5G message multichannel redundant transmission method and system based on AI self-healing mechanism

Abstract

The invention discloses a 5G message multi-channel redundant transmission method and system based on an AI self-healing mechanism, and relates to the technical field of data communication, and the method comprises the steps: deploying a distributed probe array between terminal equipment and a base station, collecting multi-channel transmission parameters in real time, inputting a hybrid prediction model to predict the transmission quality attenuation trend of each channel, and based on the prediction score, performing reinforcement learning by using a self-healing decision engine, dynamically generating a multi-system channel combination, and performing transmission path reconstruction and switching. And deploying a lightweight block chain node to verify the actual transmission performance of each channel, and performing compensation updating. The technical problems that an existing 5G message transmission method depends on a static redundant channel and cannot deal with network fluctuation and faults in real time, so that the transmission quality is unstable and the resource utilization rate is low are solved, and the purposes that the transmission path is adjusted in real time through combination optimization of an AI self-healing mechanism and a dynamic channel, and the transmission efficiency is improved are achieved. And the stability of 5G message transmission and the resource utilization efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data communication, and particularly to a 5G message multi-channel redundant transmission method and system based on an AI self-healing mechanism. Background Art

[0002] With the wide application of 5G technology, 5G Messaging plays an important role in fields such as smart home, Internet of Things, and autonomous driving. However, with the increasing complexity and diversity of the network environment, traditional 5G message transmission methods face many challenges, such as network fluctuations, device failures, signal interference, etc., resulting in a decline in transmission quality, an increase in latency, and even message loss. Existing transmission methods usually rely on statically configured redundant channels, lack the ability to adapt to network changes in real time, and fail to achieve automatic repair and flexible optimization. As a result, in a complex and dynamically changing network environment, network resources cannot be fully utilized, and it is difficult to ensure the reliability and stability of transmission. Summary of the Invention

[0003] This application provides a 5G message multi-channel redundant transmission method and system based on an AI self-healing mechanism, which is used to solve the technical problems that existing 5G message transmission methods rely on static redundant channels and cannot respond to network fluctuations and failures in real time, resulting in unstable transmission quality and low resource utilization rate.

[0004] In the first aspect of this application, a 5G message multi-channel redundant transmission method based on an AI self-healing mechanism is provided. The method includes: deploying a distributed probe array between a terminal device and a base station, and collecting multi-channel transmission parameters in real time based on the distributed probe array; inputting the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality decay trend of each channel within a preset time window, and generating a channel health score; based on the channel health score, performing reinforcement learning through a self-healing decision engine to dynamically generate a multi-mode channel combination; according to the multi-mode channel combination, reconstructing a transmission path to generate an updated transmission path, and switching the transmission path according to the updated transmission path; deploying lightweight blockchain nodes to verify the actual transmission performance of each channel in real time, and performing channel compensation update according to the actual transmission performance.

[0005] In the second aspect of the present application, a 5G message multi-channel redundant transmission system based on an AI self-healing mechanism is provided. The system includes: a dynamic perception module, which is used to deploy a distributed probe array between a terminal device and a base station, and collect multi-channel transmission parameters in real time based on the distributed probe array; an intelligent prediction module, which is used to input the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality attenuation trend of each channel within a preset time window, and generate a channel health score; a policy generation module, which is used to perform reinforcement learning through a self-healing decision engine based on the channel health score to dynamically generate a multi-mode channel combination; a self-healing execution module, which is used to reconstruct the transmission path according to the multi-mode channel combination to generate an updated transmission path, and perform a transmission path switch according to the updated transmission path; a verification feedback module, which is used to deploy lightweight blockchain nodes to verify the actual transmission performance of each channel in real time, and perform channel compensation update according to the actual transmission performance.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages: The 5G message multi-channel redundant transmission method and system based on the AI self-healing mechanism provided in the present application relate to the technical field of data communication. By deploying a distributed probe array to collect multi-channel transmission parameters in real time, and using a hybrid prediction model to predict the transmission quality attenuation trend to generate a channel health score, based on this score, a self-healing decision engine is used to perform reinforcement learning to dynamically generate a multi-mode channel combination, optimize the transmission path and switch, which solves the technical problem that the existing 5G message transmission method relies on static redundant channels and cannot respond to network fluctuations and failures in real time, resulting in unstable transmission quality and low resource utilization rate. The technical effect of realizing real-time adjustment of the transmission path through the AI self-healing mechanism and dynamic channel combination optimization, and improving the stability and resource utilization efficiency of 5G message transmission is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0008] Figure 1 It is a schematic flowchart of a 5G message multi-channel redundant transmission method based on an AI self-healing mechanism provided by an embodiment of the present application; Figure 2 It is a schematic structural diagram of a 5G message multi-channel redundant transmission system based on an AI self-healing mechanism provided by an embodiment of the present application.

