Autonomous fault-tolerant system and method for dual-mode hot switching

Abstract

The invention discloses an autonomous fault-tolerant system and method for dual-mode hot switching, and relates to the technical field of edge computing, and the method comprises the steps: switching a main communication link to a standby communication link when the main communication link has a fault, and generating a first hardware equipment migration instruction when any one of a plurality of hardware equipment has a fault risk, meanwhile, a second hardware equipment migration instruction is generated when any hardware equipment is in a fault state, the fault hardware equipment in the fault risk and the fault state is migrated to a target node based on the first hardware equipment migration instruction and the second hardware equipment migration instruction, and when the target node recovers execution of a service task, the target node is migrated to the target node. A service task is executed based on a communication link of the autonomous fault-tolerant system, the technical problems of long response time, delay and high cost generated by manual operation and maintenance and cloud redundant backup are solved, a dual-mode communication design idea is adopted, and all functional modules work cooperatively, so that a complete closed loop from fault monitoring to autonomous recovery is realized.

Description

Technical Field

[0001] The present invention relates to the field of edge computing technology, and in particular to an autonomous fault-tolerant system and method for dual-mode hot switching. Background Art

[0002] In modern industrial and smart living scenarios, a large number of physical devices are deployed at the edge. However, these devices face challenges such as unstable networks and frequent hardware failures during operation, which can lead to incomplete data transmission or delayed command delivery.

[0003] Related technologies mainly rely on manual intervention or cloud-based redundant backup. Manual intervention requires operation and maintenance personnel to arrive at the site to troubleshoot and repair faults, which has a long response time and high time and labor costs. Cloud-based redundant backup will cause large delays in data transmission to and recovery from the cloud, which cannot meet some application scenarios with high real-time requirements and urgently needs to be solved. Summary of the Invention

[0004] The present invention provides an autonomous fault-tolerant system and method for dual-mode hot switching, which at least solves the problems in related technologies such as long response time, delay and high cost caused by manual operation and maintenance and cloud redundant backup, resulting in incomplete data transmission or failure to issue instructions in a timely manner.

[0005] The present invention provides an autonomous fault-tolerant system for dual-mode hot switching, comprising: a dual-mode communication hot-swap module, which controls the communication link of the autonomous fault-tolerant system to switch to the backup communication link when detecting that the primary communication link in the autonomous fault-tolerant system is in a fault state; a fault prediction module, configured to predict faults based on the operating status of each hardware device and generate a first hardware device migration instruction when any hardware device among the plurality of hardware devices has a risk of failure; A lightweight containerized fault-tolerant engine is configured to generate a second hardware device migration instruction when any hardware device among the multiple hardware devices is in a faulty state, and based on the first hardware device migration instruction and the second hardware device migration instruction, migrate the faulty hardware device at risk of failure and the faulty hardware device in the faulty state to a target node, and while the target node resumes execution of the business task, execute the business task based on the communication link of the autonomous fault-tolerant system.

[0006] The present invention also provides an autonomous fault-tolerant method for dual-mode hot switching, comprising the following steps: monitoring a primary communication link in the autonomous fault-tolerant system, and controlling the communication link of the autonomous fault-tolerant system to switch to a backup communication link when the primary communication link is in a fault state; Monitor the operating status of each hardware device, perform fault prediction based on the operating status of each hardware device, trigger a first hardware device migration instruction when the hardware device is at risk of failure, and generate a second hardware device migration instruction when any hardware device among the multiple hardware devices is detected to be in a faulty state; According to the first hardware device migration instruction and the second hardware device migration instruction, the faulty hardware device in the faulty state is migrated to the target node, and the execution of the business task is resumed at the target node, while the business task is continued to be executed based on the backup communication link.

[0007] Through the present invention, when the main communication link fails, the dual-mode communication hot-swap module is used to switch the main communication link to the backup communication link, and when any hardware device among multiple hardware devices is at risk of failure, a first hardware device migration instruction is generated based on the fault prediction module, and at the same time, a second hardware device migration instruction is generated when any hardware device is in a faulty state. Based on the first and second hardware device migration instructions, the faulty hardware devices at risk of failure and in a faulty state are migrated to the target node based on a lightweight containerized fault-tolerant engine, and while the business task execution is restored at the target node, the business task is executed based on the communication link of the autonomous fault-tolerant system, which solves the technical problems of long response time, delay and high cost caused by manual operation and maintenance and cloud redundant backup. By adopting the dual-mode communication design concept, each functional module works together to achieve a complete closed loop from fault monitoring to autonomous recovery. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the embodiments of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0009] Figure 1 A block diagram of an autonomous fault-tolerant system with dual-mode hot switching provided by an embodiment of the present invention; Figure 2 A schematic diagram of a dual-mode communication hot switching mechanism provided by one embodiment of the present invention; Figure 3 A schematic diagram of a lightweight containerized fault-tolerant engine provided by one embodiment of the present invention; Figure 4 A schematic diagram of an AI-driven fault prediction model provided by one embodiment of the present invention; Figure 5 4 is a flow chart of an autonomous fault-tolerant method for dual-mode hot switching according to an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0011] It should be noted that, in the description of the present invention, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. The terms "first," "second," etc., in the present invention are used to distinguish similar objects, and are not used to describe a particular order or precedence.

[0012] In order to enable those skilled in the art to better understand the solutions of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0013] An embodiment of the present invention provides an autonomous fault-tolerant system for dual-mode hot switching. The system is described in detail in conjunction with an execution system of the autonomous fault-tolerant system for dual-mode hot switching.

