated network element device to the target network based on the historical network topology information; Step b4: Using a simulated terminal device to perform VoNR service, continuously probe and record the historical network status information at each historical probe moment until a fault occurs in the target network. Determine the historical network status information and corresponding node status tags at the target historical probe moment, as well as the historical network status information at historical probe moments prior to the target historical probe moment. If no fault occurs for a period of time T, the latest probed historical network status information will replace the historical network status information of the previous historical probe moment in the storage queue. The time T can be set according to actual needs. There is a certain time interval between historical probe moments, which can be set according to actual needs, for example, 1 minute.

[0043] For example, the node status label corresponding to each network node at the target's historical detection time is "0" or "1", where "0" indicates that the network node is normal; "1" indicates that the network node is faulty (i.e., the network node is an abnormal network node). It should be noted that the above node status label is not limited to "0" or "1", but can also be "normal" or "abnormal", etc., without specific limitation.

[0044] Step B2: Based on the historical network state information at the target's historical detection time and the historical network state information at the historical detection time before the target's historical detection time, generate time slice data and store the time slice data and node state labels in the network state database.

[0045] Step B3: Based on the time slice data and node status labels in the network state database, train the network fault prediction model to obtain the trained network fault prediction model.

[0046] Specifically, multiple historical network state information can be selected as multiple time slices from the historical network state information at the target's historical detection time and the historical network state information at previous historical detection times, according to a preset time interval. Then, each time slice data and the node state label at the target's historical detection time are used to form a training data pair. Finally, the network fault prediction model is trained based on the training data pairs to obtain the trained network fault prediction model. The preset time interval can be set according to actual needs and is not specifically limited.

[0047] For example, a preset time interval is used as the historical detection time t. i Taking the time interval between i as an example, historical network state information S at i historical detection times was selected. i As i time slices of data, then each time slice of data s i The node state labels L at historical detection times of the target constitute the training data pair {S}. i Afterwards, the training data pairs can be stored in the network state database. That is, the process of constructing the network state database can be specifically described as follows: Figure 3 As shown.

[0048] Specifically, by repeating step b4 above multiple times, the advantages of active probing can be fully utilized to obtain a large amount of historical network state information and node state labels, which can then be used to form training data pairs.

[0049] Compared to the "passive" detection technology of related network fault early warning technologies, which can only obtain network status information when a significant fault has been detected, thus failing to acquire network status information corresponding to the early stages of a fault, this embodiment actively detects and acquires the historical network status information and corresponding node status labels of each network node in the target network at the target historical detection time, as well as the historical network status information at historical detection times before the target historical detection time. That is, the historical network status information throughout the entire time period is retained during the above-mentioned active detection process. At the same time, the historical network status information of multiple faults in the target network can be obtained through simulation, thereby ensuring that the constructed training data pairs are sufficiently comprehensive. In addition, the preset time interval can be easily set according to actual needs, thereby increasing the number of time slices and constructing more granular training data pairs and network status databases. The preset time interval can also be randomly set within a reasonable range, thereby adding randomness and making the constructed training data pairs richer.

[0050] Furthermore, at the beginning of training, the learnable parameters in the network fault prediction model are randomly initialized. During training, training data pairs are randomly extracted from the network state database and input into the network fault prediction model to calculate the feedforward loss. Since the probability of network node failure in the target network is very low, there is a severe imbalance between abnormal and normal network nodes. To address this, a weighted classification loss function can be used during training, appropriately increasing the weight of the classification error of abnormal network nodes while decreasing the weight of the classification error of normal network nodes. After obtaining the feedforward loss, the gradients corresponding to the network parameters in the network fault prediction model are calculated using the backpropagation algorithm, and the parameters are updated accordingly. This process is repeated until the network fault prediction model converges, resulting in the trained network fault prediction model.

[0051] In one implementation, by simulating VoNR service using a terminal device, the network status information of the network nodes in the target network under the VoNR service is obtained (i.e., S108), and the following steps C1 to C2 can be executed: Step C1: Simulate the terminal device initiating the corresponding VoNR call request under different VoNR service scenarios, and receive the signaling message returned by the second network element device of the target network based on the VoNR call request.

[0052] Specifically, on the simulated network element device, the simulated terminal device initiates corresponding VoNR call requests under different VoNR service scenarios.