[0009] Description of reference numerals: Dynamic perception module 11, intelligent prediction module 12, policy generation module 13, self-healing execution module 14, verification feedback module 15. Detailed implementation manners

[0010] This application provides a 5G message multi-channel redundant transmission method and system based on an AI self-healing mechanism, which is used to solve the technical problems that the existing 5G message transmission method relies on static redundant channels and cannot respond to network fluctuations and faults in real time, resulting in unstable transmission quality and low resource utilization.

[0011] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the protection scope of this application.

[0012] It should be noted that the terms "first", "second", etc. in the specification of this application and the above accompanying drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, methods, products or devices.

[0013] Embodiment 1, as Figure 1 shown, this application provides a 5G message multi-channel redundant transmission method based on an AI self-healing mechanism, and the method includes: P10: Deploy a distributed probe array between the terminal device and the base station, and collect multi-channel transmission parameters in real time based on the distributed probe array. Among them, the multi-channel transmission parameters include physical layer parameters, protocol layer parameters, and network layer parameters.

[0014] It should be understood that a distributed probe array is deployed between the terminal device and the base station. The probe array is composed of multiple probe nodes, which are distributed at multiple positions in the network and can realize real-time monitoring and data collection of multiple communication channels. The advantage of the distributed probe array is that it can cover a wide area and obtain real-time transmission information of multiple channels, providing accurate network status feedback. These probe nodes collect real-time transmission data of each communication link through communication with the base station and the terminal device.

[0015] The collected multi-channel transmission parameters include data at three main levels: physical layer parameters, protocol layer parameters, and network layer parameters. Specifically, physical layer parameters refer to the basic attributes of wireless signals, such as signal strength, frequency, bandwidth, modulation method, bit error rate, etc., which directly affect the signal transmission quality. By monitoring physical layer parameters in real time, the system can evaluate the signal quality of the current communication link and analyze problems such as signal attenuation and interference.

[0016] Protocol layer parameters mainly refer to various types of data in the communication protocol stack, including transmission delay, packet loss rate, number of retransmissions, throughput, etc. These parameters reflect the efficiency and reliability of the protocol in data transmission. By analyzing protocol layer data, it can be determined whether packets encounter congestion, retransmission, or other situations with low transmission efficiency during transmission, thus providing guidance for network optimization.

[0017] Network layer parameters include routing information, network topology, routing delay, network load, etc. These parameters help evaluate whether the data transmission path in the network is unobstructed and whether there are problems such as congestion and path switching. In multi-channel redundant transmission, network layer parameters are particularly important because they determine the scheduling and selection of messages between multiple transmission channels, thus affecting the final data transmission efficiency.

[0018] In actual operation, the distributed probe array is connected to the terminal device and the base station through a high-speed data interface, collects the above multi-channel transmission parameters in real time, and transmits the collected data to the backend processing system. To ensure the accuracy and real-time nature of the data, the probe nodes adopt high-precision sensors and fast data processing chips, which can complete data collection and transmission within milliseconds. At the same time, to cope with complex communication environments, the probe array also has an adaptive adjustment function, which can dynamically adjust monitoring parameters and frequencies according to different network states and service requirements.

[0019] By collecting these multi-channel transmission parameters in real time through the distributed probe array, the status of each layer of the network can be comprehensively understood. These real-time data will serve as the basis for subsequent prediction, decision-making, and optimization, helping the system dynamically adjust network resource allocation, optimize the message transmission path, and quickly respond for self-healing processing when the transmission quality deteriorates or network anomalies occur.

[0020] P20: Input the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality attenuation trend of each channel within a preset time window, and generate a channel health score.

[0021] Furthermore, step P20 of the embodiment of the present application further includes: P21: Perform frequency-domain feature decomposition on the physical layer parameters, and extract the high-frequency fluctuation components as the early warning indicators for quality degradation; P22: Based on the high-frequency fluctuation components, perform time-series analysis of the transmission quality to generate a quality decay trend; P23: Dynamically allocate index weights for the quality degradation early warning indicators; P24: Based on the quality decay trend, combined with the index weights, calculate and generate a channel health score.

[0022] It should be understood that the collected multi-channel transmission parameters, including various data of the physical layer, protocol layer, and network layer, are input into the hybrid prediction model for processing. Through comprehensive analysis of these transmission parameters, combined with historical data and real-time information, the model predicts the transmission quality decay trend of each channel within a preset time window. The decay trend reflects the changes of parameters such as signal strength, delay, and bit error rate of the channel over time. Based on these prediction results, a channel health score is generated, which reflects the current health status of each channel and provides a basis for subsequent optimization and decision-making.

[0023] Specifically, first perform frequency-domain feature decomposition on the physical layer parameters, and convert these parameters from the time domain to the frequency domain through mathematical tools such as Fourier transform. This process can effectively extract the high-frequency fluctuation components, which are the key early warning indicators for quality degradation. For example, in the millimeter-wave band, high-frequency fluctuations may indicate that the signal is affected by environmental factors (such as building occlusion or weather changes), resulting in a decline in transmission quality. By extracting these high-frequency fluctuation components, potential quality problems can be captured in advance.

