Automatic fault detection and recovery method and system for FTTR network

Abstract

The invention relates to an FTTR network automatic fault detection and recovery method and system, and belongs to the technical field of network communication, and the method comprises the steps: constructing an FTTR network topology structure based on network nodes of each room; collecting network connection quality data of each network node; calculating a current state parameter of each network node and carrying out anomaly analysis; marking the network nodes with abnormal current state parameters as potential abnormal nodes, and determining corresponding abnormal levels; marking the potential abnormal nodes of which the abnormal levels exceed a preset level threshold value as current fault nodes; performing fault mode analysis on the current fault node, and determining a fault source; detecting a standby network path according to the current network path of the fault source, and switching the current network path of the fault source to the standby network path; and obtaining recovery result feedback information and updating the recovery result feedback information to the node fault historical database. According to the invention, efficient and accurate fault detection and recovery in the FTTR network can be realized, and the stability and reliability of the network are improved.

Description

Technical Field

[0001] The present application relates to the field of network communication technology, and in particular to a method and system for automatic fault detection and recovery of an FTTR network. Background Art

[0002] With the increasing popularity of smart homes and high-bandwidth applications, FTTR (fiber-to-the-room) network technology has become an important solution for achieving high-speed network connectivity in homes and businesses. However, in practice, the operation of FTTR networks faces various technical challenges, particularly in terms of network stability and troubleshooting.

[0003] In traditional FTTR networks, the number and types of network nodes are increasing. Wireless access points and other network devices deployed in various rooms create a complex network topology. Due to the diversity of devices and varying operating environments, network connection quality is susceptible to various factors, such as signal interference, device failures, and upstream network anomalies. These issues can lead to users experiencing poor network performance, such as connection interruptions and speed drops. Furthermore, when network failures occur, users often struggle to quickly locate the problem, resulting in extended network recovery times and impacting normal service. Therefore, achieving efficient and accurate fault detection and recovery in FTTR networks, while improving network stability and reliability, has become a pressing issue. Summary of the Invention

[0004] In order to improve the stability and reliability of the network, the present application provides a method and system for automatic fault detection and recovery of an FTTR network.

[0005] In a first aspect, the present application provides a method for automatic fault detection and recovery of an FTTR network, which adopts the following technical solution: A method for automatic fault detection and recovery of an FTTR network, the method comprising: Build FTTR network topology based on network nodes in each room; Collect network connection quality data of each network node in real time; Calculating a current state parameter of each network node based on the network connection quality data; Based on the node historical status database, performing abnormality analysis on the current status parameters of each of the network nodes to obtain abnormality analysis results; According to the abnormality analysis result, the network node with abnormal current state parameters is marked as a potential abnormal node, and the abnormality level corresponding to the potential abnormal node is determined; Marking the potential abnormal node whose abnormality level exceeds the preset level threshold as the current fault node; Based on the node fault history database and the FTTR network topology, performing a fault mode analysis on the current fault node to determine the corresponding fault source; Based on the FTTR network topology, detecting a backup network path according to the current network path of the fault source, and switching the current network path of the fault source to the backup network path; Obtain recovery result feedback information and update it to the node failure history database.

[0006] By adopting the above technical solution, in home networks and smart home environments, network connection quality data of network nodes in each room are collected in real time, and anomaly analysis is performed in combination with the historical status database, thus achieving fine-grained hierarchical detection of network nodes. During the fault analysis process, the historical fault data of nodes is used to accurately locate the source of the fault based on the FTTR network topology structure. Compared with existing technologies, the root cause of the problem can be identified more quickly and accurately, and the fault repair response speed can be improved. After a fault is detected, an alternative path is selected through the topology structure to ensure that the data flow automatically bypasses the faulty node and switches to an alternative path, thus achieving automatic detection and recovery of the faulty node, effectively reducing the impact of the fault on the user experience and improving the reliability and stability of the network.

[0007] Optionally, the step of performing a failure mode analysis on the current faulty node based on the node failure history database and the FTTR network topology to determine the corresponding fault source includes: Building and training a fault analysis model based on the node fault history database; Determining an upstream node of the current faulty node based on the FTTR network topology; Obtaining current state parameters of the current faulty node and the upstream node, and inputting the parameters into the fault analysis model for fault analysis to identify fault characteristics of the current faulty node and the upstream node; Determining whether the upstream node has a fault characteristic; If not, determining that the fault source is the current fault node; If so, determining each downstream node corresponding to the upstream node based on the FTTR network topology; Obtaining current state parameters of each of the downstream nodes and inputting them into the fault analysis model to perform fault analysis and identify fault characteristics of the downstream nodes; Counting the proportion of downstream nodes having fault characteristics among the downstream nodes, and if the proportion is greater than a preset proportion threshold, determining that the fault source is the upstream node; If the ratio is less than a preset ratio threshold, calculating the similarity of the fault characteristics of the current fault node and the upstream node; If the fault feature similarity is greater than a preset similarity threshold, determining that the fault source is the upstream node; If the fault feature similarity is less than a preset similarity threshold, it is determined that the fault source includes the current fault node and the upstream node.

[0008] By adopting the above technical solution, a fault analysis model based on historical data is established. Combined with the FTTR network topology, the status of the faulty node is gradually analyzed and determined, enabling precise location of the fault source. This technical solution not only improves the efficiency and accuracy of fault detection, but also avoids misjudgments by comprehensively analyzing the status of upstream and downstream nodes, ensuring high network availability and stability. By combining proportional statistics and similarity analysis, a multi-level basis for determining the fault source is provided. Proportional statistics provide a quantitative assessment of the fault scope, while similarity analysis further confirms the fault source. This logically clear and systematic judgment method ensures the precise location of the fault source in complex network environments, significantly improving fault response capabilities and making network management more intelligent and automated.

[0009] Optionally, the steps of constructing and training a fault analysis model based on the node fault history database include: Based on the node fault history database, a historical sample fault feature set is obtained and preprocessed; wherein the historical sample feature set includes node historical state parameters and corresponding fault feature labels; Perform feature extraction on the preprocessed historical sample feature set to obtain a feature vector set; Dividing the feature vector set into training samples and verification samples; Inputting the training samples into a pre-built random forest model for initial training, optimizing model parameters, and obtaining an initially trained fault analysis model; The fault analysis model is verified and model parameters are corrected based on the verification samples until the loss function of the model meets a preset condition or the number of model iterations reaches a preset number, thereby obtaining the trained fault analysis model.

[0010] By employing this technical solution, a fault analysis model built and trained based on a historical database of node faults can effectively learn fault characteristics from historical data, automatically identify potential fault sources, and improve the accuracy and efficiency of fault detection. This model is not only highly adaptable and can handle faults in a variety of network environments, but also provides important support for subsequent fault recovery and network optimization.

[0011] Optionally, when the fault source is the upstream node, the steps of performing backup network path detection according to the current network path of the fault source based on the FTTR network topology and switching the current network path of the fault source to the backup network path include: Determining each downstream node of the upstream node based on the FTTR network topology structure to obtain a current network path of the upstream node; wherein the current network path includes a network connection path between the upstream node and each downstream node; Obtaining the historical average traffic demand of the upstream node; Determining adjacent upstream nodes of the upstream node based on the FTTR network topology, and obtaining network status parameters of each of the adjacent upstream nodes; Performing traffic carrying analysis and performance evaluation analysis on each adjacent upstream node according to the network status parameters to obtain traffic carrying results and performance evaluation results of each adjacent upstream node; Filtering adjacent upstream nodes whose traffic carrying results meet the historical average traffic demand to obtain a list of candidate adjacent upstream nodes; sorting the candidate adjacent upstream node list according to the performance evaluation result, and selecting the adjacent upstream node with the best performance evaluation result in the candidate adjacent upstream node list as a backup adjacent upstream node; using the network connection path between the backup adjacent upstream node and each of the downstream nodes as the backup network path; The data flow of the upstream node is switched to the backup adjacent upstream node.