[0014] Specifically, before introducing the embodiments of the present invention, we first introduce the autonomous fault-tolerant system in the related technology. In modern industrial and smart life scenarios, a large number of physical devices such as relay protection devices and smart meters are deployed on the edge side. However, these devices face many challenges during operation. Network fluctuations are one of the common problems. For example, in some remote areas or complex industrial environments, network signals are unstable, and packet loss and delays often occur, which will lead to incomplete data transmission or instructions cannot be issued in time; hardware failures also occur from time to time, and long-term operation of equipment may cause problems such as aging and damage of components, which are mainly manifested in the following aspects: (1) Inadequate response to network fluctuations: When network interruption or quality degradation occurs, there is a lack of an efficient local communication switching mechanism, which can easily lead to data loss or interruption of command transmission, affecting the normal operation of the equipment; (2) Delayed response to hardware failures: Relying on manual intervention to troubleshoot and repair hardware failures, the response time is long and cannot meet real-time requirements, especially in harsh or remote environments, where it is difficult and costly for manual personnel to arrive at the site; (3) Limitations of cloud backup: Although cloud redundant backup can ensure data security, there is a large delay in the data upload and recovery process, which is difficult to adapt to scenarios with high real-time requirements, and the cloud operation cost is high; (4) High operation and maintenance costs: Manual intervention requires a lot of manpower, and cloud backup also comes with economic costs. The existing solutions have not implemented an efficient and autonomous fault-tolerant mechanism, resulting in high overall operation and maintenance costs.

[0015] The solutions in related technologies mainly rely on manual intervention or cloud-based redundant backup. Manual intervention requires operation and maintenance personnel to arrive at the site to troubleshoot and repair the problem, which takes a long time to respond. Moreover, in some harsh environments or remote areas, the time and labor costs of arriving at the site are very high. Although cloud-based redundant backup can ensure data security to a certain extent, it will cause large delays in the process of transmitting data to the cloud and restoring data from the cloud, which cannot meet some application scenarios with high real-time requirements.

[0016] The current technological evolution of the edge side of physical devices shows five core trends, namely: (1) Deep penetration of edge AI (Artificial Intelligence): The integration of AI reasoning capabilities into edge devices has become standard. More than 70% of newly released MCUs (Microcontroller Units) / MPUs (Microprocessor Units) are equipped with NPUs (Neural Processing Units) / TPUs (Tensor Processing Units) (such as Renesas RZ / V2N), supporting INT8 (8-bit Integer) / FP16 (16-bit Floating Point) quantization models, reducing reasoning latency to microseconds, and power consumption is only 1 / 20 of that of CPUs (Central Processing Units). In industrial scenarios, edge AI accelerator cards (such as Huawei Atlas 500) achieve millisecond-level real-time decision-making, promoting predictive maintenance and production line optimization; (2) Low power and zero energy innovation: Semiconductor manufacturers use the 18nm FD-SOI process to integrate IoT MCUs. Standby power consumption is reduced to 10μA / MHz (such as STM32U3), and passive devices (such as backscatter Bluetooth tags) and solar power supply solutions are popularized, reducing dependence on traditional power supplies; (3) Wireless protocol integration: The Matter protocol promotes multi-mode chips (such as Silicon Labs EFR32MG24) to support Wi-Fi 6, Bluetooth 5.4, Thread and other protocols; (4) Security protection upgrade: 80% of mid-to-high-end MCUs integrate post-quantum encryption algorithms (such as the Latent Gemini cipher), and the Broadcom BK7236 chip resists quantum attacks through hardware-level security design. Zero trust architecture and dynamic authentication have become standard in industrial scenarios; (5) Deepening cloud-edge collaboration: The intelligent edge computing platform realizes cloud training and edge inference of AI algorithms, supports containerized deployment and OTA (Over-The-Air Technology) upgrades, and reduces data cloud bandwidth consumption by 90%.

[0017] However, the evolution of physical device edge technologies has the following drawbacks: (1) High risk of network fluctuations and communication interruptions: Traditional solutions rely on a single communication link (such as fiber optics and 4G) and lack a local redundant switching mechanism. When the network is interrupted, the equipment cannot maintain command issuance and data transmission, resulting in paralysis of the production system. For example, when a fiber optic failure occurs in a substation, the protection logic fails due to communication interruption and requires manual intervention to repair, which can take up to several hours to recover. In addition, heterogeneous protocols (such as Modbus and 104 protocols) have poor compatibility, and interconnection between devices relies on cloud-based transit, further exacerbating latency and bandwidth pressure.

[0018] (2) Hardware failure response is delayed and manual: After equipment failure, the operation and maintenance process requires manual troubleshooting and hardware replacement, and the average repair time exceeds 24 hours, which cannot meet the real-time requirements of the industry. For example, after a sensor failure on a factory production line, operation and maintenance personnel are required to conduct on-site inspections, which leads to high manpower and maintenance costs. The existing solution lacks the ability to automatically migrate business, and hardware failures directly lead to service interruptions. A rapid recovery mechanism based on containerization or virtualization has not been implemented.

[0019] (3) Insufficient cloud backup and fault prediction capabilities: Although cloud-based redundant backup can ensure data security, data upload and recovery require network transmission, with a latency of hundreds of milliseconds, making it unsuitable for millisecond-level response scenarios (such as autonomous driving decisions). Existing technologies rely on statistical analysis of historical data and can only trigger responses after a fault occurs. They lack AI prediction models based on real-time operating data (such as temperature rise and vibration spectrum) and cannot provide 24-hour advance warning of potential faults. Furthermore, the centralized cloud processing model requires uploading more than 90% of the original data, resulting in high bandwidth costs and significant privacy risks.

[0020] Therefore, based on the above-mentioned problems, the embodiments of the present invention solve many long-standing technical bottlenecks in the edge computing field based on the dynamic resource allocation algorithm based on business priority, the container rapid migration method and the architecture design that supports dynamic loading of protocol modules, and provide strong support for achieving a more stable and efficient edge computing environment.

[0021] Specifically, Figure 1 A block diagram of an autonomous fault-tolerant system with dual-mode hot switching provided by an embodiment of the present invention.

[0022] like Figure 1 As shown, the autonomous fault-tolerant system 10 for dual-mode hot switching includes: a dual-mode communication hot switching module 100 , a lightweight containerized fault-tolerant engine 200 and a fault prediction module 300 .

[0023] Among them, the dual-mode communication hot switching module 100 controls the communication link of the autonomous fault-tolerant system to switch to the backup communication link when it detects that the main communication link in the autonomous fault-tolerant system is in a fault state; the fault prediction module 300 is used to predict faults based on the operating status of each hardware device, and generate a first hardware device migration instruction when any hardware device among multiple hardware devices is at risk of failure; the lightweight containerized fault-tolerant engine 200 is used to generate a second hardware device migration instruction when any hardware device among multiple hardware devices is in a fault state, and based on the first hardware device migration instruction and the second hardware device migration instruction, migrate the faulty hardware device at risk of failure and the faulty hardware device in a fault state to the target node, and while the target node resumes the execution of business tasks, execute business tasks based on the communication link of the autonomous fault-tolerant system.