[0024] Based on the extracted high-frequency fluctuation components, further perform time-series analysis of the transmission quality. By establishing a time series model (such as an ARIMA model or a long short-term memory network LSTM), analyze the change trend of the high-frequency fluctuation components within a preset time window. This process can generate a decay trend curve of the transmission quality of each channel, clearly showing the change of transmission quality over time. For example, if the high-frequency fluctuation components continue to increase within a certain period, it indicates that the transmission quality of the channel may be deteriorating. Through this time-series analysis, the change of transmission quality in the future period can be predicted more accurately, providing an early warning for the self-healing mechanism.

[0025] To enable the model to more flexibly adapt to different scenarios and conditions, this step also introduces a dynamic weight allocation mechanism. The index weights are dynamically allocated for the quality deterioration warning indicators. According to historical data and real-time monitoring results, machine learning algorithms (such as reinforcement learning) are used to dynamically adjust the weights of each indicator. For example, in a high-interference environment, the weight of the high-frequency fluctuation component may be increased, while in a stable environment, the weights of other indicators (such as protocol layer parameters) may be increased accordingly. This dynamic adjustment can ensure that the model can accurately reflect the actual situation of the transmission quality in different scenarios, thereby improving the adaptability and robustness of the model.

[0026] Finally, based on the generated quality decay trend and the dynamically allocated indicator weights, the channel health score is calculated. The channel health score is a comprehensive indicator used to quantify the transmission quality status of each channel. The specific calculation method is to multiply the value of each indicator by its corresponding weight and then perform a weighted sum. The higher the score value, the better the transmission quality of the channel; conversely, it indicates that the transmission quality is poor. In this way, the health status of each channel can be intuitively evaluated, providing an important basis for subsequent self-healing decisions.

[0027] In practical applications, this step can be implemented in the following way: First, standardize the multi-channel transmission parameters collected to ensure that the data format input to the model is consistent. Then, use historical data to train the hybrid prediction model and deploy it in an actual communication system. The model can run on edge computing nodes or cloud servers and make predictions based on real-time data. The channel health score output by the model can be fed back to the system in real time for subsequent self-healing decisions. At the same time, the model can also perform online incremental learning based on real-time data to continuously optimize the prediction performance.

[0028] Furthermore, to construct the hybrid prediction model, Embodiment of the present application further includes step P20a, and step P20a further includes: P21a: Adopt a hybrid neural network architecture to parallelly extract the millimeter-wave frequency band fluctuation characteristics of the physical layer parameters and the topological correlation characteristics of the protocol layer parameters; P22a: Based on the millimeter-wave frequency band fluctuation characteristics and topological correlation characteristics, inject adversarial samples simulating burst interference scenarios for model robustness training to generate the hybrid prediction model.

[0029] Optionally, the construction process of the hybrid prediction model can be further refined to enhance the model's prediction ability and robustness. Specifically, first, a hybrid neural network architecture is adopted to extract the millimeter-wave band fluctuation features in the physical layer parameters and the topological correlation features in the protocol layer parameters in parallel. The extraction of these two features is the key to the model's performance. The millimeter-wave band fluctuation features reflect the changes in the signal in the millimeter-wave band, which are usually related to signal attenuation and interference, while the topological correlation features involve the connection relationships between nodes in the network, affecting the stability and efficiency of data transmission. The hybrid neural network architecture can be a parallel architecture of an LSTM network and a temporal convolutional network, which respectively extract the temporal features and spatial correlations of the transmission parameters, that is, the millimeter-wave band fluctuation features in the physical layer parameters and the topological correlation features in the protocol layer parameters. By extracting these two features in parallel, the model can simultaneously capture the fluctuations of physical signals and the network topology changes in the protocol layer communication, thereby comprehensively evaluating the transmission quality.

[0030] After the feature extraction is completed, it enters the robustness training stage of the model. To enhance the model's adaptability to network anomalies and interference, the system injects adversarial samples in the simulated burst interference scenario into the training process. Adversarial samples are created by artificially generating signal noise or interference to simulate a harsh environment in the network. These adversarial samples can help the model maintain high robustness and prediction accuracy when facing actual interference. By introducing these adversarial samples in training, the hybrid prediction model can learn how to handle various complex and sudden network situations and improve its performance in the real network environment.

[0031] In actual operation, the hybrid neural network architecture can consist of multiple sub-networks, and each sub-network specializes in processing a type of parameter. For example, one sub-network can focus on analyzing the millimeter-wave band fluctuation features in the physical layer parameters, while another sub-network is responsible for extracting the topological correlation features of the protocol layer parameters. These sub-networks generate a comprehensive prediction result through parallel computing and information sharing. During the training process, by introducing adversarial samples, the model can learn how to adjust its prediction strategy when facing sudden interference, thereby improving the accuracy and stability of the prediction.

