A method, system, device, processor, and medium for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM.

Abstract

This invention relates to a method for self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM, comprising the following steps: a range extender detects the connection status of the Wi-Fi Mesh backhaul link in real time; multi-dimensional status data corresponding to the backhaul link is collected and preprocessed; the preprocessed status data is encrypted and sent to the cloud-based LLM inference engine via a communication module; fault diagnosis inference is performed and a corresponding repair strategy is generated; corresponding hierarchical repair actions are executed; and the status of the repaired backhaul link is verified. The method, system, device, processor, and computer-readable storage medium for self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM of this invention improve fault diagnosis accuracy, significantly shorten the average recovery time, intelligently select the optimal repair strategy, reduce user experience jitter, reduce user fault perception time, significantly improve user satisfaction, and provide continuous optimization capabilities. The feedback closed-loop mechanism makes the system increasingly intelligent with use, and the diagnostic accuracy continues to improve with usage.

Description

Technical Field

[0001] This invention relates to the field of Wi-Fi Mesh network technology, and more particularly to the field of intelligent operation and maintenance. Specifically, it refers to a method, system, device, processor, and computer-readable storage medium for achieving self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM. Background Technology

[0002] 1) Current status of Wi-Fi Mesh network technology development: With the increasing prevalence of smart home devices, users are demanding higher coverage and stability from Wi-Fi networks. Wi-Fi Mesh networks extend Wi-Fi signal coverage by having multiple nodes work together, solving the problem of insufficient coverage from traditional single routers. The backhaul link is the core connection in a Mesh network, responsible for data transmission between nodes.

[0003] In traditional Wi-Fi Mesh networks, backhaul link failure self-healing mechanisms typically employ fixed threshold detection and a fixed number of retries. For example, when a link is lost, wpa_supplicant attempts to reconnect at fixed time intervals, performing the same number of retries and the same recovery process regardless of the cause of the failure.

[0004] Although some manufacturers have tried to introduce machine learning methods for fault diagnosis, these methods are usually based on traditional machine learning algorithms (such as decision trees and random forests) to classify known fault types. They require a large amount of labeled data for training and cannot handle unknown types of faults, resulting in poor adaptability.

[0005] (2) Deficiencies in the prior art: Traditional backhaul link self-healing relies on fixed rules, which has the following specific drawbacks: 1. Inability to intelligently distinguish fault types, resulting in many invalid operations: Traditional mechanisms cannot distinguish between different fault types such as signal interference, authentication failure, driver abnormality, and channel conflict. They can only uniformly execute the same repair process. For example, regardless of the fault, a fixed number of wpa_supplicant reconnections must be executed. It is impossible to select a targeted repair strategy based on the cause of the fault, resulting in invalid retries and increased network burden.

[0006] When the fault is caused by temporary interference, waiting at fixed intervals prolongs the recovery time; When the failure is caused by a driver malfunction, simply retrying will not solve the problem; it will only increase the system load. When the failure is caused by a failed authentication negotiation, only re-authentication is required to recover, without resetting the driver. The existing mechanism actually leads to an unnecessary long recovery time.

[0007] 2. Long recovery time: The traditional mechanism of fixed 3 retries with fixed intervals cannot dynamically adjust the repair order and recovery process according to the actual fault situation, resulting in an average recovery time of minutes.

[0008] 3. Poor user experience: When a failure occurs, users have to wait a long time to restore the network, and even after restoration, the root cause cannot be identified, and the problem may recur.

[0009] 4. Unpredictable maintenance: It can only respond passively after a failure occurs, and cannot detect potential problems in advance based on historical data analysis.

[0010] In summary, the core technical problem faced by existing Wi-Fi Mesh backhaul link self-healing is that the causes of failures are complex and diverse, including signal interference, channel conflicts, authentication failures, driver anomalies, and other types. Different types of failures require different repair strategies, but traditional fixed rules cannot make intelligent inference decisions based on real-time collected multi-dimensional operational status data. This results in: inaccurate diagnosis of failure types and inability to adaptively select the optimal repair strategy, leading to long recovery times and many invalid operations.