[0024] Specifically, in order to solve the problem of data loss and instruction interruption caused by network fluctuations / interruptions, the embodiment of the present invention realizes local redundant switching through dual-mode communication. It is intended to monitor that when the main communication link in the autonomous fault-tolerant system is in a fault state, the communication link of the autonomous fault-tolerant system can be controlled to switch to the backup communication link to ensure the continuous issuance of instructions and maintain the continuity of data communication. Therefore, a multi-protocol communication middleware is used to realize heterogeneous protocol fusion, support dynamic resource allocation, and ensure low-latency transmission of key instructions. At the same time, a fault prediction module 300 is integrated, such as a low-power wide area network communication model (such as TinyML (Tiny Machine Learning, micro machine learning) model), can predict and monitor whether hardware devices have failure risks 24 hours in advance, reduce 90% of cloud data transmission requirements, and reduce bandwidth costs and privacy risks. If any hardware device among multiple hardware devices is at risk of failure, a first hardware device migration instruction is generated, and the faulty hardware device at risk of failure is migrated to the target node according to the first hardware device migration instruction. At the same time, the lightweight containerized fault tolerance engine 200 is used to calculate that when any hardware device among multiple hardware devices is in a fault state, a second hardware device migration instruction is generated, and the faulty hardware device in the fault state is migrated to the target node, which can greatly reduce the service interruption time caused by hardware failure, and while the target node resumes the execution of business tasks, the business tasks are executed based on the communication link of the autonomous fault tolerance system.

[0025] According to one embodiment of the present invention, the dual-mode communication hot-switching module 100 includes: a multi-protocol communication middleware, which is used to allocate communication resources according to the priority of the multi-protocol business tasks of the autonomous fault-tolerant system, and to schedule tasks based on the priority of each business task; a link quality real-time monitoring module, which is used to monitor the health index of the main communication link in real time, and generate switching instructions for the pre-switching mechanism based on the health index; and an automatic switching execution module, which is used to monitor the network operation status of the main communication link, and execute the link switching action of the autonomous fault-tolerant system based on the network operation status.

[0026] Specifically, if Figure 2 As shown, in an embodiment of the present invention, the dual-mode communication hot-swap module 100 first uses a multi-protocol communication middleware to implement heterogeneous protocol fusion. For example, the multi-protocol communication middleware supports a heterogeneous protocol fusion engine of MQTT (Message Queuing Telemetry Transport) / Modbus (Modbus Protocol) / 104 protocols, and dynamically allocates communication resources through a priority queue (such as a PQoS (Priority Quality of Service) algorithm). For example, power protection instructions with high real-time requirements (such as the 104 protocol) are given the highest priority (weight 0.8), and data acquisition tasks (such as Modbus / MQTT) are scheduled using polling to ensure that the transmission delay of key instructions is ≤10ms. The protocol parser supports dynamic loading, and a plug-in architecture is used to achieve 72-hour rapid adaptation of new protocols.

[0027] Secondly, a multi-dimensional monitoring indicator system is designed through the real-time link quality monitoring module. It collects real-time data on signal strength (unit: dBm), packet loss rate (threshold: 3%), and round-trip delay (RTT (Round-Trip Time), threshold: 50ms). A sliding window algorithm (window size: 10s) is used to calculate the health index (0-100). When the health index is less than 60, the pre-switching mechanism is triggered to establish long-distance link connections in advance, shortening the switching time to less than 200ms.

[0028] Finally, the automatic switching execution module integrates a long-distance self-organizing network module (such as the LoRa (Long Range) self-organizing network module, which supports Class A and Class B device models) to monitor the network operation status of the main communication link and execute link switching actions of the autonomous fault-tolerant system based on the network operation status. For example, it uses a dynamic spreading factor (SF7-SF12) adaptive algorithm to complete protocol switching within 500ms when the fiber / 4G is interrupted. For example, in the substation scenario, the gateway broadcasts commands to devices within 3 kilometers via LoRa, using frequency hopping technology (FHSS (Frequency Hopping Spread Spectrum, 16 channels) to avoid electromagnetic interference, and the command reliable transmission rate reaches 99.5%.

[0029] Therefore, the dual-mode communication hot-swap module 100 is used to ensure that when the main communication link of the network (such as optical fiber or 4G) fails, it can quickly and smoothly switch to the backup communication link (such as LoRa), thereby ensuring the continuity of key business instructions and services and avoiding service interruptions caused by network interruptions. At the same time, combined with multi-protocol communication middleware, the dual-mode communication hot-swap module 100 can support multiple industrial communication protocols (such as MQTT, Modbus, IEC 104, etc.), enhancing the compatibility and adaptability of the system, and based on the priority queue manager (such as the PQoS algorithm), it can also dynamically adjust the allocation of communication resources according to the needs of different tasks to ensure that high-priority tasks are processed in a timely manner.

[0030] According to one embodiment of the present invention, the multi-protocol communication middleware includes: a priority queue management unit, which is used to prioritize multi-protocol business tasks based on a preset priority service quality algorithm to obtain a priority queue of multi-protocol business tasks; a protocol parser plug-in unit, which is used to parse the business tasks corresponding to each protocol based on a plug-in architecture, so that the business tasks corresponding to each protocol after parsing are adapted to the new protocol; a communication resource allocation unit, which is used to allocate communication resources based on the priority of the multi-protocol business tasks; and a task scheduling unit, which is used to select target business tasks according to the priority of the multi-protocol business tasks based on the priority queue, and call the corresponding protocol parser plug-in unit for parsing according to the protocol type of the selected target business task.

[0031] The preset priority quality of service algorithm may be defined by those skilled in the art according to actual test requirements and is not specifically defined herein.