[0032] Through the above steps, the finally generated hybrid prediction model will have strong adaptive capabilities. It can not only predict the transmission quality based on various features of the physical layer and the protocol layer, but also maintain high prediction accuracy and stability when facing interference or emergencies, providing reliable data support for subsequent transmission optimization and path adjustment.

[0033] Furthermore, the embodiment of the present application further includes P23a: setting a dual-threshold trigger mechanism. When the signal fluctuation variance exceeds the stability threshold or the delay jitter reaches the reliability threshold, the online incremental learning unit is activated to update the hybrid prediction model.

[0034] In a possible embodiment of the present application, to further improve the real-time adaptation ability of the hybrid prediction model, a dual-threshold triggering mechanism can be set. The role of this mechanism is to monitor signal fluctuations and delay jitters in the network. When these network quality parameters exceed the preset thresholds, the online incremental learning unit is automatically activated to update the hybrid prediction model to adapt to the current changes in the network environment.

[0035] Specifically, the signal fluctuation variance refers to the degree of signal strength fluctuation over time, which reflects the signal stability in the network. If the signal fluctuation variance exceeds the stability threshold, this usually means there are large fluctuations in the network, which may lead to a decline in transmission quality. On the other hand, delay jitter refers to the variation in the arrival time intervals of data packets. Excessive delay jitter may affect real-time data transmission, such as delay-sensitive applications like video calls or online games. If the delay jitter exceeds the reliability threshold, it indicates a reduction in network reliability, which may result in data packet loss or long delays, affecting the user experience.

[0036] When either of these two parameters exceeds the set threshold, the dual-threshold triggering mechanism is activated to start the online incremental learning unit. The role of the online incremental learning unit is to dynamically update the hybrid prediction model based on newly collected real-time transmission data, enabling it to better cope with changes in the network state. Through incremental learning, the model does not need to be retrained but adjusts its parameters and prediction strategies according to the new data, thus quickly adapting to fluctuations in the network environment without interrupting the service.

[0037] The activation process of the online incremental learning unit is as follows: First, the system collects the latest transmission parameters of the current communication link, which include real-time data from the physical layer, protocol layer, and network layer. Then, these new data are input into the online incremental learning unit. The online incremental learning unit uses these new data to fine-tune the hybrid prediction model, updating the model's parameters and weights. This process does not require retraining the entire model but adapts to new data changes through local adjustments. This method not only saves computing resources but also can quickly respond to changes in the link state.

[0038] Through the combination of the dual-threshold triggering mechanism and the online incremental learning unit, the hybrid prediction model can achieve dynamic adaptive updates. For example, in a high-interference environment, the signal fluctuation variance may suddenly increase, triggering the dual-threshold mechanism. At this time, the online incremental learning unit will update the model based on the latest high-interference data, enabling the model to better predict the attenuation trend of transmission quality under this interference. Similarly, when the network becomes congested, resulting in an increase in delay jitter, the model will also be adjusted through online incremental learning to adapt to the new network state.

[0039] This mechanism can effectively ensure that the prediction model can quickly respond when facing sudden network problems, optimize the subsequent transmission paths and channel combinations, and ensure the efficient and reliable transmission of 5G messages.

[0040] P30: Based on the channel health score, perform reinforcement learning through the self-healing decision engine to dynamically generate multi-mode channel combinations.

[0041] Furthermore, step P30 of the embodiment of the present application further includes: P31: The self-healing decision engine includes a multi-objective reward function, an exploration rate dynamic decay mechanism, and heterogeneity constraint conditions; P32: Based on the quality of service requirements of delay-sensitive services and bandwidth-intensive services, construct a multi-objective reward function; P33: Adopt a double deep reinforcement learning architecture and set an exploration rate dynamic decay mechanism; P34: Set heterogeneity constraint conditions, and combine the multi-objective reward function and the exploration rate dynamic decay mechanism to optimize the channel combination strategy according to the channel health score, and generate multi-mode channel combinations. Among them, the heterogeneity constraint condition is that the channel combination includes at least three channels of different communication modes, and the proportion of non-terrestrial network channels does not exceed a preset upper limit.

[0042] Specifically, based on the channel health score, perform reinforcement learning through the self-healing decision engine to dynamically generate multi-mode channel combinations to optimize the transmission quality and stability of 5G messages. The self-healing decision engine uses the health score as input, combines the network state and service requirements, and dynamically adjusts the channel combination strategy to achieve the best transmission effect.

[0043] First of all, the self-healing decision engine includes several key components: a multi-objective reward function, an exploration rate dynamic decay mechanism, and heterogeneity constraint conditions. The design of the multi-objective reward function aims to balance multiple objectives, including the quality of service requirements of delay-sensitive services and bandwidth-intensive services. Delay-sensitive services (such as video calls, real-time control, etc.) require low latency, while bandwidth-intensive services (such as high-definition video streams, file transfers, etc.) require high bandwidth support. Therefore, the reward function needs to comprehensively consider the requirements of these two types of services and optimize the transmission effects of different services.