[0011] The technical problem to be solved by this invention is: how to use the natural language reasoning and generalization capabilities of cloud-based LLM to intelligently diagnose and adaptively repair backhaul link faults based on multi-dimensional state data, thereby overcoming the limitations of traditional fixed rules. Summary of the Invention

[0012] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device, processor and computer-readable storage medium for achieving Wi-Fi Mesh backhaul link self-healing based on cloud LLM with high accuracy, high predictability and wide applicability.

[0013] To achieve the above objectives, the present invention provides a method, system, apparatus, processor, and computer-readable storage medium for implementing Wi-Fi Mesh backhaul link self-healing based on cloud-based LLM, as follows: The main feature of this method for achieving self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM is that the method includes the following steps: (1) The range extender detects the Wi-Fi Mesh backhaul link connection status in real time; (2) After the backhaul link is detected to be disconnected, the status acquisition process is triggered to acquire the multi-dimensional status data corresponding to the backhaul link and to preprocess the multi-dimensional status data. (3) The preprocessed state data is encrypted and sent to the cloud LLM inference engine through the communication module; (4) The cloud-based LLM inference engine analyzes the state data, performs fault diagnosis inference, and generates corresponding repair strategies; (5) Based on the diagnostic results and repair strategies output by the cloud-based LLM inference engine, execute the corresponding graded repair actions; (6) Verify the status of the repaired backhaul link. If the verification fails, the failure result is fed back to the cloud LLM inference engine for iterative decision-making or switching of the repair strategy. If the verification is successful, the success result is fed back to the cloud storage for subsequent optimization.

[0014] Preferably, the multidimensional state data in step (2) includes at least RSSI signal strength, SNR signal-to-noise ratio, packet loss rate, channel interference, device status, authentication handshake information, and running time; the channel interference is obtained through full-channel scanning results, including the interference intensity value of each channel.

[0015] Preferably, the preprocessing of the multidimensional state data in step (2) specifically includes the following steps: Data aggregation, anomaly filtering, data standardization, and feature extraction are performed to organize the raw collected data into a structured format.

[0016] Preferably, step (4) involves fault diagnosis reasoning and generating a corresponding repair strategy, specifically including the following steps: (4.1) Receive multidimensional state data; (4.2) Construct structured prompts, which include system prompt templates, fault contexts, constraint injection, and complete prompts; (4.3) Perform inference analysis on the structured prompts using an LLM cluster and output a response result in JSON format; (4.4) Parse the response results, perform result verification, policy formatting, instruction issuance and log recording to obtain diagnostic results and repair strategies.

[0017] Preferably, step (5) specifically includes the following steps: (5.1) Establish a mapping library between fault types and graded repair strategies, or use reinforcement learning for strategy selection; (5.2) Dynamically select the optimal repair solution based on the diagnosis results, and attempt repairs in order of increasing impact on users.

[0018] Preferably, the cloud-based LLM inference engine is deployed on a cloud server, or on an edge computing gateway or a local high-performance embedded device; the cloud-based LLM inference engine uses a general large language model, a distilled lightweight LLM model, or a hybrid expert model.

[0019] This system, which implements Wi-Fi Mesh backhaul self-healing based on cloud-based LLM, is characterized by the following: The terminal sensing module is used to receive the operating status of the range extender, collect multi-dimensional status data and preprocess it, and output the preprocessed status data. The network transmission layer, connected to the terminal sensing module, is used to receive the preprocessed state data, encrypt and transmit the data via the Internet, and output it to the cloud inference layer. The cloud-based inference layer, connected to the network transmission layer, is used to receive the preprocessed state data, analyze the state data through the Large Language Model (LLM) inference engine, perform fault diagnosis inference and generate a repair strategy, and output the diagnosis results and repair strategy to the terminal perception module. The terminal perception module also receives the diagnostic results and repair strategies, executes corresponding graded repair actions, verifies the repair effect, and outputs feedback results to the cloud inference layer.