[0032] Specifically, the multi-protocol communication middleware is mainly composed of the following core parts: a priority queue management unit, a protocol parser plug-in unit, a communication resource allocation unit and a task scheduling unit, wherein the priority queue management unit is used to prioritize multi-protocol business tasks based on a preset priority quality of service algorithm (such as a PQoS algorithm) to obtain a priority queue of multi-protocol business tasks; the protocol parser plug-in unit is used to parse the business tasks corresponding to each protocol based on a plug-in architecture, so that the business tasks corresponding to each protocol after parsing are adapted to the new protocol, that is, it can support dynamic loading of parsers for different protocols to facilitate rapid adaptation of new protocols; the communication resource allocation unit is used to dynamically allocate communication resources based on the priority of multi-protocol business tasks to ensure low-latency transmission of key instructions; the task scheduling unit is used to select target business tasks according to the priority of multi-protocol business tasks based on the priority queue, and call the corresponding protocol parser plug-in unit for parsing according to the protocol type of the selected target business task.

[0033] Specifically, the priority queue management unit uses the PQoS algorithm to implement the priority queue, giving the power protection instructions of the 104 protocol the highest priority (weight 0.8). The data collection tasks of Modbus and MQTT are scheduled using polling. The priority queue can be implemented using a heap data structure to ensure efficient insertion and deletion operations; the protocol parser plug-in unit adopts a plug-in architecture. Each protocol parser is an independent plug-in. The plug-in interacts with the main program through a unified interface. The main program can dynamically load and unload plug-ins, so that it can quickly adapt to new protocols within 72 hours; the communication resource allocation unit allocates communication resources to different tasks based on the results of the priority queue. For the power protection instructions of the 104 protocol, bandwidth and processing time are allocated first to ensure that the transmission delay is ≤ 10ms; the task scheduling unit takes tasks from the priority queue, calls the corresponding protocol parser for processing according to the protocol type of the task, and sends the processing results.

[0034] The following is a simplified Python code example that implements some of the functionality of the above architecture: # Protocol parser interface class ProtocolParser(ABC): @abstractmethod def parse(self, data): pass # 104 Specification parser example class IEC104Parser(ProtocolParser): def parse(self, data): # Implement 104 specification parsing logic print("Parsing IEC 104 data:", data) return data Modbus parser example class ModbusParser(ProtocolParser): def parse(self, data): # Implement Modbus parsing logic print("Parsing Modbus data:", data) return data # MQTT Parser Example class MQTTParser(ProtocolParser): def parse(self, data): # Implement MQTT parsing logic print("Parsing MQTT data:", data) return data # Priority Queue Manager class PriorityQueueManager: def __init__(self): self.queue = [] def add_task(self, priority, task): heapq.heappush(self.queue, (-priority, task)) def get_task(self): if self.queue: _, task = heapq.heappop(self.queue) return task return None # Protocol parser plugin system class ProtocolParserPluginSystem: def __init__(self): self.parsers = {} def register_parser(self, protocol, parser): self.parsers[protocol] = parser def get_parser(self, protocol): return self.parsers.get(protocol) # Communication resource allocator class CommunicationResourceAllocator: def allocate_resources(self, task): # Allocate resources according to task priority if isinstance(task, IEC104Parser): print("Allocating high-priority resources for IEC 104 task") else: print("Allocating normal resources for task") # Task scheduler class TaskScheduler: def __init__(self, queue_manager, parser_system, resource_allocator): self.queue_manager = queue_manager self.parser_system = parser_system self.resource_allocator = resource_allocator def run(self): while True: task = self.queue_manager.get_task() if task: protocol = type(task).__name__.replace("Parser", "").lower() protocol = type(task).__name__.replace("Parser", "").lower() if parser: self.resource_allocator.allocate_resources(task) data = task.parse("Sample data") #Simulate sending processing results print("Sending processed data:", data) # Main program if __name__ == "__main__": # Initialize each component queue_manager = PriorityQueueManager() parser_system = ProtocolParserPluginSystem() resource_allocator = CommunicationResourceAllocator() scheduler = TaskScheduler(queue_manager, parser_system, resource_allocator) # Register protocol parser parser_system.register_parser("iec104", IEC104Parser()) parser_system.register_parser("modbus", ModbusParser()) parser_system.register_parser("mqtt", MQTTParser()) # Add tasks to the priority queue queue_manager.add_task(0.8, IEC104Parser()) queue_manager.add_task(0.2, ModbusParser()) queue_manager.add_task(0.2, MQTTParser()) # Start the task scheduler scheduler.run() Therefore, multi-protocol communication middleware can support the processing of multiple industrial communication protocols on the same gateway and dynamically schedule resources according to task priority, enhancing the system's compatibility and interoperability, enabling it to seamlessly connect with equipment of different types and brands, simplifying the access process of new protocols, reducing development and integration costs, and improving resource management and scheduling efficiency as well as the ability to quickly adapt to new protocols.

[0035] According to one embodiment of the present invention, a link quality real-time monitoring module includes: a first data acquisition unit, used to collect status parameters of the main communication link, and obtain the network operation status according to the status parameters, wherein the status parameters include at least one of network signal strength, packet loss rate and data round-trip delay; an indicator calculation unit, used to calculate a first statistical value of network signal strength, a second statistical value of packet loss rate and a third statistical value of data round-trip delay based on a preset sliding window algorithm; a health assessment unit, used to calculate the health index of the main communication link based on the first statistical value of network signal strength, the second statistical value of packet loss rate and the third statistical value of data round-trip delay; a pre-switching trigger unit, used to trigger the pre-switching mechanism when the health index of the communication link is less than a preset health index.

[0036] The preset health index may be set by those skilled in the art according to actual test requirements, or may be obtained through a limited number of computational simulations, and is not specifically limited here.

[0037] Specifically, the link quality real-time monitoring module includes a first data acquisition unit, an indicator calculation unit, a health assessment unit, and a pre-switching trigger unit. The first data acquisition unit is responsible for real-time collection of multi-dimensional data such as signal strength, a second statistical value of packet loss rate, and a third statistical value of data round-trip time (RTT); the indicator calculation unit uses a sliding window algorithm (window size 10s) to process the collected data and calculate the statistical values ​​of various indicators; the health assessment unit calculates the health index (0-100) of the main communication link based on the first statistical value of network signal strength, the second statistical value of packet loss rate, and the third statistical value of data round-trip time delay; and the pre-switching trigger unit triggers the pre-switching mechanism when the health index of the communication link is less than a preset health index (for example, the health index of the communication link is less than 60), and establishes a long-distance link connection in advance.