[0053] Step C2: Decode the signaling message to obtain the routing path of the second network element device through which each VoNR call request passes, and obtain the network status information of the second network element device corresponding to multiple routing paths.

[0054] Network status information of the second network element device, including one or more of the aforementioned key indicators of the second network element device.

[0055] Based on the network status information, a VoNR network status diagram (i.e., S108) is generated, and the following steps C3 can be performed: Step C3: Generate a VoNR network status diagram based on network status information and routing paths.

[0056] Additionally, the connectivity between network nodes in the VoNR network status graph can be determined based on the data flow of VoNR services.

[0057] In this embodiment, by simulating a terminal device to perform VoNR services, the user service behavior of the target network can be simulated. This enables accurate and rapid active detection of the VoNR services carried by the target network, obtaining the routing path of the second network element device through which each VoNR call request passes, as well as the network status information of the second network element device corresponding to multiple routing paths, and thus generating a VoNR network status diagram.

[0058] In one implementation, based on network topology information, the VoNR network state graph is merged to generate the target VoNR network state graph (i.e., step A1), which can be achieved by performing the following steps D1: Step D1: Based on the user network edge devices (CustomerEdge, CE), firewalls, and / or the location of the data center connected to the network nodes in the network topology information, merge the nodes in the VoNR network state diagram to generate the target VoNR network state diagram.

[0059] The target VoNR network state diagram can be one or more.

[0060] Specifically, the VoNR network state diagram can be merged based on one or more of the following: the CE, firewall, and the location of the data center connected to the second network element device corresponding to the network node, thereby generating one or more coarse-grained target VoNR network state diagrams.

[0061] In this embodiment, considering the possibility of multiple second network element devices malfunctioning due to problems with the data center transmission lines, the VoNR network state diagram is merged based on the CE, firewall, and / or the location of the data center to which the network node is connected, which can reasonably reduce the dimensionality of the VoNR network state diagram.

[0062] In one implementation, the trained network fault prediction model includes Graph Convolutional Networks (GCN).

[0063] The target VoNR network state graph naturally possesses graph structure properties. Therefore, the fault prediction task is abstracted into a graph node classification task, that is, classifying each network node in the target VoNR network state graph as a normal network node or an abnormal network node through discrimination. GCN has a natural advantage in this regard. The core idea of ​​GCN is information aggregation. Specifically, each graph convolution operation in GCN prompts each network node to aggregate effective information from its neighboring network nodes (including itself) to update its own feature representation. Through layer-by-layer information propagation, GCN can capture higher-order neighborhood information of network nodes, and provide classification prediction results based on this.

[0064] For example, the target VoNR network state graph G includes N network nodes, i.e., G = (V, E), where V is the set of network nodes and E is the set of edges (i.e., the connectivity relationships between second network element devices). Based on the target VoNR network state graph G, an adjacency matrix A∈R is constructed. N×N and characteristic matrix X∈R N×F Here, F represents the number of key metrics for each network node. GCN will learn by using the adjacency matrix A and the feature matrix X to predict the category (normal or abnormal) of each network node. To ensure GCN's effective operation, self-loops need to be added to the adjacency matrix A, followed by normalization. Specifically, adding self-loops involves superimposing the identity matrix I onto the adjacency matrix, resulting in the adjacency matrix with added self-loops: =A+I, so that GCN can perceive information about the surrounding network node without ignoring the network node's own information; then, by using the adjacency matrix after adding self-loops. degree matrix Normalization is performed to obtain the normalized adjacency matrix. : ; where the degree matrix The off-diagonal elements are 0, and the degree matrix is... The i-th diagonal element It is calculated using the following formula (1): (1) Where N is the number of network nodes; This represents the element in the i-th row and j-th column of the adjacency matrix after adding the self-loop.

[0065] After normalization, the adjacency matrix is ​​normalized. The sum of each row is 1, thus achieving balance for different network nodes.

[0066] Subsequently, multi-layer graph convolution is performed on the target VoNR network state graph, as shown in formula (2): ( ) in, Indicates the first Network state information of layer network nodes, specifically, when hour , For the first The trainable parameters of the layer, This is the activation function.

[0067] In the above formula (2), The process involves aggregating the features of each network node and its neighboring network nodes using a normalized adjacency matrix. The weights required for this aggregation process are derived from learnable parameters. During the initial training phase, the learnable parameters in GCN are assigned values ​​through random initialization.