[0020] Preferably, the terminal sensing module includes: The status acquisition module is used to acquire multi-dimensional status data in response to the detection of a backhaul link disconnection. The multi-dimensional status data includes at least parameters indicating signal quality, parameters indicating channel environment, and parameters indicating device operating status. The data preprocessing module, connected to the state acquisition module, is used to receive the multidimensional state data, preprocess the multidimensional state data, and output structured data. A communication module, connected to the data preprocessing module, is used to receive structured data output by the data preprocessing module and encrypt and transmit it to the cloud. The execution control module, connected to the communication module, is used to receive repair instructions from the cloud, parse the instructions and execute hierarchical repair actions, and output the repair execution results.

[0021] Preferably, the cloud inference layer includes: The API gateway is used to receive data sent by the communication module and forward it. The LLM inference engine is used to receive data forwarded by the API gateway, analyze the multi-dimensional state data to perform fault diagnosis inference, generate repair strategies corresponding to the fault causes, and output diagnostic results and repair strategies. The knowledge base and policy base, connected to the LLM inference engine, are used to provide a mapping between fault types and graded repair policies.

[0022] The device for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM is characterized in that the device comprises: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the above-described method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM.

[0023] The processor for implementing Wi-Fi Mesh backhaul link self-healing based on cloud-based LLM is characterized in that the processor is configured to execute computer-executable instructions, and when the computer-executable instructions are executed by the processor, the various steps of the above-mentioned method for implementing Wi-Fi Mesh backhaul link self-healing based on cloud-based LLM are implemented.

[0024] The computer-readable storage medium is characterized in that it stores a computer program thereon, which can be executed by a processor to implement the various steps of the above-described method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM.

[0025] The present invention, employing a cloud-based LLM-based method, system, device, processor, and computer-readable storage medium for Wi-Fi Mesh backhaul link self-healing, significantly improves fault diagnosis accuracy. Compared to traditional fixed-rule methods, the present invention increases the diagnostic accuracy in mixed fault scenarios from approximately 60% to over 90%, effectively distinguishing between different types of faults such as signal interference, authentication failure, and driver anomalies, avoiding blind retries. The average recovery time is significantly reduced by intelligently selecting the optimal repair strategy, attempting repairs in order of low impact to high impact, reducing the average network recovery time from 2-3 minutes to 15-30 seconds. Invalid retries and invalid operations are reduced, accurately locating the cause of the fault. Invalid driver reset operations are reduced by approximately 70%, avoiding unnecessary system disturbances, extending equipment lifespan, and reducing user experience jitter. User experience is improved through rapid network service recovery, significantly reducing user fault perception time and substantially increasing user satisfaction. Predictive maintenance capabilities are provided through cloud-based historical data analysis, enabling early detection of potential faults and predictive maintenance. Continuous optimization capabilities are achieved through a feedback loop mechanism that makes the system increasingly intelligent with use, continuously improving diagnostic accuracy. Attached Figure Description

[0026] Figure 1 This is a flowchart of the method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to the present invention.

[0027] Figure 2 This is a schematic diagram of the system architecture for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to the present invention.

[0028] Figure 3 This is a schematic diagram of the terminal sensing module of the system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM according to the present invention.

[0029] Figure 4 This is a schematic diagram of the structure of the LLM inference engine of the system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM according to the present invention. Detailed Implementation

[0030] To more clearly describe the technical content of the present invention, the following description is provided in conjunction with specific embodiments.

[0031] The method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM of the present invention includes the following steps: (1) The range extender detects the Wi-Fi Mesh backhaul link connection status in real time; (2) After the backhaul link is detected to be disconnected, the status acquisition process is triggered to acquire the multi-dimensional status data corresponding to the backhaul link and to preprocess the multi-dimensional status data. (3) The preprocessed state data is encrypted and sent to the cloud LLM inference engine through the communication module; (4) The cloud-based LLM inference engine analyzes the state data, performs fault diagnosis inference, and generates corresponding repair strategies; (5) Based on the diagnostic results and repair strategies output by the cloud-based LLM inference engine, execute the corresponding graded repair actions; (6) Verify the status of the repaired backhaul link. If the verification fails, the failure result is fed back to the cloud LLM inference engine for iterative decision-making or switching of the repair strategy. If the verification is successful, the success result is fed back to the cloud storage for subsequent optimization.