[0012] By adopting the above technical solution, when the fault source is an upstream node, efficient backup path switching is achieved through analysis and selection of the upstream node's adjacent upstream nodes, thereby ensuring that network services are quickly restored when an upstream node fails, reducing the impact on other downstream nodes, and improving the connection stability and user experience of home smart devices.

[0013] Optionally, the step of obtaining recovery result feedback information and updating the node fault history database includes: receiving recovery result feedback information of the current faulty node and the downstream node, and uploading the information to the node fault history database.

[0014] By adopting the above technical solution, after the data flow of the upstream node is switched to the backup adjacent upstream node, the current faulty node and the downstream nodes are switched to connect to the backup adjacent upstream node. Therefore, by obtaining the network recovery results of the current faulty node and its downstream nodes, the real-time and effectiveness of network fault processing are ensured.

[0015] Optionally, when the fault source is the current fault node, the steps of performing backup network path detection according to the current network path of the fault source based on the FTTR network topology and switching the current network path of the fault source to the backup network path include: Determining a current network path of the current fault node based on the FTTR network topology; wherein the current network path includes a network connection path between an upstream node of the current fault node and the current fault node; Obtaining the historical average traffic demand of the current fault node; Determine each downstream node corresponding to the upstream node based on the FTTR network topology, and obtain network status parameters of each downstream node; Performing traffic carrying analysis and performance evaluation analysis on each downstream node according to the network status parameters to obtain traffic carrying results and performance evaluation results of each downstream node; Filter downstream nodes whose traffic carrying results meet the historical average traffic demand to obtain a list of candidate downstream nodes; sorting the candidate downstream node list according to the performance evaluation result, and selecting the downstream node with the best performance evaluation result in the candidate downstream node list as a backup downstream node; Using the network connection path between the backup downstream node and the upstream node as the backup network path; The data flow of the current faulty node is switched to the backup downstream node.

[0016] By adopting this technical solution, when the fault source is the current faulty node, the system can quickly identify a suitable backup path based on the faulty node's location in the FTTR network topology and effectively switch traffic, thereby ensuring the continuity and reliability of network services. This structured fault handling approach not only improves fault response speed but also optimizes network resource utilization through historical data analysis and real-time monitoring, improving the overall user experience.

[0017] Optionally, when the fault source includes the current fault node and the upstream node, the steps of performing backup network path detection according to the current network path of the fault source based on the FTTR network topology and switching the current network path of the fault source to the backup network path include: Determining a current network path between the faulty node and the upstream node based on the FTTR network topology; wherein the current network path includes a network connection path between the current faulty node and the upstream node; Obtaining the historical average traffic demand of the current network path; Determine each adjacent upstream node of the upstream node based on the FTTR network topology, and determine corresponding adjacent downstream nodes based on the adjacent upstream nodes to obtain each adjacent network path; wherein the adjacent network path includes a network connection path between the adjacent upstream node and the corresponding adjacent downstream node; Acquiring network status parameters of each of the adjacent network paths, and performing traffic carrying analysis and performance evaluation analysis on each of the adjacent network paths to obtain traffic carrying results and performance evaluation results of each of the adjacent network paths; Filtering adjacent network paths whose traffic carrying results meet the historical average traffic demand to obtain a list of candidate adjacent network paths; sorting the candidate adjacent network path list according to the performance evaluation result, and selecting the adjacent network path with the best performance evaluation result in the candidate adjacent network path list as the backup network path; The data flow of the current network path between the faulty node and the upstream node is switched to the backup network path.

[0018] By adopting the above technical solution, when both the current faulty node and the upstream node are faulty, the network path of the current faulty node and its upstream node is identified, and historical traffic demand and status evaluation of adjacent network paths are combined to ensure that a suitable backup network path can be quickly switched to when a fault occurs. This not only improves network reliability, but also enhances the efficiency of fault recovery and minimizes the impact on user experience.

[0019] Optionally, the step of obtaining recovery result feedback information and updating the node fault history database includes: generating recovery result feedback information according to the network switching result, and updating the node fault history database.

[0020] By adopting the above technical solution, when the fault source is only the current fault node or includes the current fault node, the current fault node needs to be switched to other downstream nodes during recovery. After the switch is completed, there is no need to obtain the recovery result of the current fault node. The success or failure of the switching operation can be determined directly by obtaining the network switching result, and then the switching result is updated to the node fault history database, thereby realizing systematic management of network fault processing.

[0021] Optionally, after the step of marking the potential abnormal node whose abnormality level exceeds a preset level threshold as a current fault node, the method further includes: Acquire the household appliance corresponding to the current fault node according to the current fault node; Obtaining operating status parameters of the household appliance; determining whether the household appliance has a fault according to the working status parameter; If there is no fault, keep the mark of the current faulty node; If there is a fault, a maintenance prompt message is sent to the user terminal according to the location information of the home appliance; receiving home appliance recovery information fed back by the user terminal, performing fault analysis on the current fault node, and determining whether there are fault characteristics; If yes, keep the mark of the current faulty node; If not, the mark of the current fault node is canceled.

[0022] By adopting the above technical solutions, the system can effectively distinguish between network failures and home appliance failures, and quickly and accurately locate the source of the problem through status monitoring, feature matching, and cross-validation, thereby formulating targeted recovery strategies and improving the reliability and user experience of the smart home system.

[0023] In a second aspect, the present application provides an FTTR network automatic fault detection and recovery system, which adopts the following technical solutions: A FTTR network automatic fault detection and recovery system, configured to execute the FTTR network automatic fault detection and recovery method described in the first aspect, the system comprising: A network building module for building a FTTR network topology based on the network nodes in each room; Network data acquisition module, used to collect network connection quality data of each network node in real time; A state parameter calculation module, configured to calculate the current state parameter of each network node based on the network connection quality data; An abnormality analysis module is used to perform abnormality analysis on the current state parameters of each network node based on the node historical state database to obtain abnormality analysis results; an abnormal node marking module, configured to mark the network node with abnormal current state parameters as a potential abnormal node according to the abnormal analysis result, and determine the abnormality level corresponding to the potential abnormal node; A fault node marking module, configured to mark a potential abnormal node whose abnormality level exceeds a preset level threshold as a current fault node; A failure mode analysis module is used to perform a failure mode analysis on the current fault node based on a node failure history database and the FTTR network topology to determine the corresponding fault source; A backup network switching module, configured to detect a backup network path based on the current network path of the fault source and switch the current network path of the fault source to the backup network path based on the FTTR network topology; The recovery feedback information updating module is used to obtain the recovery result feedback information and update it to the node fault history database.

[0024] In summary, the present application includes at least one of the following beneficial technical effects: utilizing historical node fault data and accurately locating the source of the fault based on the FTTR network topology structure, it can identify the root cause of the problem more quickly and accurately compared to the existing technology, and improve the fault repair response speed; after detecting a fault, an alternative path is selected through the topology structure to ensure that the data flow automatically bypasses the faulty node and switches to an alternative path, realizing automatic detection and recovery of the faulty node, effectively reducing the impact of the fault on the user experience, and improving the reliability and stability of the network. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a first flow chart of a method for automatic fault detection and recovery in an FTTR network according to one embodiment of the present application.

[0026] Figure 2 This is a second flow chart of a method for automatic fault detection and recovery in an FTTR network according to one embodiment of the present application.

[0027] Figure 3 This is a third flow chart of an FTTR network automatic fault detection and recovery method according to one embodiment of the present application.

[0028] Figure 4 This is a fourth flow chart of a method for automatic fault detection and recovery in an FTTR network according to one embodiment of the present application.

[0029] Figure 5 This is a fifth flow chart of an FTTR network automatic fault detection and recovery method according to one embodiment of the present application.

[0030] Figure 6 This is a sixth flow chart of an FTTR network automatic fault detection and recovery method according to one embodiment of the present application.