[0068] Finally, after performing the above multi-layer graph convolution, the obtained high-level node features can be unfolded in one dimension, and the category to which each network node belongs, i.e., the node label, can be given through a non-linear mapping composed of several fully connected network layers and non-linear activation layers, as shown in the following formula (3): (3) in, This refers to a fully connected layer. This indicates the predicted node label.

[0069] It should be noted that this example only includes two fully connected layers. In actual use, the number of fully connected layers can be adjusted according to the specific situation, and there is no specific limit to this.

[0070] Figure 4 This is a flowchart illustrating another network fault early warning method provided in an embodiment of this application. Figure 4 As shown, the method includes: Step 402: Obtain the historical network state information and corresponding node state labels of each network node in the target network at the target historical detection time, as well as the historical network state information at historical detection times before the target historical detection time.

[0071] The target network includes the core network in the VoNR service bearer network; the target historical detection time is the historical detection time when the target network has a fault; the node status label indicates whether the network node has a fault.

[0072] Step 404: Based on the historical network state information at the target's historical detection time and the historical network state information at the historical detection time before the target's historical detection time, generate time slice data and store the time slice data and node state labels in the network state database.

[0073] Step 406: Based on the time slice data and node status labels in the network state database, train the network fault prediction model to obtain the trained network fault prediction model.

[0074] The trained network fault prediction model includes a graph convolutional neural network.

[0075] Step 408: Obtain the network topology information of the target network.

[0076] Step 410: Simulate the base station used to access the target network and the terminal equipment used to access the base station to obtain the simulated base station and the simulated terminal equipment, and then connect the simulated terminal equipment to the simulated base station.

[0077] Step 412: Simulate the first network element device in the network topology information to obtain the simulated network element device, and connect the simulated base station and the simulated network element device to the target network based on the network topology information.

[0078] Step 414: Simulate the terminal device initiating the corresponding VoNR call request under different VoNR service scenarios, and receive the signaling message returned by the second network element device of the target network based on the VoNR call request.

[0079] Step 416: Decode the signaling message to obtain the routing path of the second network element device through which each VoNR call request passes, and obtain the network status information of the second network element device corresponding to multiple routing paths.

[0080] Step 418: Generate a VoNR network status diagram based on network status information and routing paths.

[0081] Step 420: Based on the network topology information, merge nodes in the VoNR network state diagram to generate the target VoNR network state diagram.

[0082] Step 422: Based on the target VoNR network state graph, perform coarse-grained fault prediction using the trained network fault prediction model to obtain initial prediction results characterizing whether there are abnormal network nodes in the target VoNR network state graph.

[0083] Step 424: If the initial prediction result indicates the presence of abnormal network nodes, based on the initial prediction result, simulate the forward network element device of the target network element device corresponding to the abnormal network node, and perform fine-grained fault prediction by connecting the forward network element device to the target network element device to obtain the target prediction result indicating whether the network node of the target network has a fault.

[0084] The specific processes of steps 402 to 424 above have been described in detail in the above embodiments, and will not be repeated here.

[0085] In this embodiment, network topology information of the target network is obtained, including the core network in the VoNR service bearer network. Simulations are performed on base stations and terminal devices used to access the target network to obtain simulated base stations and simulated terminal devices, and the simulated terminal devices are then connected to the simulated base stations. Simulations are also performed on the first network element device in the network topology information to obtain simulated network element devices, and based on the network topology information, the simulated base stations and simulated network element devices are connected to the target network. By simulating VoNR services using the simulated terminal devices, network status information of network nodes in the target network under the VoNR service is obtained, and a VoNR network status diagram is generated based on this network status information. Based on the VoNR network status diagram, network fault prediction is performed to obtain a target prediction result indicating whether network nodes in the target network have faults. Compared to related network fault early warning technologies, this application actively probes the target network through simulation for VoNR services, obtains network status information of network nodes in the target network under the VoNR service, generates a VoNR network status diagram, and then proactively discovers potential problems in the target network based on the VoNR network status diagram, thereby enabling rapid fault location and assisting network operation and maintenance departments in improving fault handling efficiency. It solves the problem of delayed fault detection in related network fault early warning technologies.

[0086] Corresponding to the network fault early warning method provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides a network fault early warning device. Figure 5 This is a schematic diagram of a network fault early warning device according to an embodiment of the present invention. The network fault early warning device is used to perform... Figures 1 to 4 The described network fault early warning method, such as Figure 5 As shown, the network fault early warning device includes: an acquisition module 510, a first simulation module 520, a second simulation module 530, a generation module 540, and a prediction module 550.