[0044] Secondly, the exploration rate dynamic decay mechanism realizes real-time adaptation to network state changes by setting a double deep reinforcement learning architecture. Deep reinforcement learning tries different channel combination strategies during the exploration stage and optimizes according to the effects of the strategies in the subsequent learning process. The exploration rate refers to the system's exploration ability for new strategies during the learning process. As the learning progresses, the exploration rate will gradually decrease so that the system can focus more on using known optimal strategies in a stable network environment. This mechanism ensures that the system can self-optimize in different network environments and continuously improve the accuracy of prediction and decision-making.

[0045] In addition, the self-healing decision engine also sets heterogeneity constraints to ensure the diversity and stability of channel combinations. Specifically, the heterogeneity constraints require that the channel combination contains at least three channels of different communication systems, and the proportion of non-terrestrial network channels does not exceed a preset upper limit. This constraint ensures that when the system generates a channel combination, it can make full use of different types of networks (e.g., terrestrial networks and non-terrestrial networks), avoiding relying on a single type of network and affecting the overall transmission performance. At the same time, restricting the proportion of non-terrestrial network channels can prevent over-reliance on high-latency and low-bandwidth channels such as satellite networks or aerial base stations.

[0046] In practical applications, the self-healing decision engine first screens out eligible channel combinations based on the channel health score, and then uses a multi-objective reward function to evaluate the performance of these combinations. By traversing all possible combinations and adjusting the exploration rate in combination with the exploration rate dynamic decay mechanism, the system finally selects the optimal channel combination.

[0047] By combining the above mechanisms, the system can optimize the channel combination strategy according to the channel health score and dynamically generate a multi-system channel combination suitable for the current network environment. This optimization process can select the best channel combination according to factors such as delay, bandwidth requirements, and network stability, ensuring the transmission quality of 5G messages and the efficient utilization of network resources.

[0048] Furthermore, step P34 of the embodiment of the present application further includes: P34-1: Based on the heterogeneity constraints and the channel health score, perform channel combination screening to generate a multi-system channel combination set that meets the requirements of delay-sensitive services and bandwidth-intensive services; P34-2: Use the multi-objective reward function to traverse the multi-system channel combination set, adjust the exploration rate according to the exploration rate dynamic decay mechanism, and perform dynamic combination strategy optimization to obtain the optimal multi-system channel combination.

[0049] Optionally, it is possible to further refine the generation of multi-system channel combinations that meet different service requirements based on the heterogeneity constraints and the channel health score, and obtain the optimal channel combination through dynamic combination strategy optimization.

[0050] First, channel combination screening is carried out based on heterogeneity constraints and channel health scores. Combinations that meet the heterogeneity constraints are screened out from all possible channel combinations, that is, the combination contains at least three channels of different communication systems, and the proportion of non-terrestrial network channels does not exceed a preset upper limit. This screening process ensures the diversity and stability of the channel combination, avoids the system's over-reliance on a single communication system, and thus enhances the system's risk resistance ability. At the same time, combined with the channel health scores, channel combinations in good health are further screened out to generate a multi-system channel combination set that adapts to the requirements of time-sensitive services and bandwidth-intensive services, providing diverse transmission path options for different types of services.

[0051] Next, the multi-objective reward function is used to evaluate and optimize the screened multi-system channel combination set. Each channel combination in the combination set is traversed, and the reward value of each combination is calculated according to the multi-objective reward function. The multi-objective reward function comprehensively considers the quality of service requirements of time-sensitive services and bandwidth-intensive services. By quantifying these requirements, a comprehensive reward value is assigned to each channel combination. This reward value reflects the overall performance of the combination in meeting the service requirements. During the traversal process, the system adjusts the exploration rate according to the exploration rate dynamic decay mechanism. In the initial stage of reinforcement learning, a higher exploration rate enables the system to widely explore different channel combinations and discover potential optimal combinations. As the learning process progresses, the exploration rate gradually decreases, and the system makes more use of the known optimal combinations, thereby improving the stability and efficiency of decision-making. Through this dynamic adjustment mechanism, the system can achieve a balance between exploration and exploitation and finally obtain the optimal multi-system channel combination. This optimal combination not only meets the requirements of time-sensitive services for low latency and high reliability but also meets the requirements of bandwidth-intensive services for high bandwidth and low packet loss rate, achieving the optimal matching of service requirements and system performance.

[0052] In practical applications, this process can be achieved through the following steps: First, the system screens out eligible combinations from all possible channel combinations according to heterogeneity constraints and channel health scores to form a multi-system channel combination set. Then, the system uses the multi-objective reward function to evaluate each combination in the combination set and calculate its reward value. During the evaluation process, the exploration rate is adjusted according to the exploration rate dynamic decay mechanism to balance exploration and exploitation. Through continuous iteration and optimization, the optimal multi-system channel combination is finally determined and applied to the actual communication link to optimize the transmission path and improve communication efficiency and reliability.