[0031] Figure 7 This is the seventh flow chart of a method for automatic fault detection and recovery in an FTTR network according to one embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-7 It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0033] With the rapid development of smart homes and the Internet of Things (IoT), FTTR (Fiber To The Room) networks are widely used in homes and offices due to their high bandwidth and low latency. However, frequent network failures severely impact user experience and device performance.

[0034] Based on this, an embodiment of the present application discloses a method for automatic fault detection and recovery in an FTTR network.

[0035] Reference Figure 1 , a FTTR network automatic fault detection and recovery method, comprising: Step S101, constructing an FTTR network topology structure based on the network nodes in each room; Fiber to the Room (FTTR) networks typically consist of multiple network nodes (such as wireless access points or fiber optic terminals) distributed across various rooms. By analyzing the network node information in each room, a complete network topology, i.e., the connection relationships between nodes, can be established.

[0036] For example, in a home environment, network nodes are set up in the living room, bedroom and study respectively, and each network node is connected to the central router. The system forms a home network topology by recording the connection status and position relationship of each network node.

[0037] As you can see, building an FTTR network topology provides the foundational information for subsequently detecting the connection status of each node, identifying potential anomalies, and implementing path switching. Through this precise network structure, the system can quickly locate each room's nodes and their upstream relationships, improving the accuracy of fault location and path switching.

[0038] Step S102, collecting network connection quality data of each network node in real time; This involves real-time monitoring of each network node's network connection quality data, such as signal strength, latency, and jitter, to continuously monitor network health. This data serves as a basis for determining network health and detecting anomalies. For example, a network node in a room records signal strength, latency, and jitter every second, which is then transmitted to a central controller via the network. If the signal strength at a given moment falls below a preset threshold, the system records that node as having poor signal quality.

[0039] Step S103, calculating the current state parameters of each network node based on the network connection quality data; By processing the collected network connection quality data, the system calculates each node's current status parameters, such as average signal strength and average latency. This simplifies the data, making it easier for the system to determine node status. For example, if the average signal strength of a network node in a room over the past 10 seconds was -60dBm and the average latency was 50ms, the system would set the node's current status parameters to these two values ​​to facilitate comparison with historical data.

[0040] It is understandable that by calculating the current status parameters, complex real-time data can be converted into concise indicators, which makes it easier for the system to quickly analyze whether the node is operating normally, helping to shorten the time for fault diagnosis.

[0041] Step S104, based on the node historical status database, performing abnormality analysis on the current status parameters of each network node to obtain abnormality analysis results; The system compares each node's current state parameters with data in a historical state database, calculating the difference from the historical average to determine whether the current parameters are abnormal. For example, if the historical average latency of a network node in a room is 30ms, but the current latency is 50ms, the system will determine that the latency is abnormal and mark the node as a potential abnormal node.

[0042] Step S105: Based on the abnormality analysis results, the network nodes with abnormal current state parameters are marked as potential abnormal nodes, and the abnormality levels corresponding to the potential abnormal nodes are determined; Specifically, based on the anomaly analysis results, the system labels any network nodes detected as "potentially abnormal nodes." The system also assigns a level to these nodes based on the severity of the abnormality (e.g., the size of the difference) to facilitate subsequent troubleshooting. For example, if the latency parameter of a network node in a room is within 30% of the historical average, it is marked as "mildly abnormal." If it is above 50%, it is marked as "severely abnormal."

[0043] As you can see, by categorizing abnormal nodes into different levels—for example, mild (network fluctuations), moderate (network interruptions), and severe (complete loss of network connectivity)—the system can more effectively determine the severity of the fault and take appropriate recovery measures. This grading mechanism ensures that fault handling resources are allocated appropriately.

[0044] Step S106: Mark the potential abnormal node whose abnormality level exceeds the preset level threshold as a current fault node; When the abnormality level of a potential abnormal node exceeds a preset threshold, the system identifies it as a current fault node and makes it the direct target of subsequent fault handling. For example, if the latency abnormality of a network node in a room exceeds the severe abnormality threshold of 50%, the system will mark it as a "current fault node."

[0045] It should be noted that the preset level threshold can be pre-set according to the actual situation. By setting the threshold, the faulty nodes that need to be dealt with immediately can be accurately screened out to ensure that the most serious network problems are solved first when resources are limited.

[0046] Step S107: Based on the node fault history database and the FTTR network topology, a fault mode analysis is performed on the current fault node to determine the corresponding fault source; The fault source includes the current fault node or the upstream node of the current fault node; In some embodiments, the system can pre-establish a fault analysis model using a fault history database. This model, combined with the FTTR network topology, analyzes the faulty node's fault pattern to determine whether the fault lies at the current node or its upstream node. Because an upstream node failure can cause other downstream nodes to malfunction, it's important to distinguish the source of the fault between the current node and the upstream node to improve repair efficiency and avoid wasted resources due to misjudgments. For example, if the system detects similar abnormal patterns on a network node in a particular room and an upstream node (router), the upstream node is likely at fault.

[0047] Step S108: Based on the FTTR network topology, perform backup network path detection according to the current network path of the fault source, and switch the current network path of the fault source to the backup network path; After confirming the source of a fault, the system detects viable backup paths based on the network topology, ensuring that data flows can maintain connectivity via these alternative paths. This provides an alternative solution for fault recovery, ensuring that network services can be restored promptly in the event of a node failure, minimizing the impact on users. For example, if a network node in a room fails and is identified as the source of the fault, the system will check the connectivity of other adjacent nodes with the central routing to determine whether a viable alternative path exists.

[0048] It should be noted that when determining backup network paths, it is also necessary to perform a performance evaluation on all backup network paths. By using the collected network status parameters (such as bandwidth, latency, load, etc.) for analysis and comparing the performance data of each backup network path, the backup network path that can meet traffic requirements and has good performance is prioritized based on the performance evaluation results.

[0049] Furthermore, once the backup path is determined, the system will switch the network connection of the fault source to the backup network path to restore the normal operation of the network. This process can be achieved through dynamic routing or redirection technology. By automatically switching to the backup path, the system can quickly restore the network connection and improve the user experience.

[0050] In some specific embodiments, if the fault source is an upstream node of the currently faulty node, then when switching to an alternate network path, the data flow of the currently faulty node can be switched from the path of the faulty upstream router node to the network path of another, normally functioning, adjacent upstream router node, allowing data to be transmitted to the core network via the alternate path. If the network of the upstream node of the currently faulty node is normal, and the network of other downstream nodes of the upstream node is also normal, and all other downstream nodes are normally connected to the upstream node, then the fault source is the currently faulty node. In this case, when switching to an alternate network path, the data flow can be switched from the currently faulty node to another, normally functioning downstream node. This can be achieved by setting a dynamic routing policy or modifying the network configuration.

[0051] In addition, if both the current faulty node and the upstream node are faulty, the fault source includes both the current faulty node and the upstream node. When switching to the backup network path, the data flow can be switched from the current faulty node to other normal downstream nodes to ensure that the user's network experience is not affected. At the same time, the data flow of the upstream node can be switched to other normal routers or switches to prevent its failure from affecting other parts of the network.

[0052] Step S109: Obtain recovery result feedback information and update it to the node failure history database.

[0053] Specifically, after network path switching, recovery result feedback information is obtained, such as the fault source location, fault level, switching path, network recovery result, recovery time, etc., and this information is recorded in the node fault history database to update historical fault records, forming a data closed loop to facilitate preventive maintenance of future network faults and improve the system's fault tracing and analysis capabilities.

[0054] In the above implementation, in a home network and smart home environment, the network connection quality data of the network nodes in each room are collected in real time, and anomaly analysis is performed in combination with the historical status database, thereby realizing fine-grained hierarchical detection of network nodes; in the fault analysis process, the historical fault data of the node is used to accurately locate the source of the fault based on the FTTR network topology structure, which can identify the root cause of the problem more quickly and accurately than the existing technology, thereby improving the fault repair response speed; after a fault is detected, an alternative path is selected through the topology structure to ensure that the data flow automatically bypasses the faulty node and switches to an alternative path, thereby realizing automatic detection and recovery of the faulty node, effectively reducing the impact of the fault on the user experience, and improving the reliability and stability of the network.