[0038] Specifically, the first data acquisition unit primarily obtains the signal strength (in dBm) of the current link through a network interface or related hardware devices, sends a certain number of test data packets, counts the number of lost data packets, calculates the packet loss rate, sends a test data packet and records the sending time. After receiving a response data packet, it records the receiving time and calculates the difference between the two to obtain the round-trip delay. The indicator calculation unit primarily uses a sliding window algorithm (window size 10s) to process the collected data. The sliding window continuously moves forward, retaining only data within the most recent 10s. At each time point, the statistical values ​​of the data within the window, such as the average, maximum, and minimum values, are calculated. The health assessment unit calculates the link health index based on the calculated statistical values ​​of each indicator using a weighted summation method. For example, different weights can be assigned to signal strength, packet loss rate, and round-trip delay. The normalized values ​​of each indicator are then multiplied by the corresponding weights and added to obtain the health index. The pre-handover trigger unit triggers a pre-handover mechanism when the link health index is less than 60. The pre-handover mechanism involves establishing a long-distance link connection in advance to shorten the handover time to less than 200ms when it is needed.

[0039] Therefore, by combining multiple key performance indicators (such as signal strength, packet loss rate, and round-trip delay) and taking a weighted sum, a comprehensive health index is generated to provide a more comprehensive and accurate link quality assessment. When the health index falls below the preset threshold, the pre-switching mechanism is automatically triggered to ensure that preventive measures are taken before link failure occurs to avoid service interruption.

[0040] According to one embodiment of the present invention, the automatic switching execution module includes: a network status monitoring unit, which is used to monitor the optical fiber connection status and the preset network interface connection status of the main communication link based on the heartbeat packet and the network interface status, and determine that the network operation status is in an interrupted state when no heartbeat packet is received for a preset number of consecutive times or the preset network interface is detected to be disconnected; a long-distance self-organizing network unit, which is used to dynamically adjust the spreading factor according to the health index of the main communication link by using a dynamic spreading factor adaptive algorithm when the network operation status is in an interrupted state; a protocol switching control unit, which is used to initialize the long-distance wireless communication protocol (such as the LoRa protocol) when the network operation status is in an interrupted state, and configure the parameters of the long-distance wireless communication protocol to control the communication link of the autonomous fault-tolerant system to switch to the backup communication link according to the configured long-distance wireless communication protocol; and an instruction broadcast unit, which is used to send broadcast instructions to target nodes within a preset range based on the long-distance wireless communication protocol.

[0041] The preset network interface and the preset range can be set according to actual test requirements and are not specifically limited here.

[0042] Specifically, the automatic switching execution module mainly includes a network status monitoring unit, a long-distance self-organizing network unit, a protocol switching control unit and an instruction broadcast unit. Among them, the network status monitoring unit monitors the connection status of the optical fiber and 4G network in real time, and notifies the protocol switching control module when a network interruption is detected; the long-distance self-organizing network unit integrates the long-distance self-organizing network function that supports Class A / Class B device models, and adopts a dynamic spreading factor (SF7-SF12) adaptive algorithm to dynamically adjust the spreading factor according to the link quality; after receiving the network interruption notification, the protocol switching control unit completes the switch from the optical fiber / 4G protocol to the long-distance wireless communication protocol within 500ms; after switching to the long-distance wireless communication protocol, the instruction broadcast unit broadcasts instructions to devices within 3 kilometers through LoRa, and adopts frequency hopping technology (FHSS, 16 channels) to avoid electromagnetic interference.

[0043] Specifically, if Figure 3 As shown, the network status monitoring unit monitors the connection status of the optical fiber and 4G networks in real time by regularly sending heartbeat packets or monitoring the status of the preset network interface. When no heartbeat response is received for a preset number of times (e.g., 3 times) or the preset network interface is detected to be disconnected, the network is determined to be disconnected. The long-distance self-organizing network unit supports Class A and Class B LoRa device models to adapt to different application scenarios, and dynamically adjusts the spreading factor (SF7-SF12) based on the health index of the main communication link, that is, the link quality (such as signal strength, packet loss rate, etc.). When the link quality is good, a smaller spreading factor is used to increase the transmission rate; when the link quality is poor, a larger spreading factor is used to improve the transmission reliability. The protocol switching control unit is used to initialize the long-distance wireless communication protocol and configure the parameters of the long-distance wireless communication protocol when the network operation status is interrupted, so as to control the communication link of the autonomous fault-tolerant system to complete the switching from the optical fiber / 4G protocol to the backup communication link long-distance wireless communication protocol within 500ms according to the configured long-distance wireless communication protocol. In addition, after receiving the network interruption notification, the current optical fiber / 4G protocol is quickly stopped. The system communicates with the long-distance wireless communication protocol, initializes the long-distance wireless communication protocol, and configures the parameters of the long-distance wireless communication protocol (such as frequency, spreading factor, transmission power, etc.) to control the communication link of the autonomous fault-tolerant system according to the configured long-distance wireless communication protocol to complete protocol switching within 500ms; the instruction broadcast unit is used to send broadcast instructions to the target nodes within a preset range based on the long-distance wireless communication protocol, and after switching to the long-distance wireless communication protocol, broadcast instructions to devices within a preset range (for example, 3 kilometers) through LoRa, using frequency hopping technology (FHSS, 16 channels) to send instructions in turn on different channels to avoid electromagnetic interference and improve the reliable transmission rate of instructions.

[0044] Therefore, the automatic switching execution module completes the switching from the main link to the backup link in a very short time (for example, within 500ms) after the main communication link of the network fails or its performance degrades, ensuring that key business tasks (such as power protection instructions, data collection, etc.) are not interrupted. At the same time, by using heartbeat packets and interface status monitoring technology, it can monitor the status of the main communication link in real time, and when no heartbeat packets are received for multiple consecutive times or the interface is detected to be disconnected, it can quickly determine that the network operation status is in an interrupted state. Once it is determined that switching is required, the pre-switching process is immediately started, which greatly shortens the time window from fault detection to recovery and reduces the possibility of business interruption.