[0087] The acquisition module 510 is used to acquire network topology information of the target network; wherein, the target network includes the core network in the VoNR service bearer network; The first simulation module 520 is used to simulate the base station used to access the target network and the terminal equipment used to access the base station, to obtain the simulated base station and the simulated terminal equipment, and to connect the simulated terminal equipment to the simulated base station. The second simulation module 530 is used to simulate the first network element device in the network topology information to obtain the simulated network element device, and to connect the simulated base station and the simulated network element device to the target network based on the network topology information. The generation module 540 is used to simulate VoNR services through a simulated terminal device, obtain network status information of network nodes in the target network under VoNR services, and generate a VoNR network status diagram based on the network status information. The prediction module 550 is used to predict network faults based on the VoNR network state diagram and obtain the target prediction result that characterizes whether the network nodes of the target network have faults.

[0088] In one implementation, the prediction module 550 includes: The merging unit is used to merge nodes in the VoNR network state diagram based on network topology information to generate the target VoNR network state diagram. The coarse-grained prediction unit is used to perform coarse-grained fault prediction based on the target VoNR network state graph and the trained network fault prediction model to obtain the initial prediction result characterizing whether there are abnormal network nodes in the target VoNR network state graph. The fine-grained prediction unit is used to simulate the forward network element device of the target network element device corresponding to the abnormal network node based on the initial prediction result when the initial prediction result indicates the existence of abnormal network nodes. Then, it connects the forward network element device to the target network element device to perform fine-grained fault prediction and obtain the target prediction result.

[0089] In one implementation, the network fault early warning device further includes a database construction module. The database construction module is used for: Obtain the historical network state information and corresponding node state labels of each network node in the target network at the target historical detection time, as well as the historical network state information at historical detection times before the target historical detection time; wherein, the target historical detection time is the historical detection time when the target network has a fault; the node state label indicates whether the network node has a fault; Based on the historical network state information at the target's historical detection time, and the historical network state information at the historical detection time before the target's historical detection time, time slice data is generated, and the time slice data and node state labels are stored in the network state database. Based on time slice data and node state labels in the network state database, a network fault prediction model is trained to obtain the trained network fault prediction model.

[0090] In one implementation, the generation module 540 is specifically used for: Simulates terminal devices initiating corresponding VoNR call requests under different VoNR service scenarios, and receives signaling messages returned by the second network element device of the target network based on the VoNR call request; Decode the signaling messages to obtain the routing path of the second network element device through which each VoNR call request passes, and obtain the network status information of the second network element device corresponding to multiple routing paths; Based on network state information, a VoNR network state diagram is generated, including: Generate a VoNR network status diagram based on network status information and routing paths.

[0091] In one implementation, the merging unit is specifically used for: Based on the CE, firewall, and / or data center location of the network nodes in the network topology information, the VoNR network state diagram is merged to generate the target VoNR network state diagram.

[0092] In one implementation, the trained network fault prediction model includes a graph convolutional neural network.

[0093] The network fault early warning device provided in this embodiment of the invention acquires network topology information of a target network, wherein the target network includes the core network in the VoNR service bearer network; simulates a base station used to access the target network and a terminal device used to access the base station to obtain a simulated base station and a simulated terminal device, and connects the simulated terminal device to the simulated base station; simulates a first network element device in the network topology information to obtain a simulated network element device, and connects the simulated base station and the simulated network element device to the target network based on the network topology information; simulates the terminal device to perform VoNR services to obtain network status information of network nodes in the target network under the VoNR service, and generates a VoNR network status diagram based on the network status information; and performs network fault prediction based on the VoNR network status diagram to obtain a target prediction result characterizing whether network nodes in the target network have faults. Compared to related network fault early warning technologies, this application uses simulation to actively probe the target network for VoNR services, obtaining network status information of network nodes in the target network under VoNR services. This results in the generation of a VoNR network status diagram, which, based on this diagram, can proactively identify potential problems in the target network, enabling rapid fault location and improving fault handling efficiency for network operations and maintenance departments. This solves the problem of delayed fault detection inherent in related network fault early warning technologies.