[0055] In addition, the final uploaded recovery result feedback information provides data support for subsequent fault prediction and mode optimization, giving the system the potential for self-learning and optimization, and ensuring the long-term stability of the FTTR network.

[0056] Reference Figure 2 As an implementation of step S107, based on the node fault history database and the FTTR network topology, the steps of performing a fault mode analysis on the current fault node and determining the corresponding fault source include: Step S201: construct and train a fault analysis model based on a node fault history database; Specifically, a machine learning model is built using data from a node fault history database to identify fault characteristics and patterns. By training the model, the system can automatically learn and adapt to various fault scenarios. This provides a data-driven analysis tool that effectively identifies potential fault sources and improves the accuracy of fault detection.

[0057] For example, by using historical failure data, the model can learn the failures that are often accompanied by packet loss or connection interruption when the network delay exceeds a certain threshold.

[0058] Step S202: determining an upstream node of the current faulty node based on the FTTR network topology; An upstream node refers to a node on the data transmission path within a data flow, responsible for sending data to downstream nodes. Because the status of an upstream node can directly affect downstream nodes, identifying the immediate upstream node of a faulty node based on the FTTR network topology facilitates subsequent fault source analysis. For example, in a three-layer network, if the faulty node is located in Room A, its upstream node might be Router B.

[0059] Step S203: obtaining current state parameters of the current faulty node and upstream nodes, and inputting them into a fault analysis model for fault analysis to identify fault characteristics of the current faulty node and upstream nodes; Specifically, the status parameters of the current node and upstream nodes, such as delay and packet loss rate, are monitored in real time, and these parameters are input into the fault analysis model for analysis to identify fault characteristics.

[0060] Step S204, determine whether the upstream node has fault characteristics; if not, jump to step S205; if so, jump to step S206; The direction of subsequent analysis is determined by quickly identifying the status of the upstream node. For example, if the model determines that the delay of the upstream node exceeds the normal range, it is considered that the upstream node has fault characteristics.

[0061] Step S205, determining that the fault source is the current fault node; Specifically, if the upstream node is normal, the fault source is directly determined to be the current fault node. By quickly locking the fault source, subsequent analysis steps are reduced and response efficiency is improved.

[0062] Step S206, determining each downstream node corresponding to the upstream node based on the FTTR network topology; Among them, the downstream node refers to the node that receives data from the upstream node, which is usually the node closer to the user device. If there is an abnormality in the upstream node, all downstream nodes corresponding to the upstream node are further analyzed to further determine the impact range of the upstream node and determine whether the current faulty node is caused by the abnormality of the upstream node.

[0063] Step S207: obtaining the current state parameters of each downstream node and inputting them into the fault analysis model to perform fault analysis and identify the fault characteristics of the downstream node; Specifically, the state parameters of all downstream nodes are obtained and input into the model for fault analysis, providing data support for subsequent fault source determination; Step S208, counting the proportion of downstream nodes with fault characteristics among the downstream nodes; Step S209, determining whether the ratio is greater than a preset ratio threshold, if so, jump to step S210; if not, jump to step S211; Specifically, the ratio of downstream nodes exhibiting fault characteristics to all downstream nodes is calculated. This quantifies the impact of upstream node failures and helps to more accurately determine the source of the fault. The preset ratio threshold can be pre-configured and adjusted based on actual conditions, for example, it can be set to 50%.

[0064] Step S210, determining that the fault source is an upstream node; In some embodiments, if four out of five downstream nodes display fault characteristics (e.g., exhibiting delays and packet loss, etc.), the ratio is 80%, which exceeds a preset ratio threshold. In this case, it can be determined that the fault of the upstream node has a wide impact, and the upstream node can be determined to be the source of the fault.

[0065] Step S211, calculating the similarity of the fault characteristics of the current fault node and the upstream node; Among them, the fault feature similarity can be based on the calculation of state parameters (such as delay, packet loss, etc.) to reflect whether the fault manifestations of two nodes are consistent.

[0066] It can be understood that if the abnormal proportion of downstream nodes is less than the preset proportion threshold, it is inferred that the upstream node may have a local fault or an occasional fault, which has not affected most of the downstream nodes. Therefore, the fault source is further confirmed by calculating the similarity of the fault characteristics of the current fault node and the upstream node.

[0067] Step S212, determine whether the fault feature similarity is greater than a preset similarity threshold; if so, jump to step S210; if not, jump to step S213; Specifically, the preset similarity threshold can be pre-configured and adjusted based on actual conditions. If the fault signatures of the current faulty node and the upstream node are similar, it can be determined that the current faulty node was affected by the upstream node, and therefore the upstream node is the fault source. For example, if the fault signature similarity between the two nodes is calculated to be 75% and the similarity threshold is set to 70%, the upstream node can be determined to be the fault source.

[0068] It can be understood that similarity analysis ensures that the judgment of the fault source does not only rely on absolute data, but also considers the similarity of fault manifestations between different nodes, making the analysis more comprehensive and accurate, and improving the reliability of fault source identification.

[0069] Step S213: Determine that the fault source includes the current faulty node and the upstream node.

[0070] Specifically, if the abnormality ratio of downstream nodes does not exceed the threshold, and the similarity of the fault characteristics of the current faulty node and the upstream node does not exceed the threshold, this indicates that both the current faulty node and the upstream node are likely to be the source of the fault. This dual judgment mechanism can more accurately locate the source of the fault, avoid misjudgments and missed detections, and ensure network reliability.

[0071] In the above implementation, a fault analysis model based on historical data is established. Combined with the FTTR network topology, the status of the faulty node is gradually analyzed and determined, enabling the precise location of the fault source. This technical solution not only improves the efficiency and accuracy of fault detection, but also avoids misjudgments by comprehensively analyzing the status of upstream and downstream nodes, ensuring high network availability and stability. By combining proportional statistics with similarity analysis, a multi-level basis for determining the source of the fault is provided. Proportional statistics provide a quantitative assessment of the fault scope, while similarity analysis further confirms the attribution of the fault source. This logically clear and systematic judgment method ensures the precise location of the fault source in complex network environments, significantly improving fault response capabilities and making network management more intelligent and automated.

[0072] It should be noted that when a local fault occurs in an upstream node, such as a configuration error, temporary network congestion, or hardware problem, these problems may not affect the downstream node; for example, if the upstream router temporarily loses packets due to overload, but the connection to the downstream node is normal, the downstream node may not show any abnormal characteristics. In addition, the fault characteristics of the upstream node may also be intermittent, such as network interference or signal instability, which leads to occasional abnormalities. If the downstream node is in a normal state during monitoring, the downstream node will not show abnormalities. Special situations similar to the above are a challenge for network management and fault detection. Since they may lead to misjudgment and neglect of downstream nodes, in order to accurately diagnose the source of the fault, the present application can detect and analyze this situation and combine multiple monitoring indicators to comprehensively evaluate the network status.

[0073] Reference Figure 3 As an implementation of step S201, the steps of constructing and training a fault analysis model based on a node fault history database include: Step S301: Based on the node fault history database, a historical sample fault feature set is obtained and preprocessed; Among them, the historical sample feature set includes the node historical state parameters and the corresponding fault feature labels; Specifically, relevant historical data is extracted from the node fault history database, including node status parameters (such as delay, packet loss rate, bandwidth, etc.) and corresponding fault feature labels (such as normal, fault type, etc.).

[0074] In addition, the preprocessing step is mainly to clean the data, remove noise and outliers, ensure that the samples used for training are of high quality and accuracy, and lay a good foundation for subsequent feature extraction and model training.

[0075] Step S302: extracting features from the pre-processed historical sample feature set to obtain a feature vector set; Among them, key features are extracted from the cleaned data to form feature vectors. This step can determine important features through statistical analysis, correlation analysis and domain knowledge.