[0045] According to one embodiment of the present invention, the fault prediction module 300 includes: a second data acquisition unit, which is used to collect the operating status data of each hardware device; a data processing unit, which is used to process the operating status data to obtain target operating status data, and input the target operating status data into the low-power wide area network communication model; and a fault prediction unit, which is used to predict the fault of the operating status of each hardware device according to the low-power wide area network communication model.

[0046] According to one embodiment of the present invention, the data processing unit includes: a data feature extraction subunit, which is used to perform fast Fourier transform and sliding mean filtering on the operating status data, and extract key data features of the operating status data to obtain target operating status data based on the key data features.

[0047] Specifically, the fault prediction module 300 predicts potential fault risks by real-time monitoring and analyzing the operating status data of the hardware devices, and triggers preventive measures when necessary.

[0048] Specifically, first, the second data acquisition unit collects operating status data of each hardware device from various sensors, PLCs (programmable logic controllers), relay protection devices and other hardware devices, including but not limited to multi-dimensional data such as temperature, vibration spectrum, current fluctuation, and voltage change. Second, the data feature extraction sub-unit performs fast Fourier transform and sliding mean filtering on the operating status data collected, and extracts key data features of the operating status data to obtain target operating status data based on the key data features. At the same time, it can smooth noise, reduce data fluctuations, and improve the accuracy of feature extraction. Third, the target operating status data is input into the low-power wide area network communication model, and the fault prediction unit performs fault prediction on the operating status of each hardware device based on the low-power wide area network communication model. The low-power wide area network communication model can be trained offline using historical fault data sets, and the model parameters are updated regularly to ensure that it can reflect the latest device operating status. The processed feature data is then received online, and the trained low-power wide area network communication model is used to perform real-time prediction, outputting the device's health score and failure probability. If the failure probability of a hardware device exceeds a preset threshold, an early warning signal is immediately generated.

[0049] Therefore, the fault prediction module 300 plays an important role in ensuring the healthy operation of hardware equipment through its real-time monitoring, data processing, fault prediction and early warning mechanism. It not only improves the overall performance and stability of the system, but also provides strong support for various complex industrial application scenarios. These features work together to form an intelligent and efficient fault prediction and prevention system, which meets the strict requirements of reliability and real-time performance in modern edge computing environments.

[0050] According to one embodiment of the present invention, a lightweight containerized fault-tolerant engine 200 includes: a monitoring unit for monitoring the operating status of each hardware device, and when the operating status is a fault state or a fault risk state, using a preset migration technology to migrate the faulty hardware devices at risk of failure and the faulty hardware devices in a fault state to a target node, and while the target node resumes the execution of business tasks, the business tasks are executed based on the communication link of the autonomous fault-tolerant system.

[0051] The preset migration technology can be set by those skilled in the art according to real-time testing requirements and is not specifically limited here. Specifically, if Figure 4As shown, the monitoring unit in the lightweight containerized fault-tolerant engine 200 can monitor the operating status of each hardware device. For example, the operating status data of the hardware device (such as temperature, voltage, current, etc.) can be collected through sensors or other monitoring means, and the changes in these data can be continuously monitored to determine whether there is a potential failure risk. When an actual failure of the hardware device is detected (such as overheating, short circuit, etc.), an emergency migration instruction is immediately triggered. When the first hardware device migration instruction or the second hardware device migration instruction in a faulty state is received from the AI-driven fault prediction module, the faulty hardware device with failure risk and the faulty hardware device in a faulty state are migrated to the target node in advance. Then, according to the load of the current node, the resource allocation strategy is dynamically adjusted to optimize the performance during the migration process. The lightweight orchestration framework is used to implement intelligent scheduling of tasks to ensure that high-priority tasks are processed first, and while the target node resumes the execution of business tasks, the business tasks are executed based on the communication link of the autonomous fault-tolerant system.

[0052] Therefore, when a hardware failure or potential risk is detected, the lightweight containerized fault-tolerant engine can migrate the affected container to the target node in a very short time (such as within 5 seconds), ensuring that critical business tasks are not interrupted. At the same time, through the instant migration mechanism, the service downtime caused by hardware failure is significantly reduced, thereby improving the overall availability of the system.

[0053] The following will be described with two specific scenario embodiments: Example 1: Smart City and IoT Deployment In smart campuses and intelligent buildings, edge computing gateways connect IoT devices such as water meters, electricity meters, and fire sensors to enable data collection and command control. When 4G / 5G networks are interrupted by base station failures, remote ad hoc networks ensure real-time transmission of emergency commands such as "fire alarm" and "elevator emergency stop," preventing safety incidents caused by communication interruptions. The containerized engine supports the rapid migration of business logic such as "lighting control" and "air conditioning scheduling" when gateway hardware fails due to aging, ensuring uninterrupted smart campus services. AI models analyze elevator vibration data and abnormal water meter flow fluctuations to proactively identify potential equipment failures and guide property management personnel in precise maintenance, reducing inspection workload and spare parts inventory costs by 60%.

[0054] Example 2: Smart Grid and Power Automation In scenarios such as substations and distribution stations, edge computing gateways are deployed between relay protection devices and measurement and control equipment to enable real-time issuance of protection commands and data collection. When fiber-optic communications are interrupted due to lightning strikes, construction, or other reasons, the dual-mode communication hot-swap mechanism automatically activates the long-distance ad hoc network, ensuring that key commands such as "trip" and "reclose" are broadcast to adjacent devices via the local network within 200ms. This maintains the integrity of regional protection logic and prevents the escalation of power grid accidents caused by communication interruptions. A lightweight containerized engine can migrate protection control logic to a backup gateway within 5 seconds in the event of a hardware failure in the measurement and control device, ensuring uninterrupted real-time services such as "differential protection." An AI-driven fault prediction model analyzes transformer temperature rise and circuit breaker vibration data to predict equipment failures 24 hours in advance, guiding operations and maintenance personnel in precise repairs and reducing manual inspection costs by 90%.