[0053] Finally, through the process of dynamic optimization, the system can adaptively optimize the transmission path of 5G messages according to real-time network conditions and the requirements of different services, ensuring that various requirements such as latency and bandwidth are reasonably balanced, and improving the overall transmission performance and reliability.

[0054] P40: Reconstruct the transmission path according to the multi-mode channel combination to generate an updated transmission path, and perform a transmission path switch based on the updated transmission path.

[0055] Exemplarily, reconstruct the transmission path according to the previously generated multi-mode channel combination. First, the system analyzes the current network environment and the characteristics of the selected multi-channel combination, and redesigns the data transmission path according to the quality and performance requirements of different channels. By considering the health score, delay, bandwidth, and network stability of each channel, evaluate the most suitable transmission path for the current conditions, reallocate the data stream to multiple transmission paths, or give priority to transmitting traffic to channels with better performance to ensure the efficiency and reliability of data transmission.

[0056] When the design of the new transmission path is completed, the system will perform a transmission path switch, that is, switch the data from the current transmission path to the updated path. This process needs to ensure a smooth transition of the data, avoid interruption, packet loss, or excessive delay to ensure the continuity of communication. During the switch, the system will monitor the real-time performance of the new path to ensure that the data transmission meets the predetermined quality standards. Once it is found that there are problems with the new path or it cannot meet the service requirements, the system will trigger path reconstruction and switch again to ensure that the network is always in the optimal transmission state.

[0057] In practical applications, the reconstruction and switch of the transmission path can be achieved through the following steps: First, the system analyzes and adjusts the existing transmission path by using network management software and hardware devices according to the output of the self-healing decision engine. Then, generate an updated transmission path through intelligent algorithms and verify and test it. After passing the verification, the system gradually migrates the traffic from the old path to the new path according to the preset smooth switch strategy. During the entire switch process, the system monitors the link status in real time and records relevant parameters and events for subsequent analysis and optimization.

[0058] The goal of the whole process is to dynamically adjust the transmission path according to the real-time data of the multi-mode channel combination and the network quality, so that in the face of network fluctuations, congestion, or failures, it can quickly respond and maintain the stability and efficiency of data transmission.

[0059] P50: Deploy lightweight blockchain nodes to verify the actual transmission performance of each channel in real time, and perform channel compensation update according to the actual transmission performance.

[0060] Furthermore, step P50 of the embodiment of the present application further includes: P51: Lightweight blockchain nodes are used to record the transmission data status information of each channel in real time; P52: Based on the transmission data status information, through the consensus mechanism of the blockchain, the actual transmission performance of each channel is verified to generate a transmission performance deviation, and channel compensation update is performed based on the transmission performance deviation.

[0061] It should be understood that by deploying lightweight blockchain nodes, the transmission data status information of each channel is recorded in real time. These blockchain nodes ensure the immutability and transparency of data in a decentralized manner. Whenever a new data transmission event occurs, the blockchain nodes will capture and record the relevant transmission status information in real time, including but not limited to signal strength, latency, packet loss rate, and other network performance metrics. This process ensures accurate tracking and monitoring of the transmission status of all channels, providing a reliable data source for subsequent verification and optimization.

[0062] Based on the transmission data status information recorded in real time, the actual transmission performance of each channel is verified through the consensus mechanism of the blockchain. The consensus mechanism is a decentralized verification method that ensures the consistency and authenticity of transmission performance data among all nodes. On this basis, the system will calculate the transmission performance deviation of each channel, that is, the difference between the actual transmission quality and the expected transmission quality. The transmission performance deviation may reflect unstable channel performance, faults, or failure to meet the design requirements.

[0063] Based on the generated transmission performance deviation, corresponding channel compensation update is performed. The goal of the compensation update is to correct or adjust channels with degraded performance to ensure that data transmission can be restored to the predetermined quality standard. For example, if the transmission latency of a certain channel is too high, the system may adjust the routing strategy of the channel or increase bandwidth resources; if the packet loss rate is too high, the system may optimize the signal transmission protocol or enhance the signal strength. These compensation update measures are automatically executed through smart contracts, ensuring the efficiency and transparency of the update process. A smart contract is an automatically executed blockchain contract that triggers corresponding operations automatically when preset conditions are met, without manual intervention.

[0064] In practical applications, the deployment of lightweight blockchain nodes can be achieved through the following steps: First, deploy lightweight blockchain nodes at key nodes of the communication network, which can be edge computing devices or dedicated blockchain hardware devices. Then, configure the monitoring and recording functions of the nodes to ensure that they can obtain and store transmission data status information in real time. Next, verify the recorded data through the consensus mechanism of the blockchain to generate a transmission performance deviation report. Finally, according to the deviation report, automatically execute channel compensation update through smart contracts to optimize the transmission performance of the channels.

[0065] Through this process, the lightweight blockchain nodes not only ensure the real-time recording and security of the performance data of each channel, but also verify and compensate for the transmission deviation of the channel through the consensus mechanism, realizing the automatic repair and optimization of network transmission, and further improving the reliability and adaptive ability of the overall system.