[0076] For example, by analyzing node status parameters, we identify "network delay," "packet loss rate," and "bandwidth utilization" as important features, forming a feature vector. For example, feature vector = [delay, packet loss rate, bandwidth utilization]. By extracting effective feature vectors, we can enhance the model's ability to predict faults and better identify potential faults.

[0077] Step S303, dividing the feature vector set into training samples and verification samples; Specifically, the feature vector set is divided into different data sets to ensure the effectiveness of model training, verification and final testing. Generally, 70% is used as training samples and 30% is used as verification samples. By reasonably dividing the data set, overfitting can be effectively avoided and the model's generalization ability on unseen data can be ensured.

[0078] Step S304: input the training samples into a pre-built random forest model for initial training, optimize the model parameters, and obtain an initially trained fault analysis model; Specifically, the random forest algorithm is used to perform preliminary training on the training samples. The random forest model improves the accuracy of prediction by integrating multiple decision trees. By inputting the training samples into the model and randomly selecting features to construct multiple decision trees, the model can obtain the fault prediction results for each sample through a voting mechanism and optimize the model parameters based on the results, so that the model can learn the patterns in the data, thereby improving the ability to identify faults.

[0079] In one of the embodiments of the present application, a random forest model is pre-constructed as the basis of the fault analysis model. The random forest model can help identify the key parameters that cause faults by analyzing the importance of features, provide a clear basis for fault analysis, and can be used for classification and regression tasks. It has strong adaptability and can be adjusted as needed to meet different analysis requirements.

[0080] Specifically, the model structure of the fault analysis model based on the random forest model in the embodiment of the present application mainly includes the following components: a data output layer for receiving preprocessed feature vectors, including network node status parameters (such as latency, packet loss rate, bandwidth utilization, etc.); a feature selection mechanism: during training, the random forest model randomly selects feature subsets to construct multiple decision trees, thereby increasing model diversity and reducing correlation; a decision tree ensemble layer, consisting of multiple decision trees, each of which independently learns and makes predictions based on random sampling and feature selection of its training set; a voting mechanism: each decision tree performs classification or regression on the input sample, ultimately obtaining the prediction result of the overall model through voting (classification) or averaging (regression); an output layer for generating fault analysis results, including the predicted fault probability and fault type label (such as "normal," "delay fault," "packet loss fault," etc.) for each node; and model evaluation and optimization: cross-validation and validation set are used to evaluate model performance, and parameter tuning (such as the number and depth of trees) is performed based on indicators such as model accuracy and recall.

[0081] Among them, the random forest model was selected for fault analysis due to its high accuracy, noise resistance and interpretability. Its model structure constitutes an effective fault analysis system through data input layer, feature selection mechanism, decision tree integration layer, voting mechanism and output layer.

[0082] Step S305 , verifying the fault analysis model and correcting model parameters based on the verification sample until the loss function of the model meets the preset conditions or the number of model iterations reaches the preset number, thereby obtaining a trained fault analysis model.

[0083] Validation samples are used to evaluate model performance and are adjusted based on the loss function. This process continues until the preset conditions are met or the maximum number of iterations is reached. Through continuous validation and parameter correction, the model's performance consistency across different datasets is ensured, thereby improving its robustness and accuracy. For example, if the model's accuracy on the validation samples falls below 80% during validation, model parameters (such as the number and depth of trees) are adjusted and the model retrained until accuracy improves.

[0084] In the above implementation, the fault analysis model built and trained based on the node fault history database can effectively learn fault characteristics from historical data, automatically identify potential fault sources, and improve the accuracy and efficiency of fault detection. This model is not only highly adaptable and can handle faults in various network environments, but also provides important support for subsequent fault recovery and network optimization.

[0085] Reference Figure 4As one implementation of step S108, when the fault source is an upstream node, based on the FTTR network topology, performing backup network path detection according to the current network path of the fault source and switching the current network path of the fault source to the backup network path includes: Step S401, determining each downstream node of the upstream node based on the FTTR network topology structure, and obtaining the current network path of the upstream node; The current network path includes the network connection path between the upstream node and each downstream node; Specifically, obtaining the current network path of the upstream node is to clarify the data flow direction and its impact range, providing a basis for subsequent traffic analysis and path switching.

[0086] Step S402, obtaining the historical average traffic demand of the upstream node; Among them, through historical data, the system can understand the traffic demand of upstream nodes during normal operation, which is the key to evaluating whether the backup path can meet the demand.

[0087] Step S403: determining adjacent upstream nodes of the upstream node based on the FTTR network topology, and obtaining network status parameters of each adjacent upstream node; Among them, adjacent upstream nodes refer to other upstream nodes that are directly connected to the currently faulty upstream node. These nodes can also transmit data to downstream nodes and maintain connections with other upstream nodes. Specifically, network status parameters such as latency, bandwidth, packet loss rate, etc., can help with subsequent traffic carrying and performance evaluation analysis by fully understanding the status of adjacent upstream nodes.

[0088] Step S404: performing traffic carrying analysis and performance evaluation analysis on each adjacent upstream node according to the network state parameters to obtain traffic carrying results and performance evaluation results of each adjacent upstream node; The selected backup path is determined to be able to support traffic demand and meet performance requirements based on the traffic carrying results and performance evaluation results of the adjacent upstream nodes.

[0089] Step S405 , screening adjacent upstream nodes whose traffic carrying results meet the historical average traffic demand, and obtaining a list of candidate adjacent upstream nodes; Among them, the selection range is narrowed by filtering the traffic carrying results, thereby improving the success rate after switching.

[0090] Step S406 , sorting the candidate adjacent upstream node list according to the performance evaluation result, and selecting the adjacent upstream node with the best performance evaluation result in the candidate adjacent upstream node list as a backup adjacent upstream node; The performance of the selected candidate adjacent upstream nodes is evaluated, including bandwidth utilization, delay, and packet loss rate. The optimal node is selected as the backup adjacent upstream node by sorting the performance evaluation results.

[0091] Step S407, using the network connection path between the backup adjacent upstream node and each downstream node as a backup network path; It should be noted that the downstream node in the above steps refers to the downstream node corresponding to the currently faulty upstream node, and the backup network path here means that after the currently faulty upstream node is replaced with the backup adjacent upstream node, each downstream node must be connected to the backup adjacent upstream node to ensure that the traffic can be transmitted smoothly, thereby forming a backup network path.

[0092] Step S408: Switch the data flow of the upstream node to the backup adjacent upstream node.

[0093] After determining the backup network path, the data flow of the upstream node is switched to the new backup adjacent upstream node to ensure the continuity of the network service.

[0094] In the above implementation, when the fault source is an upstream node, efficient backup path switching is achieved by analyzing and selecting the adjacent upstream nodes of the upstream node, thereby ensuring that network services are quickly restored when a fault occurs in the upstream node, reducing the impact on other downstream nodes, and improving the connection stability and user experience of home smart devices.

[0095] Furthermore, as an implementation method of step S109, when the fault source is an upstream node, the step of obtaining recovery result feedback information and updating it to the node fault history database includes: receiving recovery result feedback information of the current fault node and the downstream node, and uploading it to the node fault history database.

[0096] In the above embodiment, after the data flow of the upstream node is switched to the backup adjacent upstream node, the current faulty node and the downstream nodes are both switched to connect to the backup adjacent upstream node. Therefore, by obtaining the network recovery results of the current faulty node and its downstream nodes, the real-time and effectiveness of network fault processing are ensured.

[0097] It should be noted that recording this information in the node fault history database not only provides data support for subsequent fault analysis, but also provides an important basis for optimizing network management and improving network reliability. By continuously updating the database, the operations team can identify and resolve potential network problems in advance, enhancing the overall stability and reliability of the network.

[0098] Reference Figure 5As another implementation of step S108, when the fault source is the current fault node, the steps of detecting a backup network path according to the current network path of the fault source based on the FTTR network topology and switching the current network path of the fault source to the backup network path include: Step S501, determining the current network path of the current fault node based on the FTTR network topology; The current network path includes the network connection path between the upstream node of the current fault node and the current fault node; Specifically, the FTTR (Fiber To The Room) network topology is used to identify the location of the faulty node and its upstream node. For example, if the faulty node is an access point (AP) in a room, its upstream node is the router connected to the AP.

