High-speed mobile terminal routing addressing method and communication device

Abstract

The invention relates to the technical field of data processing, in particular to a high-speed mobile terminal routing addressing method and a communication device, and the method comprises the steps: collecting three-dimensional space coordinates, channel state indication information and adjacent node connection states, and constructing a network state data set with a timestamp; on the basis of node positions and connection states, a routing algorithm is combined with a topology prediction model to pre-judge a motion track, an optimal path containing a time effectiveness factor is calculated, and convergence delay is reduced through an increment updating mechanism; based on the path quality and the channel state, path stability and channel conditions are analyzed through a deep neural network, and a coding scheme is dynamically selected from a forward error correction coding library in combination with parallel simulation verification; evaluating energy efficiency of a resource allocation algorithm and a task scheduling algorithm in a digital twin simulation environment according to routing and coding requirements, and selecting a resource allocation strategy; through cross-layer optimization of an interface collaborative application path, a coding scheme and a resource strategy, a transmission index is monitored in real time and parameters are dynamically adjusted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a high-speed mobile terminal routing addressing method and communication device. BACKGROUND

[0002] In the wind power generation scenario, the high-speed mobile terminal such as unmanned aerial vehicle or automatic inspection equipment causes the network topology to change dynamically due to frequent position changes, and the existing routing protocol based on static assumption is difficult to effectively handle; therefore, the mobile ad hoc network architecture is introduced, the topology change is dynamically adapted through the distributed neighbor discovery and periodic routing update mechanism, the optimal forwarding path is calculated, and the multi-hop reliable transmission of data packets is realized, so as to support the real-time data acquisition and remote control demand of the wind power field.

[0003] In the actual application process of the existing mobile ad hoc network routing and error control integration technology, the following technical pain points exist, specifically, the network topology changes dynamically frequently under the wind power generation scenario due to the high-speed movement of the inspection unmanned aerial vehicle or automatic equipment, the routing protocol needs to continuously update the path but the convergence is slow, causing the data transmission delay; the wireless channel presents the rapid time-varying characteristic due to the environmental interference, the signal quality fluctuation aggravates the bit error rate rise; the error control relies on the retransmission mechanism, but the high bit error rate forces frequent retransmission, reducing the transmission efficiency; the distributed algorithm runs on the mobile terminal with limited computing resources, the topology update calculation load is large, and the processing capacity is insufficient, causing the difficulty in energy efficiency optimization, for example, when the unmanned aerial vehicle flies at high speed in the wind power field, the position changes fast, the routing table information expires, causing the data packet loss, the repeated transmission consumes the additional bandwidth and energy, and the real-time monitoring data reliability is affected. SUMMARY

[0004] In view of the technical problems of the prior art, the present application provides a high-speed mobile terminal routing addressing method and communication device, which solves the technical problems of frequent network topology dynamic changes, wireless channel rapid time variation and terminal computing resource limitation caused by high-speed movement of the terminal in the wind power generation scenario, resulting in routing convergence delay, bit error rate rise, retransmission mechanism efficiency reduction and system energy efficiency decline.

[0005] To solve the above technical problems, the present application is as follows:

[0006] In a first aspect, the present application provides a high-speed mobile terminal routing addressing method, comprising:

[0007] Step 1, collecting three-dimensional space coordinates, channel state indication information and adjacent node connection state, and generating a network state data set;

[0008] Step 2, based on the node position and connection state in the network state data set, using a routing algorithm combined with a topology prediction model to predict the node motion trajectory, calculating the optimal forwarding path, and using an incremental update mechanism;

[0009] Step 3, based on the quality indicators of the optimal forwarding path and the channel state information in the network state data set, select an encoding scheme from the library of forward error correction encoding schemes, analyze path stability and channel conditions through a deep neural network model, dynamically select an encoding scheme, and verify the selected encoding scheme through parallel simulation;

[0010] Step 4, according to the routing calculation requirements and the encoding processing requirements, adopt resource allocation algorithms and task scheduling algorithms to evaluate the energy efficiency performance in the simulation environment, and select a resource allocation strategy to allocate computing resources;

[0011] Step 5, the optimal forwarding path, the selected encoding scheme and the resource allocation strategy are applied cooperatively, and the path quality information and the allocated computing resources are shared through a cross-layer optimization interface for multi-hop data transmission.

[0012] Further, the method for routing addressing of the high-speed mobile terminal, the three-dimensional space coordinates, channel state indication information and adjacent node connection state are collected to generate a network state data set, comprising:

[0013] The three-dimensional space coordinate data is obtained through the positioning module;

[0014] The channel state indication information is measured through the wireless transceiver;

[0015] The adjacent node connection state is recorded by scanning through the neighbor discovery protocol;

[0016] The three-dimensional space coordinate data, the channel state indication information and the adjacent node connection state are integrated;

[0017] Timestamps are attached to all collected data, and spatiotemporal correlation analysis is performed to generate a network state data set.

[0018] Further, the method for routing addressing of the high-speed mobile terminal, based on the node position and connection state in the network state data set, adopts a routing algorithm combined with a topology prediction model to predict node motion trajectory, calculates an optimal forwarding path, and adopts an incremental update mechanism, comprising:

[0019] The network state data set is processed by a link state routing algorithm;

[0020] The historical position data in the network state data set is input into a topology prediction model to predict the future position change trend of the node;

[0021] Based on the prediction result and the current network state, a time effectiveness factor is introduced when calculating the path, and a dynamic validity period identifier is assigned to each path;

[0022] When a topology change is detected, only a local route update is triggered according to a change range.