[0032] As a preferred embodiment of the present invention, the multidimensional state data in step (2) includes at least RSSI signal strength, SNR signal-to-noise ratio, packet loss rate, channel interference, device status, authentication handshake information and running time; the channel interference is obtained through full channel scanning results, including the interference intensity value of each channel.

[0033] In a preferred embodiment of the present invention, step (2) involves preprocessing the multidimensional state data, specifically including the following steps: Data aggregation, anomaly filtering, data standardization, and feature extraction are performed to organize the raw collected data into a structured format.

[0034] In a preferred embodiment of the present invention, step (4) involves fault diagnosis reasoning and generating a corresponding repair strategy, specifically including the following steps: (4.1) Receive multidimensional state data; (4.2) Construct structured prompts, which include system prompt templates, fault contexts, constraint injection, and complete prompts; (4.3) Perform inference analysis on the structured prompts using an LLM cluster and output a response result in JSON format; (4.4) Parse the response results, perform result verification, policy formatting, instruction issuance and log recording to obtain diagnostic results and repair strategies.

[0035] In a preferred embodiment of the present invention, step (5) specifically includes the following steps: (5.1) Establish a mapping library between fault types and graded repair strategies, or use reinforcement learning for strategy selection; (5.2) Dynamically select the optimal repair solution based on the diagnosis results, and attempt repairs in order of increasing impact on users.

[0036] In a preferred embodiment of the present invention, the cloud-based LLM inference engine is deployed on a cloud server, or on an edge computing gateway or a local high-performance embedded device; the cloud-based LLM inference engine adopts a general large language model, a distilled lightweight LLM model, or a hybrid expert model.

[0037] The present invention discloses a system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM, wherein the system comprises: The terminal sensing module is used to receive the operating status of the range extender, collect multi-dimensional status data and preprocess it, and output the preprocessed status data. The network transmission layer, connected to the terminal sensing module, is used to receive the preprocessed state data, encrypt and transmit the data via the Internet, and output it to the cloud inference layer. The cloud-based inference layer, connected to the network transmission layer, is used to receive the preprocessed state data, analyze the state data through the Large Language Model (LLM) inference engine, perform fault diagnosis inference and generate a repair strategy, and output the diagnosis results and repair strategy to the terminal perception module. The terminal perception module also receives the diagnostic results and repair strategies, executes corresponding graded repair actions, verifies the repair effect, and outputs feedback results to the cloud inference layer.

[0038] In a preferred embodiment of the present invention, the terminal sensing module includes: The status acquisition module is used to acquire multi-dimensional status data in response to the detection of a backhaul link disconnection. The multi-dimensional status data includes at least parameters indicating signal quality, parameters indicating channel environment, and parameters indicating device operating status. The data preprocessing module, connected to the state acquisition module, is used to receive the multidimensional state data, preprocess the multidimensional state data, and output structured data. A communication module, connected to the data preprocessing module, is used to receive structured data output by the data preprocessing module and encrypt and transmit it to the cloud. The execution control module, connected to the communication module, is used to receive repair instructions from the cloud, parse the instructions and execute hierarchical repair actions, and output the repair execution results.

[0039] In a preferred embodiment of the present invention, the cloud inference layer includes: The API gateway is used to receive data sent by the communication module and forward it. The LLM inference engine is used to receive data forwarded by the API gateway, analyze the multi-dimensional state data to perform fault diagnosis inference, generate repair strategies corresponding to the fault causes, and output diagnostic results and repair strategies. The knowledge base and policy base, connected to the LLM inference engine, are used to provide a mapping between fault types and graded repair policies.

[0040] The device of the present invention for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM, wherein the device comprises: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the above-described method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM.

[0041] The present invention discloses a processor for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM, wherein the processor is configured to execute computer-executable instructions, and when the computer-executable instructions are executed by the processor, the various steps of the above-described method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM are implemented.