[0055] In summary, based on the specific discussion of the above embodiments, the embodiments of the present invention can achieve the following beneficial effects: (1) Breakthrough in communication reliability: Existing technologies rely on a single link, and manual recovery is required when the network is interrupted (taking several hours). This invention uses a dual-mode communication hot-switching mechanism to automatically switch to the LoRa self-organizing network within 200ms, ensuring that commands are continuously issued. For example, when the optical fiber of a substation is interrupted, the regional protection logic can still be maintained, and the response speed is increased by 99% compared with traditional solutions; (2) Hardware fault self-healing capability: Traditional solutions rely on manual troubleshooting (MTTR exceeds 24 hours). This invention uses a lightweight containerization engine to complete fault container migration within 5 seconds, with business interruption time less than 50ms, reducing production line downtime losses by 80%, and achieving fully automatic self-healing of hardware faults. (3) Fault prediction and active defense: Existing technologies can only respond after the fact. This invention integrates the TinyML model to predict faults 24 hours in advance (with an accuracy of 92%) and trigger active migration 18 hours in advance of the threshold alarm, avoiding business interruption when a fault occurs. At the same time, it reduces cloud data transmission by 90%, reducing bandwidth costs and privacy risks. (4) An algorithm for dynamically allocating communication resources based on service priorities, especially a preemptive scheduling strategy for power protection instructions, solves the problem that traditional polling cannot meet real-time requirements. A container rapid migration method based on the Checkpoint / Restore mechanism ensures that faulty containers can be recovered within 5 seconds. An architectural design that supports dynamic loading of protocol modules, including hot-swappable interfaces, version management, and conflict detection mechanisms for protocol modules, ensures rapid adaptation to new protocols within 72 hours. Overall, this invention achieves comprehensive optimization in terms of reliability, real-time performance, and operation and maintenance costs, filling the technical gap in autonomous fault tolerance for edge devices.

[0056] According to the autonomous fault-tolerant system with dual-mode hot switching proposed in an embodiment of the present invention, when the main communication link fails, the dual-mode communication hot switching module is used to switch the main communication link to the backup communication link, and when any hardware device among multiple hardware devices is at risk of failure, a first hardware device migration instruction is generated based on the fault prediction module, and at the same time, a second hardware device migration instruction is generated when any hardware device is in a faulty state. Based on the first and second hardware device migration instructions, the faulty hardware devices at risk of failure and in a faulty state are migrated to the target node based on a lightweight containerized fault-tolerant engine, and while the business task execution is restored at the target node, the business task is executed based on the communication link of the autonomous fault-tolerant system, which solves the technical problems of long response time, delay and high cost caused by manual operation and maintenance and cloud redundant backup. By adopting the dual-mode communication design concept, each functional module works together to achieve a complete closed loop from fault monitoring to autonomous recovery.

[0057] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.

[0058] The embodiment of the present invention further provides an autonomous fault-tolerant method for dual-mode hot switching, the method adopts any of the above-mentioned autonomous fault-tolerant systems for dual-mode hot switching, such as Figure 5 As shown, the following steps are included: In step S501, the primary communication link in the autonomous fault-tolerant system is monitored, and when the primary communication link is in a fault state, the communication link of the autonomous fault-tolerant system is controlled to switch to a backup communication link; In step S502, the operating status of each hardware device is monitored, and a fault prediction is performed based on the operating status of each hardware device. When a hardware device has a fault risk, a first hardware device migration instruction is triggered, and when any hardware device among the multiple hardware devices is detected to be in a fault state, a second hardware device migration instruction is generated; In step S503, according to the first hardware device migration instruction and the second hardware device migration instruction, the faulty hardware device is migrated to the target node, and the execution of the business task is resumed at the target node, while the business task is continued to be executed based on the backup communication link.

[0059] According to one embodiment of the present invention, monitoring a primary communication link in an autonomous fault-tolerant system and controlling the communication link of the autonomous fault-tolerant system to switch to a backup communication link when the primary communication link is in a fault state includes: Collecting status parameters of the primary communication link and obtaining a network operation status based on the status parameters, wherein the status parameters include at least one of network signal strength, packet loss rate, and data round-trip delay; Calculate a first statistical value of network signal strength, a second statistical value of packet loss rate, and a third statistical value of data round-trip delay based on a preset sliding window algorithm; The health index of the main communication link is calculated based on the first statistical value of the network signal strength, the second statistical value of the packet loss rate and the third statistical value of the data round-trip delay, and the pre-switching mechanism is triggered when the health index of the communication link is less than the preset health index.

[0060] According to the autonomous fault-tolerant method of dual-mode hot switching proposed in an embodiment of the present invention, when the main communication link fails, the dual-mode communication hot switching module is used to switch the main communication link to the backup communication link, and when any hardware device among multiple hardware devices is at risk of failure, a first hardware device migration instruction is generated based on the fault prediction module, and at the same time, a second hardware device migration instruction is generated when any hardware device is in a faulty state. Based on the first and second hardware device migration instructions, the faulty hardware devices at risk of failure and in a faulty state are migrated to the target node based on a lightweight containerized fault-tolerant engine, and while the business task execution is restored at the target node, the business task is executed based on the communication link of the autonomous fault-tolerant system, which solves the technical problems of long response time, delay and high cost caused by manual operation and maintenance and cloud redundant backup. By adopting the dual-mode communication design concept, each functional module works together to achieve a complete closed loop from fault monitoring to autonomous recovery.

[0061] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0062] The above is a detailed introduction to the dual-mode hot-switch autonomous fault-tolerant system provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only intended to help understand the method and core ideas of the present invention. It should be pointed out that, for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. An autonomous fault-tolerant system with dual-mode hot switching, characterized in that: include: a dual-mode communication hot-swap module, which controls the communication link of the autonomous fault-tolerant system to switch to the backup communication link when detecting that the primary communication link in the autonomous fault-tolerant system is in a fault state; a fault prediction module, configured to predict faults based on the operating status of each hardware device and generate a first hardware device migration instruction when any hardware device among the plurality of hardware devices has a risk of failure; A lightweight containerized fault-tolerant engine is configured to generate a second hardware device migration instruction when any hardware device among the multiple hardware devices is in a faulty state, and based on the first hardware device migration instruction and the second hardware device migration instruction, migrate the faulty hardware device at risk of failure and the faulty hardware device in the faulty state to a target node, and while the target node resumes execution of the business task, execute the business task based on the communication link of the autonomous fault-tolerant system.