[0066] In summary, the embodiments of the present application at least have the following technical effects: In the present application, a distributed probe array is deployed between the terminal device and the base station to collect multi-channel transmission parameters in real time, and the multi-channel transmission parameters are input into a hybrid prediction model to predict the transmission quality attenuation trend of each channel, and a channel health score is generated. Based on this score, a self-healing decision engine is used for reinforcement learning to dynamically generate a multi-mode channel combination and optimize the transmission path. Through path reconstruction and switching, the data transmission quality is ensured. At the same time, lightweight blockchain nodes are deployed to verify the actual transmission performance of each channel and perform compensation updates according to the transmission performance deviation.

[0067] It achieves the technical effect of real-time adjusting the transmission path through the AI self-healing mechanism and dynamic channel combination optimization, and improving the stability and resource utilization efficiency of 5G message transmission.

[0068] Embodiment 2 is based on the same inventive concept as the 5G message multi-channel redundant transmission method based on the AI self-healing mechanism in the foregoing embodiment. As Figure 2 shown, the present application provides a 5G message multi-channel redundant transmission system based on the AI self-healing mechanism. The system in the embodiments of the present application and the method embodiments are based on the same inventive concept. Among them, the system includes: A dynamic perception module 11, configured to deploy a distributed probe array between the terminal device and the base station, and collect multi-channel transmission parameters in real time based on the distributed probe array.

[0069] An intelligent prediction module 12, configured to input the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality attenuation trend of each channel within a preset time window, and generate a channel health score.

[0070] A policy generation module 13, configured to perform reinforcement learning through a self-healing decision engine based on the channel health score to dynamically generate a multi-mode channel combination.

[0071] A self-healing execution module 14, configured to perform transmission path reconstruction according to the multi-mode channel combination to generate an updated transmission path, and perform transmission path switching according to the updated transmission path.

[0072] A verification feedback module 15, configured to deploy lightweight blockchain nodes to verify the actual transmission performance of each channel in real time, and perform channel compensation updates according to the actual transmission performance.

[0073] Further, the dynamic perception module 11 is further configured to perform the following steps: The multi-channel transmission parameters include physical layer parameters, protocol layer parameters, and network layer parameters.

[0074] Further, the intelligent prediction module 12 is further configured to perform the following steps: Adopt a hybrid neural network architecture to parallelly extract the millimeter-wave frequency band fluctuation characteristics of the physical layer parameters and the topological correlation characteristics of the protocol layer parameters; based on the millimeter-wave frequency band fluctuation characteristics and topological correlation characteristics, inject adversarial samples simulating burst interference scenarios for model robustness training to generate the hybrid prediction model.

[0075] Further, the intelligent prediction module 12 is further configured to perform the following steps: Set a dual-threshold trigger mechanism. When the signal fluctuation variance exceeds the stability threshold or the delay jitter reaches the reliability threshold, activate the online incremental learning unit to update the hybrid prediction model.

[0076] Further, the intelligent prediction module 12 is further configured to perform the following steps: Perform frequency-domain feature decomposition on the physical layer parameters, extract the high-frequency fluctuation components as quality degradation warning indicators; based on the high-frequency fluctuation components, perform time-series analysis of the transmission quality to generate a quality decay trend; dynamically allocate index weights for the quality degradation warning indicators; based on the quality decay trend, combine the index weights to calculate and generate a channel health score.

[0077] Further, the policy generation module 13 is further configured to perform the following steps: The self-healing decision engine includes a multi-objective reward function, an exploration rate dynamic decay mechanism, and heterogeneity constraint conditions; based on the quality of service requirements of delay-sensitive services and bandwidth-intensive services, construct a multi-objective reward function; adopt a double deep reinforcement learning architecture to set an exploration rate dynamic decay mechanism; set heterogeneity constraint conditions, and combine the multi-objective reward function and the exploration rate dynamic decay mechanism to optimize the channel combination strategy according to the channel health score to generate a multi-mode channel combination. The heterogeneity constraint condition is that the channel combination includes at least three channels of different communication systems, and the proportion of non-terrestrial network channels does not exceed a preset upper limit.

[0078] Further, the policy generation module 13 is further configured to perform the following steps: Based on the heterogeneous constraint conditions and the channel health score, perform channel combination screening to generate a multi-mode channel combination set that adapts to the requirements of delay-sensitive services and bandwidth-intensive services; use the multi-objective reward function to traverse the multi-mode channel combination set, adjust the exploration rate according to the exploration rate dynamic attenuation mechanism, and perform dynamic combination strategy optimization to obtain the optimal multi-mode channel combination.

[0079] Further, the verification feedback module 15 is further configured to perform the following steps: Based on the lightweight blockchain node, record the transmission data status information of each channel in real time; according to the transmission data status information, verify the actual transmission performance of each channel through the consensus mechanism of the blockchain, generate a transmission performance deviation, and perform channel compensation update based on the transmission performance deviation.