[0042] The computer-readable storage medium of the present invention stores a computer program thereon, which can be executed by a processor to implement the various steps of the above-described method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM.

[0043] Figure 1 This is a flowchart of a self-healing method for backhaul links in a Wi-Fi Mesh network based on LLM.

[0044] The method includes the following steps: S1: The range extender monitors the backhaul link connection status in real time; S2: After detecting a backhaul link disconnection, collect multi-dimensional status data (including RSSI signal strength, SNR signal-to-noise ratio, packet loss rate, channel interference, device status, authentication handshake information, running time, etc.). S3: Send the preprocessed state data to the cloud-based LLM inference engine via the communication module; S4: The cloud-based LLM inference engine analyzes status data, performs fault diagnosis inference, and generates repair strategies; S5: Based on the diagnostic results and repair strategies output by LLM, the terminal executes the corresponding graded repair actions; S6: Verify the repair effect. If it fails, the result is fed back to LLM for iterative decision-making or switching of repair strategies. If it succeeds, the success result is fed back to cloud storage for optimization.

[0045] Figure 2 This is a schematic diagram of the backhaul link self-healing system architecture of a Wi-Fi Mesh network based on LLM.

[0046] The system includes a terminal sensing module 100, a network transmission layer, and a cloud inference layer.

[0047] The terminal sensing module 100 is used to collect the operating status data of the range extender.

[0048] The network transport layer encrypts and transmits data to the cloud via the Internet.

[0049] The cloud-based inference layer includes an API gateway 200, an LLM inference engine 300, a knowledge base 400, a policy base 500, and a data storage 600.

[0050] Figure 3 This is a detailed structural diagram of the terminal sensing module 100.

[0051] The terminal sensing module 100 includes: Status acquisition module 110: responsible for collecting multi-dimensional operational data such as RSSI, SNR, packet loss rate, interference detection, channel scanning, device status, and authentication handshake results.

[0052] Data preprocessing module 120: responsible for data aggregation, anomaly filtering, data standardization, and feature extraction, organizing the raw collected data into a structured format.

[0053] Communication module 130: Responsible for communicating with the cloud API gateway, including API client, TLS encryption, message queue, and retry control.

[0054] Execution control module 140: responsible for parsing the repair instructions sent from the cloud and controlling wpa_supplicant and Wi-Fi driver to perform graded repair actions.

[0055] Figure 4 This is a detailed structural diagram of the LLM inference engine 300.

[0056] LLM Inference Engine 300 includes: Data input interface 310: Receives multidimensional status data from the terminal.

[0057] Prompt constructor 320: Constructs a complete Prompt that includes a system Prompt template, fault context assembly, constraint injection, and a Few-shot example.

[0058] LLM Large Model Core 330: Cloud-based LLM cluster for inference analysis, including model scheduler, inference cluster, caching layer, and monitoring logs.

[0059] Response parser 340: Parses the LLM output JSON, performs result validation, policy formatting, command issuance, and logging.

[0060] The key points of this invention are as follows: 1. Cloud-based LLM inference engine applied to Wi-Fi Mesh backhaul self-healing architecture: By deploying a large language model in the cloud, the system receives multi-dimensional state data collected from terminals and performs intelligent reasoning and decision-making, breaking through the self-healing limitations of traditional fixed rules. Existing technologies, even those using machine learning, require manual annotation and training. This invention utilizes the zero-shot or few-shot reasoning capabilities of LLM to diagnose even unseen fault types.

[0061] 2. Multidimensional state data Prompt construction method: By assembling multi-dimensional operational data such as RSSI, SNR, packet loss rate, and interference conditions into a structured Prompt, a general-purpose large model can accurately understand Wi-Fi Mesh fault scenarios and contexts, thus addressing the technical challenge of adapting a general-purpose LLM to a specific network domain. This is the core technological bridge for applying LLM to Wi-Fi Mesh self-healing in this invention.

[0062] 3. Dynamic generation mechanism for hierarchical adaptive repair strategy: Establish a mapping library between fault types and graded repair strategies, enabling LLM to dynamically select the optimal repair solution based on diagnostic results, and attempt repairs in order of increasing impact on users, avoiding unnecessary deep repair operations and significantly shortening the average recovery time.