2. The autonomous fault-tolerant system for dual-mode hot switching according to claim 1, characterized in that: The dual-mode communication hot-swap module includes: Multi-protocol communication middleware, used to allocate communication resources according to the priority of the multi-protocol business tasks of the autonomous fault-tolerant system, and to schedule tasks based on the priority of each business task; A link quality real-time monitoring module, configured to monitor the health index of the primary communication link in real time and generate a switching instruction of a pre-switching mechanism according to the health index; The automatic switching execution module is used to monitor the network operation status of the main communication link and execute the link switching action of the autonomous fault-tolerant system based on the network operation status.

3. The autonomous fault-tolerant system for dual-mode hot switching according to claim 2, characterized in that: The multi-protocol communication middleware includes: A priority queue management unit, configured to prioritize the multi-protocol service tasks based on a preset priority quality of service algorithm to obtain a priority queue for the multi-protocol service tasks; The protocol parser plug-in unit is used to parse the business tasks corresponding to each protocol based on the plug-in architecture, so that the business tasks corresponding to each protocol after parsing are adapted to the new protocol; a communication resource allocation unit, configured to allocate communication resources based on the priority of the multi-protocol service task; The task scheduling unit is used to select a target business task based on the priority queue and according to the priority of the multi-protocol business task, and call the corresponding protocol parser plug-in unit for parsing according to the protocol type of the selected target business task.

4. The autonomous fault-tolerant system for dual-mode hot switching according to claim 2, characterized in that: The link quality real-time monitoring module includes: a first data acquisition unit, configured to acquire state parameters of the primary communication link and obtain the network operation state according to the state parameters, wherein the state parameters include at least one of network signal strength, packet loss rate, and data round-trip delay; An indicator calculation unit, configured to calculate a first statistical value of the network signal strength, a second statistical value of the packet loss rate, and a third statistical value of the data round-trip delay based on a preset sliding window algorithm; a health evaluation unit, configured to calculate a health index of the primary communication link based on a first statistical value of the network signal strength, a second statistical value of the packet loss rate, and a third statistical value of the data round-trip delay; The pre-switching triggering unit is used to trigger the pre-switching mechanism when the health index of the communication link is less than a preset health index.

5. The autonomous fault-tolerant system for dual-mode hot switching according to claim 2, characterized in that: The automatic switching execution module includes: a network status monitoring unit, configured to monitor the optical fiber connection status of the primary communication link and the connection status of the preset network interface based on the heartbeat packet and the network interface status, and determine that the network operation status is in an interrupted state when the heartbeat packet is not received for a preset number of consecutive times or the preset network interface is detected to be disconnected; a long-distance self-organizing network unit, configured to dynamically adjust the spreading factor according to the health index of the main communication link using a dynamic spreading factor adaptive algorithm when the network operation state is in an interrupted state; a protocol switching control unit, configured to initialize a long-distance wireless communication protocol and configure parameters of the long-distance wireless communication protocol when the network operation state is in an interrupted state, so as to control the communication link of the autonomous fault-tolerant system to switch to a backup communication link according to the configured long-distance wireless communication protocol; The instruction broadcast unit is used to send a broadcast instruction to a target node within a preset range based on the long-distance wireless communication protocol.

6. The autonomous fault-tolerant system for dual-mode hot switching according to claim 1, characterized in that: The lightweight containerized fault-tolerant engine includes: A monitoring unit is used to monitor the operating status of each hardware device, and when the operating status is the fault state or the fault risk state, use a preset migration technology to migrate the faulty hardware devices at the fault risk and the faulty hardware devices in the fault state to a target node, and while the target node resumes the execution of the business task, execute the business task based on the communication link of the autonomous fault-tolerant system.

7. The autonomous fault-tolerant system for dual-mode hot switching according to claim 1, characterized in that: The fault prediction module includes: A second data acquisition unit is used to collect the operating status data of each hardware device; a data processing unit, configured to process the operating status data to obtain target operating status data, and input the target operating status data into a low power wide area network communication model; A fault prediction unit is used to perform fault prediction on the operating status of each hardware device according to the low-power wide area network communication model.

8. The autonomous fault-tolerant system for dual-mode hot switching according to claim 7, characterized in that: The data processing unit includes: The data feature extraction subunit is used to perform fast Fourier transform and sliding mean filtering on the operating status data, and extract key data features of the operating status data to obtain the target operating status data according to the key data features.

9. An autonomous fault-tolerant method for dual-mode hot switching, characterized in that: The method adopts the autonomous fault-tolerant system for dual-mode hot switching according to any one of claims 1 to 8, comprising the following steps: monitoring a primary communication link in the autonomous fault-tolerant system, and controlling the communication link of the autonomous fault-tolerant system to switch to a backup communication link when the primary communication link is in a fault state; Monitor the operating status of each hardware device, perform fault prediction based on the operating status of each hardware device, trigger a first hardware device migration instruction when the hardware device is at risk of failure, and generate a second hardware device migration instruction when any hardware device among the multiple hardware devices is detected to be in a faulty state; According to the first hardware device migration instruction and the second hardware device migration instruction, the faulty hardware device in the faulty state is migrated to the target node, and the execution of the business task is resumed at the target node, while the business task is continued to be executed based on the backup communication link.

10. The autonomous fault-tolerant method for dual-mode hot switching according to claim 9, characterized in that: The monitoring of the primary communication link in the autonomous fault-tolerant system and controlling the communication link of the autonomous fault-tolerant system to switch to a backup communication link when the primary communication link is in a fault state includes: Collecting state parameters of the primary communication link and obtaining a network operation state according to the state parameters, wherein the state parameters include at least one of network signal strength, packet loss rate, and data round-trip delay; Calculate the first statistical value of the network signal strength, the second statistical value of the packet loss rate, and the third statistical value of the data round-trip delay based on a preset sliding window algorithm; The health index of the main communication link is calculated based on the first statistical value of the network signal strength, the second statistical value of the packet loss rate and the third statistical value of the data round-trip delay, and the pre-switching mechanism is triggered when the health index of the communication link is less than the preset health index.

Images