[0094] Those skilled in the art will understand that the aforementioned network fault early warning device can be used to implement the network fault early warning method described above. The detailed description therein should be similar to the method description above, and will not be repeated here to avoid repetition.

[0095] Based on the same technical concept, embodiments of this application also provide an electronic device for executing the aforementioned network fault early warning method. Figure 6 This is a schematic diagram of the structure of an electronic device to implement various embodiments of this application. The electronic device can vary significantly due to differences in configuration or performance, and may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640. The processor 610, communications interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call a computer program stored in the memory 630 and executable on the processor 610 to perform the following steps: Obtain the network topology information of the target network; the target network includes the core network in the VoNR service bearer network; Simulate the base station used to access the target network and the terminal equipment used to access the base station to obtain the simulated base station and simulated terminal equipment, and then connect the simulated terminal equipment to the simulated base station; Simulate the first network element device in the network topology information to obtain the simulated network element device, and connect the simulated base station and simulated network element device to the target network based on the network topology information; By simulating the VoNR service using a terminal device, the network status information of the network nodes in the target network under the VoNR service is obtained, and a VoNR network status diagram is generated based on the network status information. Based on the VoNR network state diagram, network fault prediction is performed to obtain the target prediction result that characterizes whether the network nodes of the target network have faults.

[0096] The technical solutions of one or more embodiments of this specification involve obtaining network topology information of a target network, wherein the target network includes the core network in the VoNR service bearer network; simulating base stations and terminal devices used to access the target network to obtain simulated base stations and simulated terminal devices, and connecting the simulated terminal devices to the simulated base stations; simulating the first network element device in the network topology information to obtain simulated network element devices, and connecting the simulated base stations and simulated network element devices to the target network based on the network topology information; simulating VoNR services through simulated terminal devices to obtain network status information of network nodes in the target network under VoNR services, and generating a VoNR network status diagram based on the network status information; and performing network fault prediction based on the VoNR network status diagram to obtain target prediction results characterizing whether network nodes in the target network have faults. Compared to related network fault early warning technologies, this application uses simulation to actively probe the target network for VoNR services, obtaining network status information of network nodes in the target network under VoNR services. This results in the generation of a VoNR network status diagram, which, based on this diagram, can proactively identify potential problems in the target network, enabling rapid fault location and improving fault handling efficiency for network operations and maintenance departments. This solves the problem of delayed fault detection inherent in related network fault early warning technologies.

[0097] The specific execution steps can be found in the various steps of the above-described network fault early warning method embodiment, and can achieve the same technical effect. To avoid repetition, they will not be repeated here.

[0098] It should be noted that the electronic devices in the embodiments of this application include: servers, terminals, or other devices besides terminals.

[0099] The above electronic device structure does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or arrange them differently. For example, an input unit may include a Graphics Processing Unit (GPU) and a microphone, and a display unit may use a liquid crystal display (LCD), organic light-emitting diode (OLED), or other similar display panels. User input units include at least one of a touch panel and other input devices. A touch panel is also called a touchscreen. Other input devices may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, and joysticks, which will not be elaborated further here.

[0100] Memory can be used to store software programs and various data. Memory can primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area can store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, memory can include volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (Synchlink DRAM, SLDRAM), and direct memory bus RAM (DRRAM).

[0101] The processor may include one or more processing units; optionally, the processor integrates an application processor and a modem processor, wherein the application processor mainly handles operations related to the operating system, user interface, and applications, while the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into the processor.

[0102] This application also provides a storage medium storing computer-executable instructions. When these computer-executable instructions are executed by a processor, they implement the various processes of the above-described network fault early warning method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.

[0103] The processor is the processor in the electronic device described in the above embodiments. The storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0104] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described network fault early warning method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0105] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0106] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the various processes of the above-described network fault early warning method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0107] It should be understood that the training and prediction processes of the AI ​​models involved in the various embodiments of this specification all adhere to multiple legal and compliant principles, including legal data sources, compliant data content, compliant data governance, compliant training objectives and schemes, compliant training processes, compliant training environments and tools, and compliant ethical verification of training results, and comply with the requirements of Article 5 of the Patent Law. Among them: Data source legitimacy: All datasets used for AI model training were obtained through legal means, covering three categories: publicly authorized data, data authorized by partners, and self-collected compliant data. Publicly authorized data comes from compliant data sources following open-source licenses such as Apache 2.0, with complete copyright attribution and authorization scope clearly marked, and no unauthorized open-source code or data reuse. Data authorized by partners has been subject to formal data usage agreements, clearly defining the scope, duration, and confidentiality obligations, and possessing a complete authorization chain. For self-collected data involving personal information, strict informed consent procedures have been followed, and anonymization processes (including but not limited to field masking, feature anonymization, and differential privacy technology applications) have been implemented to remove personally identifiable information, fully complying with the requirements of relevant laws and regulations such as the "Interim Measures for the Administration of Generative Artificial Intelligence Services" and the "Personal Information Protection Law."