[0063] 4. Feedback closed-loop iterative optimization mechanism: The final repair results (success / failure information) are fed back to the cloud-based LLM, forming a complete closed loop of "perception-reasoning-execution-verification-feedback", continuously optimizing the accuracy of subsequent fault diagnosis and the ability to select repair strategies.

[0064] In specific embodiments of the present invention, the following examples are provided: Example 1: Self-healing process of link disconnection caused by signal interference In a user's home, a range extender was deployed in the living room, and the backhaul link connecting to the main router suddenly dropped.

[0065] Step 1: The terminal status acquisition module 110 detects a link disconnection and immediately acquires multi-dimensional status data: RSSI signal strength = -72dBm; SNR (Signal-to-Noise Ratio) = 22dB; Packet loss rate = 95%; Current operating channel = 6 (2.4GHz band); Full channel scan results: Channel 1 interference intensity -45dBm, Channel 6 interference intensity -30dBm, Channel 11 interference intensity -75dBm; The last four handshake authentication status = successful; The equipment can run continuously for 48 hours. Last repair record = None.

[0066] Step 2: After the data preprocessing module performs data standardization, it is encrypted and sent to the cloud API gateway through the communication module.

[0067] Step 3: The cloud-based Prompt constructor constructs a complete Prompt as follows: You are a Wi-Fi Mesh backhaul link troubleshooting expert. Based on the status information collected from the following terminals, you need to analyze the cause of the fault and select the optimal repair strategy from the given options.

[0068] Current status data: RSSI signal strength: -72dBm; SNR (Signal-to-Noise Ratio): 22dB; Current packet loss rate: 95%; Current channel: 6 (2.4GHz); Interference intensity for each channel: Channel 1: -45dBm, Channel 6: -30dBm, Channel 11: -75dBm; Recent authentication status: Authentication successful; Continuous equipment operating time: 48 hours; Recent repair history: None.

[0069] Possible causes of the malfunction are as follows: 1. Severe interference from adjacent channel signals; 2. WPA / WPA3 certification negotiation failed; 3. Wi-Fi driver malfunction; 4. Hardware failure; 5. Other unknown reasons.

[0070] Optional repair strategies: A. Immediately rescan all channels and switch to the available channel with the least interference; B. Disconnect the current connection and renegotiate the authentication. C. Reset the Wi-Fi driver module; D. Trigger a device soft reboot.

[0071] Please strictly adhere to the following JSON format when outputting results. Output only JSON and no other content: {"Cause of Fault":"Fill in your diagnostic conclusion",} "Repair Strategy": "Briefly explain the repair approach" "Explanation": Briefly explain the basis of the reasoning. "Next step": "Enter the action number you choose (A / B / C / D)" } Step 4: The cloud-based LLM performs inference, and the output response is parsed to obtain the result: { "Cause of Fault": "Severe interference from adjacent and co-channel signals" "Remediation Strategy": "Switch to a channel with less interference" "Note": "Currently, channel 6 experiences extremely strong interference, while channel 11 has significantly lower interference levels and its RSSI remains within usable limits." "Next step": "A" } Step 5: The cloud sends instructions to the terminal through the API gateway. The terminal executes the control module to parse the instructions, performs a channel scan, switches to channel 11, and re-establishes the backhaul link.

[0072] Step 6: The terminal verifies that the backhaul link has been restored to normal and data transmission is normal. The successful repair result is then sent to the cloud storage. The entire repair process takes a total of 12 seconds.

[0073] Example 2: Self-healing process of link disconnection due to authentication negotiation failure A range extender detected a backhaul link disconnection and collected the following data: RSSI = -55dBm; SNR = 30dB; Packet loss rate = 100%; Current channel interference is normal; The last four WPA handshakes failed consecutively.

[0074] Based on the structured prompt input, the cloud-based LLM analyzes and diagnoses an anomaly in the WPA authentication negotiation process, outputs repair strategy B, and after the terminal disconnects again, it re-attempts authentication negotiation. The link returns to normal after 1 second.