[0099] Step S502: Obtain the historical average traffic demand of the current fault node; Specifically, historical data can be analyzed to calculate the traffic demand of the currently faulty node. This information helps in subsequent steps to assess the traffic carrying capacity of the backup node. For example, the average traffic volume of the node over a period of time can be calculated by dividing the total traffic volume by the length of the period.

[0100] Step S503: determining each downstream node corresponding to the upstream node based on the FTTR network topology, and obtaining network status parameters of each downstream node; Among them, by tracing back from the upstream node, the downstream nodes directly connected to it are identified, and the status parameters of these nodes (such as latency, bandwidth, packet loss rate, etc.) are monitored in real time. By fully understanding the status of the downstream nodes, subsequent traffic carrying and performance evaluation analysis are facilitated.

[0101] Step S504: performing traffic carrying analysis and performance evaluation analysis on each downstream node according to the network status parameters to obtain traffic carrying results and performance evaluation results of each downstream node; Among them, the collected network status parameters are analyzed to evaluate the traffic carrying capacity of each downstream node (that is, to determine the actual available bandwidth, which can be obtained by subtracting the bandwidth in use data from the total bandwidth data) and comprehensive performance indicators (which can be obtained by integrating status parameters such as delay, packet loss rate, and bandwidth).

[0102] Specifically, multiple key performance indicators (KPIs) can be defined, such as bandwidth utilization, latency, and packet loss rate, with weights assigned to each. Next, network status parameters for each downstream node are collected in real time, and standardized values ​​for each metric are calculated. These standardized values ​​are then combined using a weighted average method to generate a comprehensive performance score reflecting the node's overall performance. This comprehensive score will help quickly identify high-performing downstream nodes during fault recovery, supporting traffic switching decisions.

[0103] Step S505 , screening downstream nodes whose traffic carrying results meet the historical average traffic demand, and obtaining a candidate downstream node list; Among them, nodes that can meet historical traffic requirements are screened out from all downstream nodes to form a candidate list to ensure that the network path after switching can meet the traffic requirements after network recovery.

[0104] Step S506: sort the candidate downstream node list according to the performance evaluation result, and select the downstream node with the best performance evaluation result in the candidate downstream node list as a backup downstream node; Specifically, candidate downstream nodes are sorted and the nodes with the best performance are selected to ensure that a reliable path is selected during fault recovery, thereby improving network stability and user experience.

[0105] Step S507: Using the network connection path between the backup downstream node and the upstream node as a backup network path; After the backup node is confirmed, a backup network path is formed based on the connection between the backup node and the upstream node.

[0106] Step S508: Switch the data flow of the current faulty node to the backup downstream node.

[0107] Specifically, a traffic switching operation is performed to redirect the data flow of the current faulty node to the selected backup downstream node, providing users with a seamless network switching experience and ensuring that users continue to enjoy network services, thereby reducing the impact of partial node failures on user experience.

[0108] In the above implementation, when the fault source is the currently faulty node, the system can quickly identify a suitable backup path based on the faulty node's location in the FTTR network topology and effectively switch traffic, thereby ensuring the continuity and reliability of network services. This structured fault handling approach not only improves fault response speed but also optimizes network resource utilization through historical data analysis and real-time monitoring, improving the overall user experience.

[0109] Reference Figure 6As another embodiment of step S108, when the fault source includes the current faulty node and an upstream node, the steps of detecting an alternative network path based on the current network path of the fault source and switching the current network path of the fault source to the alternative network path based on the FTTR network topology include: Step S601, determining the current network path between the faulty node and the upstream node based on the FTTR network topology; The current network path includes the network connection path between the current fault node and the upstream node; Specifically, in an FTTR network, the network connection path between the currently faulty node (e.g., an access point) and the upstream node (e.g., the main router) needs to be clearly identified. By analyzing the network topology, the connection relationship between these nodes can be determined. By clarifying the network path to the fault source, this provides a foundation for subsequent steps and helps quickly locate the problem.

[0110] For example, assuming that the upstream node is router A and the current faulty node is access point B, the current network path is: router A→access point B→smart device C.

[0111] Step S602, obtaining the historical average traffic demand of the current network path; Historical traffic data helps the system determine the bandwidth required under normal circumstances, allowing it to select an appropriate backup path when a failure occurs. By using historical average traffic demand as a reference, the backup path can ensure that it can meet traffic demands during failure recovery, avoiding performance bottlenecks.

[0112] Step S603: determining adjacent upstream nodes of the upstream node based on the FTTR network topology, and determining corresponding adjacent downstream nodes according to the adjacent upstream nodes to obtain adjacent network paths; Wherein, the adjacent network path includes a network connection path between an adjacent upstream node and a corresponding adjacent downstream node; Specifically, in the FTTR network topology, adjacent network paths are composed of adjacent upstream nodes of the upstream node and adjacent downstream nodes (i.e., downstream nodes of the adjacent upstream node). By analyzing the topology, these nodes are identified in order to find backup paths, thereby broadening the possible backup path options and improving network redundancy.

[0113] For example, if the adjacent upstream node of router A is router D, and router D is connected to access point E, then the adjacent network path is: router D→access point E→smart device F.

[0114] Step S604: Acquire network status parameters of each adjacent network path, and perform traffic carrying analysis and performance evaluation analysis on each adjacent network path to obtain traffic carrying results and performance evaluation results of each adjacent network path; The status of each adjacent network path is monitored, and its traffic carrying capacity and performance are evaluated to determine whether it is suitable as a backup path. For example, traffic carrying analysis and performance evaluation can be performed by collecting network status parameters (such as latency, bandwidth utilization, packet loss rate, etc.) to facilitate making more informed decisions and ensure that the selected backup path has good performance.

[0115] Step S605 , screening adjacent network paths whose traffic carrying results meet the historical average traffic demand, and obtaining a candidate adjacent network path list; After evaluating the traffic carrying capacity of adjacent network paths, it is necessary to screen out paths that can meet historical traffic needs to ensure that there is no bandwidth shortage during switching. That is, to ensure that when the fault is restored, the traffic carrying capacity of the backup path is consistent with user needs to avoid service interruption.

[0116] Step S606: sort the candidate adjacent network path list according to the performance evaluation result, and select the adjacent network path with the best performance evaluation result in the candidate adjacent network path list as the backup network path; Among them, according to the performance evaluation results, the candidate paths that meet the traffic requirements are sorted to select the adjacent network path with the best performance as the backup path to ensure that the network still maintains efficient operation after switching.

[0117] Step S607: Switch the data flow of the current network path between the faulty node and the upstream node to the backup network path.

[0118] After determining the backup path, the data flow of the current faulty node and the upstream node needs to be switched to the new backup network path to achieve fault recovery, reduce the impact of the fault on users, and ensure stable system operation.

[0119] In the above implementation, when both the current faulty node and the upstream node are faulty, the network paths of the current faulty node and its upstream nodes are identified, and the historical traffic demand and the status evaluation of the adjacent network paths are combined to ensure that a suitable backup network path can be quickly switched to when a fault occurs. This not only improves the reliability of the network, but also enhances the efficiency of fault recovery and minimizes the impact on user experience.

[0120] Furthermore, as another implementation method of step S109, when the fault source is the current fault node or the fault source includes the current fault node and the upstream node, the step of obtaining recovery result feedback information and updating it to the node fault history database includes: generating recovery result feedback information based on the network switching result, and updating it to the node fault history database.

[0121] In the above implementation, when the fault source is only the current fault node or includes the current fault node, the current fault node needs to be switched to other downstream nodes during recovery. After the switch is completed, there is no need to obtain the recovery result of the current fault node. The success or failure of the switching operation can be determined directly by obtaining the network switching result, and then the switching result is updated to the node fault history database, thereby realizing systematic management of network fault processing.