[0108] Data content compliance: The AI ​​model's dataset undergoes multiple screenings and cleaning processes to remove all content that may violate social morality or harm public interests. It contains no obscene, pornographic, violent, discriminatory, or information that endangers national or public safety, nor does it involve the illegal acquisition or use of genetic resources. For data in sensitive fields (such as healthcare and finance), an additional privacy-preserving computation module (including federated learning and secure multi-party computation technologies) ensures that the data is "usable but not visible," avoiding compliance risks during the original data transmission process and ensuring that the data application scenarios and uses comply with public order and good morals and industry regulatory requirements.

[0109] Data governance norms: A complete data traceability system is established during the AI ​​model training process to automatically record the source, collection time, annotation process, cleaning rules, and permission allocation of training data, generating traceable compliance reports to ensure that the data is verifiable throughout its entire lifecycle. The dataset annotation process for AI models is completed by a professional human R&D team, clearly defining the proportion of human creative contributions and avoiding reliance on AI-generated data that has not undergone substantial human modification, thus meeting the examination requirements for "human main contributions" in AI patent applications.

[0110] Training objectives and plans are compliant: The AI ​​model training objective focuses on network fault prediction. The training scheme and the final output results do not violate any mandatory provisions of laws and administrative regulations, do not harm the public interest or the legitimate rights and interests of others, and do not pose any potential risks of being used for illegal activities, infringing on privacy, or undermining public safety. It strictly adheres to the ethical principle of "intelligent for good".

[0111] Training process compliance: A closed-loop training framework is adopted to ensure compliance and controllability of the training process. The specific process is as follows: First, training samples are obtained through compliant data sources. After the aforementioned data cleaning and desensitization, they are input into the neural network model to generate preliminary training results. Second, an expert system is introduced to verify the preliminary results. Based on preset rules and human expert experience, the feasibility of the results is evaluated, and outputs that may pose ethical risks or compliance hazards are corrected (such as removing decision-making logic that violates public order and good morals, and adjusting model parameters that do not comply with safety regulations). Finally, the loss function weights are dynamically optimized based on expert system feedback to strengthen the model's learning of compliant results, avoid overfitting errors or non-compliant labels, and form a closed-loop control of "data input - model training - expert verification - parameter optimization - result feedback" to ensure that the entire training process complies with A5 ethical review requirements.

[0112] Training environment and tool compliance: AI model training is implemented using nationally licensed chips and a compliant training platform. All open-source frameworks and components used in the training process have obtained their corresponding licenses, and copyright statements and patent citation information are fully retained, with no instances of infringement or reuse. The training environment is built using virtual devices (containers / virtual machines) with fixed random seeds and initial parameter configurations to ensure the reproducibility of the training process. Furthermore, through access control and operation log recording, risks such as data leakage and parameter tampering during training are prevented, ensuring the security and compliance of the training process.

[0113] Training results ethical verification compliance: After the model is trained, it undergoes additional third-party ethical compliance assessment and algorithm filing review to verify that the model output does not violate social morality or harm public interests. For potentially sensitive scenarios (such as public services and intelligent decision-making), a special result verification mechanism is established to ensure that the model always complies with Article 5 of the Patent Law and relevant laws and regulations in practical applications.

[0114] In summary, the data and training process used in the AI ​​model of this specification strictly comply with the relevant provisions of Article 5 of the Patent Law and the Patent Examination Guidelines (2023 Edition), and there are no violations of laws, social ethics, public interests, or illegal use of genetic resources. It fully meets the compliance requirements for patent authorization.

[0115] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include multitasking and parallel processing according to the functions involved, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0116] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of the various embodiments of this application.