[0075] Example 3: Self-healing process of link disconnection caused by driver anomaly A range extender detects a backhaul link disconnection and collects data: RSSI = -58dBm; SNR = 28dB; Channel interference is normal. I have failed to re-authenticate twice in a row; Channel switching has been attempted but has failed.

[0076] The cloud-based LLM inference diagnosed a Wi-Fi driver error and output repair strategy C. The terminal executed a Wi-Fi driver reset operation. After the reset was completed, the connection was reconnected and the link was successfully restored. The total time was 22 seconds.

[0077] VI. Other Alternative Solutions (1) Alternative deployment locations for LLM: In addition to cloud deployment, LLM can also be deployed on edge computing gateways or local high-performance embedded devices, which is suitable for scenarios with high privacy sensitivity or network latency requirements.

[0078] (2) Model type substitution: Distilled lightweight LLM models can be used, or a hybrid expert model (MoE) architecture can be adopted to reduce the consumption of computing resources while ensuring reasoning ability.

[0079] (3) Fault classification replacement: The fault classification system can be adjusted according to the actual deployment environment, and diagnostic and repair strategies for specific fault types can be added or refined.

[0080] (4) Strategy selection alternative: Reinforcement learning can be used to replace the rule-based policy library, enabling LLM to learn the optimal repair strategy autonomously.

[0081] For the specific implementation scheme of this embodiment, please refer to the relevant descriptions in the above embodiments, which will not be repeated here.

[0082] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0083] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0084] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0085] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution device. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0086] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The corresponding program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0087] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0088] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0089] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0090] The present invention, employing a cloud-based LLM-based method, system, device, processor, and computer-readable storage medium for Wi-Fi Mesh backhaul link self-healing, significantly improves fault diagnosis accuracy. Compared to traditional fixed-rule methods, the present invention increases the diagnostic accuracy in mixed fault scenarios from approximately 60% to over 90%, effectively distinguishing between different types of faults such as signal interference, authentication failure, and driver anomalies, avoiding blind retries. The average recovery time is significantly reduced by intelligently selecting the optimal repair strategy, attempting repairs in order of low impact to high impact, reducing the average network recovery time from 2-3 minutes to 15-30 seconds. Invalid retries and invalid operations are reduced, accurately locating the cause of the fault. Invalid driver reset operations are reduced by approximately 70%, avoiding unnecessary system disturbances, extending equipment lifespan, and reducing user experience jitter. User experience is improved through rapid network service recovery, significantly reducing user fault perception time and substantially increasing user satisfaction. Predictive maintenance capabilities are provided through cloud-based historical data analysis, enabling early detection of potential faults and predictive maintenance. Continuous optimization capabilities are achieved through a feedback loop mechanism that makes the system increasingly intelligent with use, continuously improving diagnostic accuracy.

[0091] In this specification, the invention has been described with reference to specific embodiments thereof. However, it will be apparent that various modifications and variations can be made without departing from the spirit and scope of the invention. Therefore, the specification and drawings should be considered illustrative rather than restrictive.

Claims

1. A method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM, characterized in that, The method includes the following steps: (1) The range extender detects the Wi-Fi Mesh backhaul link connection status in real time; (2) After the backhaul link is detected to be disconnected, the status acquisition process is triggered to acquire the multi-dimensional status data corresponding to the backhaul link and to preprocess the multi-dimensional status data. (3) The preprocessed state data is encrypted and sent to the cloud LLM inference engine through the communication module; (4) The cloud-based LLM inference engine analyzes the state data, performs fault diagnosis inference, and generates corresponding repair strategies; (5) Based on the diagnostic results and repair strategies output by the cloud-based LLM inference engine, execute the corresponding graded repair actions; (6) Verify the status of the repaired backhaul link. If the verification fails, the failure result is fed back to the cloud LLM inference engine for iterative decision-making or switching of the repair strategy. If the verification is successful, the success result is fed back to the cloud storage for subsequent optimization.