[0122] Reference Figure 7 As a further embodiment of the present application, after marking the potential abnormal node whose abnormality level exceeds the preset level threshold as the current fault node in step S106, the following steps are further included: Step S701, obtaining the home appliance corresponding to the current fault node according to the current fault node; Among them, the current fault node can be a subnet segment of the home network, and the home appliances directly connected to the current fault node, such as smart TVs, air conditioners, etc., are identified. By identifying these devices as targets that need attention, subsequent fault analysis can target specific devices and concentrate resources to inspect and handle possible fault sources.

[0123] Step S702, obtaining the working status parameters of the home appliance; Specifically, real-time monitoring of the status parameters of household appliances can obtain the operating parameters of the equipment through network protocols or local sensors, such as power status, response time, working mode, etc., thereby providing necessary data support for judging equipment failures and ensuring that the basis for judgment is the latest and most accurate.

[0124] Step S703, judging whether the household appliance has a fault according to the working status parameters; if not, jumping to step S704; if so, jumping to step S705; Specifically, the set thresholds or models can be applied to analyze the working status parameters of household appliances to determine whether the equipment has abnormalities or faults. The accuracy of fault identification can be improved through data-driven methods, the risk of misjudgment can be reduced, and effective fault management can be ensured.

[0125] In some embodiments, when certain smart home appliances malfunction, the following abnormal conditions may occur: abnormal increase in network traffic or sending of a large number of invalid data packets, causing network node overload, thereby affecting its normal operation; causing network address conflicts (such as multiple devices requesting the same IP address), which will make the network node unable to correctly identify and process data streams, resulting in network failures; internal failures of home appliances lead to unstable power supply, which may affect network devices in the same circuit (such as routers or access points) and cause them to malfunction; error signals at the protocol layer may be generated, which will interfere with the normal communication of network nodes and cause the network nodes to fail to work normally.

[0126] Step S704: Keep the mark of the current fault node; Specifically, if it is determined that there is no abnormality in the home appliance, it is determined that the fault is in the network node itself (such as router hardware failure, firmware problem, etc.), and the node continues to be marked and waits for subsequent repair.

[0127] Step S705: Sending maintenance reminder information to the user terminal according to the location information of the home appliance; Specifically, once a fault is identified, the system should immediately notify the user, using the device's location information to generate a specific notification so that the user can take timely action. For example, if a home appliance fails, the user may need to be guided to restart, update firmware, or perform repairs.

[0128] It is understandable that by promptly detecting home appliance faults and providing troubleshooting prompts, user experience and satisfaction can be improved, and the interaction between users and smart home systems can be enhanced, enabling them to deal with problems in a timely manner.

[0129] Step S706: Receive home appliance recovery information fed back by the user terminal and perform fault analysis on the current fault node; Step S707, determine whether there is a fault feature; if so, jump to step S704; if not, jump to step S708; Specifically, after the user repairs and restores the home appliance, the system receives feedback from the user that the appliance has returned to normal, records this information, and checks whether the current faulty node still displays a faulty status.

[0130] Among them, if the home appliance has been repaired and restored to normal, but the current faulty node still has fault characteristics, it is determined that the fault is in the network node itself (such as router hardware failure, firmware problem, etc.), and the node will continue to be marked and wait for subsequent repairs.

[0131] Step S708: cancel the mark of the current fault node.

[0132] Among them, if the home appliance returns to normal after maintenance and the current faulty node also returns to normal, it means that the failure of the network node is caused by the failure of the home appliance, so the mark of the current faulty node is revoked.

[0133] In the above implementation, the system can effectively distinguish between network failures and home appliance failures, and quickly and accurately locate the source of the problem through status monitoring, feature matching, and cross-validation, thereby formulating targeted recovery strategies and improving the reliability and user experience of the smart home system.

[0134] It is understandable that in a home, the FTTR network may not only connect network devices but also include smart appliances. In this application, by comparing the data of each node, it is determined whether it is a network failure or a home appliance failure. For example, if the network node is inaccessible or faulty due to an abnormality in the home appliance, a home appliance maintenance prompt can be issued; otherwise, it is determined to be a network node failure and the subsequent network repair is awaited.

[0135] In practical applications, this application not only effectively reduces the risk of service interruption due to network failures, but also helps network managers optimize network architecture and resource allocation through historical data analysis. Furthermore, as the number of smart devices increases, automated fault detection and recovery systems can effectively reduce the burden on operations and maintenance personnel, allowing them to focus on higher-level network optimization and policy formulation, thereby promoting the further development of the smart home and IoT industries.

[0136] The embodiment of the present application also discloses an automatic fault detection and recovery system for an FTTR network.

[0137] A FTTR network automatic fault detection and recovery system is used to implement the above-mentioned FTTR network automatic fault detection and recovery method, and the system includes: A network building module for building a FTTR network topology based on the network nodes in each room; Network data acquisition module, used to collect network connection quality data of each network node in real time; A state parameter calculation module is used to calculate the current state parameters of each network node based on the network connection quality data; The abnormality analysis module is used to perform abnormality analysis on the current state parameters of each network node based on the node historical state database to obtain abnormality analysis results; The abnormal node marking module is used to mark the network nodes with abnormal current state parameters as potential abnormal nodes based on the abnormal analysis results, and determine the abnormality level corresponding to the potential abnormal nodes; A fault node marking module is used to mark potential abnormal nodes whose abnormality level exceeds a preset level threshold as current fault nodes; The fault mode analysis module is used to analyze the fault mode of the current fault node based on the node fault history database and the FTTR network topology to determine the corresponding fault source; A backup network switching module is used to detect a backup network path based on the current network path of the fault source based on the FTTR network topology structure, and switch the current network path of the fault source to the backup network path; The recovery feedback information update module is used to obtain the recovery result feedback information and update it to the node fault history database.

[0138] The above implementation achieves rapid identification and automatic recovery of network faults by building a comprehensive FTTR network automatic fault detection and recovery system. This completes a closed-loop fault handling process, encompassing network topology construction, real-time data collection, status parameter calculation, anomaly analysis, faulty node marking, and backup network switching. This system significantly reduces fault recovery time, improves network reliability and stability, and ultimately enhances overall user satisfaction.

[0139] An FTTR network automatic fault detection and recovery system of an embodiment of the present application can implement any of the above-mentioned FTTR network automatic fault detection and recovery methods, and the specific working process of each module in the FTTR network automatic fault detection and recovery system can refer to the corresponding process in the above-mentioned method embodiment.

[0140] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for example, the division of a module is merely a logical functional division, and in actual implementation, other division methods may be used, such as combining or integrating multiple modules into another system, or ignoring or not implementing certain features.

[0141] It should be noted that, in the above embodiments, the description of each embodiment has different emphases. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0142] The above are all preferred embodiments of the present application and are not intended to limit the scope of protection of this application. Unless otherwise specified, any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features. In other words, unless otherwise specified, each feature is merely an example of a series of equivalent or similar features.

Claims

1. A method for automatic fault detection and recovery of FTTR network, characterized in that: The method comprises: Build FTTR network topology based on network nodes in each room; Collect network connection quality data of each network node in real time; Calculating the current state parameters of each network node according to the network connection quality data; Based on the node historical status database, respectively perform abnormal analysis on the current status parameters of each of the network nodes to obtain abnormal analysis results; According to the abnormality analysis result, the network node with abnormal current state parameters is marked as a potential abnormal node, and the abnormality level corresponding to the potential abnormal node is determined; Marking the potential abnormal node whose abnormal level exceeds the preset level threshold as a current fault node; Based on the node fault history database and the FTTR network topology, a fault mode analysis is performed on the current fault node to determine the corresponding fault source; Based on the FTTR network topology, performing backup network path detection according to the current network path of the fault source, and switching the current network path of the fault source to the backup network path; Obtain recovery result feedback information and update it to the node failure history database.