[0080] It should be noted that the above sequence of embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. In addition, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0081] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0082] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A 5G message multi-channel redundant transmission method based on an AI self-healing mechanism, characterized in that, The method includes: Deploying a distributed probe array between the terminal device and the base station, and collecting multi-channel transmission parameters in real time based on the distributed probe array; Inputting the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality decay trend of each channel within a preset time window, and generating a channel health score; Based on the channel health score, performing reinforcement learning through a self-healing decision engine to dynamically generate a multi-mode channel combination; According to the multi-mode channel combination, reconstructing the transmission path to generate an updated transmission path, and switching the transmission path according to the updated transmission path; Deploying lightweight blockchain nodes to verify the actual transmission performance of each channel in real time, and performing channel compensation update according to the actual transmission performance.

2. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 1, characterized in that The multi-channel transmission parameters include physical layer parameters, protocol layer parameters, and network layer parameters.

3. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 2, wherein Constructing the hybrid prediction model includes: Adopting a hybrid neural network architecture to parallelly extract the millimeter wave band fluctuation characteristics of the physical layer parameters and the topological correlation characteristics of the protocol layer parameters; Based on the millimeter wave band fluctuation characteristics and topological correlation characteristics, injecting adversarial samples simulating burst interference scenarios for model robustness training to generate the hybrid prediction model.

4. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 3, characterized in that, Setting a dual-threshold trigger mechanism. When the signal fluctuation variance exceeds the stability threshold or the delay jitter reaches the reliability threshold, activate the online incremental learning unit to update the hybrid prediction model.

5. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 4, characterized in that, Inputting the multi-channel transmission parameters into the hybrid prediction model to predict the transmission quality decay trend of each channel within a preset time window, and generating a channel health score, including: Performing frequency domain feature decomposition on the physical layer parameters, and extracting high-frequency fluctuation components as quality degradation warning indicators; Based on the high-frequency fluctuation components, performing time series analysis of the transmission quality to generate a quality decay trend; Dynamically allocating index weights for the quality degradation warning indicators; Based on the quality decay trend, combined with the index weights, calculating and generating a channel health score.

6. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 1, characterized in that, Based on the channel health score, performing reinforcement learning through a self-healing decision engine to dynamically generate a multi-mode channel combination, including: The self-healing decision engine includes a multi-objective reward function, an exploration rate dynamic decay mechanism, and heterogeneity constraint conditions; Based on the quality of service requirements of delay-sensitive services and bandwidth-intensive services, constructing a multi-objective reward function; Adopting a double deep reinforcement learning architecture and setting an exploration rate dynamic decay mechanism; Setting heterogeneity constraint conditions, and combining the multi-objective reward function and the exploration rate dynamic decay mechanism to optimize the channel combination strategy according to the channel health score to generate a multi-mode channel combination.

7. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 6, wherein, The heterogeneity constraint condition is that the channel combination includes at least three channels of different communication modes, and the proportion of non-terrestrial network channels does not exceed a preset upper limit.

8. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 7, wherein, Setting heterogeneity constraint conditions, and combining the multi-objective reward function and the exploration rate dynamic decay mechanism to optimize the channel combination strategy according to the channel health score to generate a multi-mode channel combination, including: Based on the heterogeneous constraints and the channel health score, perform channel combination screening to generate a multi-mode channel combination set that adapts to the requirements of time-sensitive services and bandwidth-intensive services; Utilize the multi-objective reward function to traverse the multi-mode channel combination set, adjust the exploration rate according to the exploration rate dynamic decay mechanism, and perform dynamic combination strategy optimization to obtain the optimal multi-mode channel combination.

9. The 5G message multi-channel redundant transmission method based on the AI self-healing mechanism according to claim 1, wherein, Deploy lightweight blockchain nodes to real-time verify the actual transmission performance of each channel, and perform channel compensation update according to the actual transmission performance, including: Based on the lightweight blockchain nodes, record the transmission data status information of each channel in real time; According to the transmission data status information, verify the actual transmission performance of each channel through the consensus mechanism of the blockchain, generate a transmission performance deviation, and perform channel compensation update based on the transmission performance deviation.

10. A 5G message multi-channel redundant transmission system based on an AI self-healing mechanism, characterized in that, The system includes: A dynamic perception module, which is used to deploy a distributed probe array between the terminal device and the base station, and collect multi-channel transmission parameters in real time based on the distributed probe array; An intelligent prediction module, which is used to input the multi-channel transmission parameters into a hybrid prediction model to predict the transmission quality decay trend of each channel within a preset time window, and generate a channel health score; A strategy generation module, which is used to perform reinforcement learning through a self-healing decision engine based on the channel health score to dynamically generate a multi-mode channel combination; A self-healing execution module, which is used to reconstruct the transmission path according to the multi-mode channel combination, generate an updated transmission path, and perform transmission path switching according to the updated transmission path; A verification feedback module, which is used to deploy lightweight blockchain nodes to real-time verify the actual transmission performance of each channel, and perform channel compensation update according to the actual transmission performance.

Images