2. The method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to claim 1, characterized in that, The multidimensional status data in step (2) includes at least RSSI signal strength, SNR signal-to-noise ratio, packet loss rate, channel interference, device status, authentication handshake information, and running time; the channel interference is obtained through full channel scanning results, including the interference strength value of each channel.

3. The method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to claim 1, characterized in that, The preprocessing of the multidimensional state data in step (2) includes the following steps: Data aggregation, anomaly filtering, data standardization, and feature extraction are performed to organize the raw collected data into a structured format.

4. The method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to claim 1, characterized in that, Step (4) involves fault diagnosis reasoning and generating a corresponding repair strategy, specifically including the following steps: (4.1) Receive multidimensional state data; (4.2) Construct structured prompts, which include system prompt templates, fault contexts, constraint injection, and complete prompts; (4.3) Perform inference analysis on the structured prompts using an LLM cluster and output a response result in JSON format; (4.4) Parse the response results, perform result verification, policy formatting, instruction issuance and log recording to obtain diagnostic results and repair strategies.

5. The method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to claim 1, characterized in that, Step (5) specifically includes the following steps: (5.1) Establish a mapping library between fault types and graded repair strategies, or use reinforcement learning for strategy selection; (5.2) Dynamically select the optimal repair solution based on the diagnosis results, and attempt repairs in order of increasing impact on users.

6. The method for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM according to claim 1, characterized in that, The cloud-based LLM inference engine is deployed on a cloud server, or on an edge computing gateway or a local high-performance embedded device; the cloud-based LLM inference engine uses a general large language model, a distilled lightweight LLM model, or a hybrid expert model.

7. A system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM, characterized in that, The system includes: The terminal sensing module is used to receive the operating status of the range extender, collect multi-dimensional status data and preprocess it, and output the preprocessed status data. The network transmission layer, connected to the terminal sensing module, is used to receive the preprocessed state data, encrypt and transmit the data via the Internet, and output it to the cloud inference layer. The cloud-based inference layer, connected to the network transmission layer, is used to receive the preprocessed state data, analyze the state data through the Large Language Model (LLM) inference engine, perform fault diagnosis inference and generate a repair strategy, and output the diagnosis results and repair strategy to the terminal perception module. The terminal perception module also receives the diagnostic results and repair strategies, executes corresponding graded repair actions, verifies the repair effect, and outputs feedback results to the cloud inference layer.

8. The system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM as described in claim 7, characterized in that, The terminal sensing module includes: The status acquisition module is used to acquire multi-dimensional status data in response to the detection of a backhaul link disconnection. The multi-dimensional status data includes at least parameters indicating signal quality, parameters indicating channel environment, and parameters indicating device operating status. The data preprocessing module, connected to the state acquisition module, is used to receive the multidimensional state data, preprocess the multidimensional state data, and output structured data. A communication module, connected to the data preprocessing module, is used to receive structured data output by the data preprocessing module and encrypt and transmit it to the cloud. The execution control module, connected to the communication module, is used to receive repair instructions from the cloud, parse the instructions and execute hierarchical repair actions, and output the repair execution results.

9. The system for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM as described in claim 7, characterized in that, The cloud-based inference layer includes: The API gateway is used to receive data sent by the communication module and forward it. The LLM inference engine is used to receive data forwarded by the API gateway, analyze the multi-dimensional state data to perform fault diagnosis inference, generate repair strategies corresponding to the fault causes, and output diagnostic results and repair strategies. The knowledge base and policy base, connected to the LLM inference engine, are used to provide a mapping between fault types and graded repair policies.

10. A device for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM, characterized in that, The device includes: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the steps of the method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM as described in any one of claims 1 to 6.

11. A processor for implementing self-healing of Wi-Fi Mesh backhaul links based on cloud-based LLM, characterized in that, The processor is configured to execute computer-executable instructions, which, when executed by the processor, implement the steps of the method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM as described in any one of claims 1 to 6.

12. A computer-readable storage medium, characterized in that, It stores a computer program that can be executed by a processor to implement the steps of the method for self-healing Wi-Fi Mesh backhaul links based on cloud-based LLM as described in any one of claims 1 to 6.

Images