[0117] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for early warning of network faults, characterized in that, The method includes: Obtain network topology information of the target network; wherein, the target network includes the core network in the New Radio Voice (VoNR) service bearer network; Simulations are performed on the base station used to access the target network and the terminal device used to access the base station to obtain a simulated base station and a simulated terminal device, and the simulated terminal device is then connected to the simulated base station. The first network element device in the network topology information is simulated to obtain a simulated network element device, and based on the network topology information, the simulated base station and the simulated network element device are connected to the target network; By simulating the simulated terminal device to perform VoNR services, network status information of network nodes of the target network under the VoNR service is obtained, and a VoNR network status diagram is generated based on the network status information. Based on the VoNR network state diagram, network fault prediction is performed to obtain a target prediction result that characterizes whether the network nodes of the target network have faults.

2. The method according to claim 1, characterized in that, The process of predicting network faults based on the VoNR network state diagram to obtain target prediction results characterizing whether network nodes in the target network have faults includes: Based on the network topology information, the VoNR network state graph is merged to generate a target VoNR network state graph; Based on the target VoNR network state graph, coarse-grained fault prediction is performed using the trained network fault prediction model to obtain an initial prediction result characterizing whether there are abnormal network nodes in the target VoNR network state graph. If the initial prediction result indicates the existence of the abnormal network node, based on the initial prediction result, the forward network element device of the target network element device corresponding to the abnormal network node is simulated, and fine-grained fault prediction is performed by connecting the forward network element device to the target network element device to obtain the target prediction result.

3. The method according to claim 2, characterized in that, Before performing coarse-grained fault prediction using the trained network fault prediction model based on the target VoNR network state diagram, the method further includes: The historical network state information and corresponding node state labels of each network node in the target network at the target historical detection time are obtained, as well as the historical network state information at historical detection times before the target historical detection time; wherein, the target historical detection time is the historical detection time when the target network has a fault; the node state label indicates whether the network node has a fault; Based on the historical network state information at the target's historical detection time and the historical network state information at the historical detection time before the target's historical detection time, time slice data is generated, and the time slice data and the node state labels are stored in the network state database. Based on the time slice data and node status labels in the network state database, the network fault prediction model is trained to obtain the trained network fault prediction model.

4. The method according to claim 1, characterized in that, The step of obtaining network status information of network nodes in the target network under the VoNR service by simulating the simulated terminal device to perform VoNR service includes: The simulated terminal device initiates corresponding VoNR call requests under different VoNR service scenarios and receives signaling messages returned by the second network element device of the target network based on the VoNR call request; The signaling message is decoded to obtain the routing path of the second network element device through which each VoNR call request passes, and to obtain the network status information of the second network element device corresponding to the multiple routing paths; The step of generating a VoNR network state diagram based on the network state information includes: The VoNR network status diagram is generated based on the network status information and the routing path.

5. The method according to claim 2, characterized in that, The step of merging nodes in the VoNR network state graph based on the network topology information to generate a target VoNR network state graph includes: Based on the user network edge devices (CEs), firewalls, and / or data center locations connected to the network nodes in the network topology information, the VoNR network state diagram is merged to generate the target VoNR network state diagram.

6. The method according to claim 2, characterized in that, The trained network fault prediction model includes a graph convolutional neural network.

7. A network fault early warning device, characterized in that, The device includes: An acquisition module is used to acquire network topology information of a target network; wherein, the target network includes the core network in the VoNR service bearer network; The first simulation module is used to simulate a base station for accessing the target network and a terminal device for accessing the base station, to obtain a simulated base station and a simulated terminal device, and to connect the simulated terminal device to the simulated base station; The second simulation module is used to simulate the first network element device in the network topology information to obtain a simulated network element device, and to connect the simulated base station and the simulated network element device to the target network based on the network topology information. The generation module is used to simulate the simulated terminal device to perform VoNR service, obtain the network status information of the network nodes of the target network under the VoNR service, and generate a VoNR network status diagram based on the network status information. The prediction module is used to perform network fault prediction based on the VoNR network state diagram to obtain a target prediction result that characterizes whether the network nodes of the target network have faults.

8. An electronic device, characterized in that, include: processor; as well as A memory configured to store computer-executable instructions configured to be executed by the processor, the executable instructions including a method for performing a network fault warning as described in any one of claims 1-6.

9. A storage medium, characterized in that, The storage medium is used to store computer-executable instructions that cause the computer to perform the network fault early warning method as described in any one of claims 1-6.

10. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the network fault early warning method as described in any one of claims 1-6.

Images