2. A method for automatic fault detection and recovery of FTTR network according to claim 1, characterized in that: Based on the node fault history database and the FTTR network topology, the steps of performing a fault mode analysis on the current fault node and determining the corresponding fault source include: Building and training a fault analysis model based on the node fault history database; Determine an upstream node of the current faulty node based on the FTTR network topology; Obtaining current state parameters of the current faulty node and the upstream node, and inputting them into the fault analysis model for fault analysis, and identifying fault characteristics of the current faulty node and the upstream node; Determine whether the upstream node has a fault characteristic; If not, determining that the fault source is the current fault node; If so, determining each downstream node corresponding to the upstream node based on the FTTR network topology; Acquire current state parameters of each of the downstream nodes, and input them into the fault analysis model to perform fault analysis and identify fault characteristics of the downstream nodes; Counting the proportion of downstream nodes with fault characteristics among the downstream nodes, and if the proportion is greater than a preset proportion threshold, determining that the fault source is the upstream node; If the ratio is less than a preset ratio threshold, then calculating the similarity of the fault characteristics of the current fault node and the upstream node; If the fault feature similarity is greater than a preset similarity threshold, determining that the fault source is the upstream node; If the fault feature similarity is less than a preset similarity threshold, it is determined that the fault source includes the current fault node and the upstream node.

3. A method for automatic fault detection and recovery of FTTR network according to claim 2, characterized in that: Based on the node fault history database, the steps of building and training a fault analysis model include: Based on the node fault history database, a historical sample fault feature set is obtained and preprocessed; wherein the historical sample feature set includes node historical state parameters and corresponding fault feature labels; Perform feature extraction on the preprocessed historical sample feature set to obtain a feature vector set; Dividing the feature vector set into training samples and verification samples; Inputting the training samples into a pre-built random forest model for initial training, optimizing model parameters, and obtaining an initially trained fault analysis model; The fault analysis model is verified and model parameters are corrected based on the verification samples until the loss function of the model meets a preset condition or the number of model iterations reaches a preset number, thereby obtaining the trained fault analysis model.

4. A method for automatic fault detection and recovery of FTTR network according to claim 2, characterized in that: When the fault source is the upstream node, based on the FTTR network topology, performing backup network path detection according to the current network path of the fault source, and switching the current network path of the fault source to the backup network path comprises: Determine each downstream node of the upstream node based on the FTTR network topology structure, and obtain a current network path of the upstream node; wherein the current network path includes a network connection path between the upstream node and each downstream node; Obtaining the historical average traffic demand of the upstream node; Determine the adjacent upstream nodes of the upstream node based on the FTTR network topology structure, and obtain network status parameters of each of the adjacent upstream nodes; Performing traffic carrying analysis and performance evaluation analysis on each adjacent upstream node according to the network status parameters to obtain traffic carrying results and performance evaluation results of each adjacent upstream node; Filter the adjacent upstream nodes whose traffic carrying results meet the historical average traffic demand to obtain a list of candidate adjacent upstream nodes; sorting the candidate adjacent upstream node list according to the performance evaluation result, and selecting the adjacent upstream node with the best performance evaluation result in the candidate adjacent upstream node list as a backup adjacent upstream node; Using the network connection path between the backup adjacent upstream node and each of the downstream nodes as the backup network path; The data flow of the upstream node is switched to the backup adjacent upstream node.

5. A method for automatic fault detection and recovery of FTTR network according to claim 4, characterized in that: The step of obtaining recovery result feedback information and updating it to the node fault history database includes: receiving recovery result feedback information of the current fault node and the downstream node, and uploading it to the node fault history database.

6. A method for automatic fault detection and recovery of FTTR network according to claim 2, characterized in that: When the fault source is the current fault node, based on the FTTR network topology, performing backup network path detection according to the current network path of the fault source, and switching the current network path of the fault source to the backup network path comprises: Determine the current network path of the current fault node based on the FTTR network topology structure; wherein the current network path includes a network connection path between an upstream node of the current fault node and the current fault node; Obtaining the historical average traffic demand of the current faulty node; Determine each downstream node corresponding to the upstream node based on the FTTR network topology structure, and obtain network status parameters of each downstream node; Performing traffic carrying analysis and performance evaluation analysis on each downstream node according to the network status parameters to obtain traffic carrying results and performance evaluation results of each downstream node; Filter the downstream nodes whose traffic carrying results meet the historical average traffic demand to obtain a candidate downstream node list; Sorting the candidate downstream node list according to the performance evaluation result, and selecting the downstream node with the best performance evaluation result in the candidate downstream node list as a standby downstream node; Using the network connection path between the backup downstream node and the upstream node as the backup network path; The data flow of the current failed node is switched to the standby downstream node.

7. A method for automatic fault detection and recovery of FTTR network according to claim 2, characterized in that: When the fault source includes the current fault node and the upstream node, based on the FTTR network topology, performing backup network path detection according to the current network path of the fault source, and switching the current network path of the fault source to the backup network path comprises: Determine the current network path between the faulty node and the upstream node based on the FTTR network topology structure; wherein the current network path includes the network connection path between the current faulty node and the upstream node; Obtaining the historical average traffic demand of the current network path; Determine each adjacent upstream node of the upstream node based on the FTTR network topology structure, and determine the corresponding adjacent downstream node according to the adjacent upstream node to obtain each adjacent network path; wherein the adjacent network path includes a network connection path between the adjacent upstream node and the corresponding adjacent downstream node; Acquire network status parameters of each of the adjacent network paths, and perform traffic carrying analysis and performance evaluation analysis on each of the adjacent network paths to obtain traffic carrying results and performance evaluation results of each of the adjacent network paths; Filtering adjacent network paths whose traffic carrying results meet the historical average traffic demand to obtain a list of candidate adjacent network paths; Sorting the candidate adjacent network path list according to the performance evaluation result, and taking the adjacent network path with the best performance evaluation result in the candidate adjacent network path list as the backup network path; The data flow of the current network path between the faulty node and the upstream node is switched to the backup network path.

8. A method for automatic fault detection and recovery of FTTR network according to any one of claims 6 to 7, characterized in that: The step of obtaining recovery result feedback information and updating the node fault history database includes: generating recovery result feedback information according to the network switching result, and updating the node fault history database.

9. The method for automatic fault detection and recovery of FTTR network according to claim 1, characterized in that: After the step of marking the potential abnormal node whose abnormal level exceeds the preset level threshold as the current fault node, the step further includes: Acquire the household appliance corresponding to the current faulty node according to the current faulty node; Obtaining working status parameters of the household appliance; Determining whether the household appliance has a fault according to the working status parameter; If there is no fault, keep the mark of the current faulty node; If there is a fault, a maintenance prompt message is sent to the user terminal according to the location information of the home appliance; Receiving the household appliance recovery information fed back by the user terminal, performing fault analysis on the current fault node, and determining whether there is a fault feature; If yes, keep the mark of the current faulty node; If not, the mark of the current faulty node is cancelled.

10. An FTTR network automatic fault detection and recovery system, used to execute an FTTR network automatic fault detection and recovery method according to any one of claims 1 to 9, characterized in that: The system comprises: A network building module, used to build the FTTR network topology based on the network nodes in each room; Network data collection module, used to collect network connection quality data of each network node in real time; A state parameter calculation module, used to calculate the current state parameter of each network node according to the network connection quality data; An abnormality analysis module is used to perform abnormality analysis on the current state parameters of each network node based on the node historical state database to obtain abnormality analysis results; An abnormal node marking module is used to mark the network node with abnormal current state parameters as a potential abnormal node according to the abnormal analysis result, and determine the abnormal level corresponding to the potential abnormal node; A fault node marking module, used to mark the potential abnormal node whose abnormal level exceeds a preset level threshold as a current fault node; A fault mode analysis module, used to perform a fault mode analysis on the current fault node based on a node fault history database and the FTTR network topology to determine a corresponding fault source; A backup network switching module, configured to detect a backup network path based on the FTTR network topology and the current network path of the fault source, and switch the current network path of the fault source to the backup network path; The recovery feedback information updating module is used to obtain the recovery result feedback information and update it to the node failure history database.

Images