[0023] Further, the high-speed mobile terminal routing addressing method provided by the application selects an encoding scheme from a forward error correction encoding scheme library based on the quality index of the optimal forwarding path and channel state information in the network state data set, analyzes path stability and channel conditions through a deep neural network model, dynamically selects an encoding scheme, and verifies the selected encoding scheme through parallel simulation, including:

[0024] The forward error correction encoding scheme library includes convolutional codes, Turbo codes and LDPC codes.

[0025] Real-time channel state parameters and data service characteristics are input into a deep neural network model, and an encoding scheme selection decision is output.

[0026] A parallel simulation module is started, and transmission error rate and throughput of multiple encoding schemes are simulated based on current network state.

[0027] The neural network decision result and the simulation result are compared, and an encoding scheme is selected in combination with historical transmission records.

[0028] Further, the high-speed mobile terminal routing addressing method provided by the application selects an encoding scheme from a forward error correction encoding scheme library based on the quality index of the optimal forwarding path and channel state information in the network state data set, analyzes path stability and channel conditions through a deep neural network model, dynamically selects an encoding scheme, and verifies the selected encoding scheme through parallel simulation, including:

[0029] A reinforcement learning resource allocation algorithm and an improved genetic task scheduling algorithm are used.

[0030] A digital twin simulation environment is constructed, and actual hardware resource state data is loaded.

[0031] The two algorithms are run in parallel in the simulation environment, and CPU utilization, memory occupation and energy consumption indicators are collected.

[0032] The collected indicators are analyzed through a multi-objective optimization evaluation function, and a resource allocation scheme is selected.

[0033] According to the selected scheme, the computing tasks are dynamically allocated to different processing cores.

[0034] Further, the high-speed mobile terminal routing addressing method provided by the application selects an encoding scheme from a forward error correction encoding scheme library based on the quality index of the optimal forwarding path and channel state information in the network state data set, analyzes path stability and channel conditions through a deep neural network model, dynamically selects an encoding scheme, and verifies the selected encoding scheme through parallel simulation, including:

[0035] The route path quality information is transmitted to the encoding processing unit through a cross-layer optimization interface.

[0036] The resource allocation strategy is used to allocate computing resources for route calculation and encoding processing.

[0037] In the data transmission process, the end-to-end delay and packet loss rate indicators are collected;

[0038] The routing parameters, encoding parameters and resource allocation parameters are adjusted according to the collected delay and packet loss rate indicators.

[0039] Further, the method for routing addressing of the high-speed mobile terminal provided by the application records the connection state of the adjacent node by scanning through the neighbor discovery protocol, and comprises the following steps:

[0040] The adjacent node is periodically scanned by using an optimized link state routing protocol;

[0041] The connection state and the link quality of the adjacent node are recorded;

[0042] The recorded connection state and link quality data are stored in a network state data set.

[0043] Further, the method for routing addressing of the high-speed mobile terminal provided by the application uses historical position data in the network state data set as input to predict the future position change trend of the node by using a topology prediction model, and comprises the following steps:

[0044] The historical position data is extracted from the network state data set;

[0045] The extracted historical position data is input to a prediction model based on a filtering algorithm;

[0046] The historical motion trajectory data is analyzed by the prediction model, and a future position prediction result of the node is output.

[0047] Further, the method for routing addressing of the high-speed mobile terminal provided by the application inputs real-time channel state parameters and data service characteristics into a deep neural network model, and outputs an encoding scheme selection decision, and comprises the following steps:

[0048] The signal-to-noise ratio, Doppler shift and data priority parameters are extracted from the channel state information and data service characteristics;

[0049] The extracted parameters are input to the input layer of the deep neural network model;

[0050] The input parameters are processed by the deep neural network model, and an encoding scheme selection decision is generated in the output layer.

[0051] In the second aspect, the application provides a communication device for routing addressing of a high-speed mobile terminal, which is applied to the method for routing addressing of the high-speed mobile terminal, and comprises the following steps:

[0052] A data collection module is configured to collect three-dimensional space coordinates, channel state indication information and adjacent node connection states, and generate a network state data set;

[0053] A path calculation module is configured to, based on the node positions and connection states in the network state dataset, predict node motion trajectories by using a routing algorithm in combination with a topology prediction model, calculate an optimal forwarding path, and use an incremental updating mechanism;

[0054] An encoding selection module is configured to, based on the quality indicators of the optimal forwarding path and the channel state information in the network state dataset, select an encoding scheme from a forward error correction encoding scheme library, analyze path stability and channel conditions by using a deep neural network model, dynamically select an encoding scheme, and verify the selected encoding scheme by using parallel simulation.

[0055] A resource allocation module is configured to, according to routing calculation requirements and encoding processing requirements, use a resource allocation algorithm and a task scheduling algorithm to evaluate energy efficiency performance in a simulation environment, and select a resource allocation strategy to allocate computing resources.

[0056] A strategy execution module is configured to cooperatively apply the optimal forwarding path, the selected encoding scheme and the resource allocation strategy, share path quality information and allocated computing resources through a cross-layer optimization interface, and is used for multi-hop data transmission.

[0057] The present application has the following advantages:

[0058] The present application periodically collects three-dimensional space coordinates, channel state indication information and adjacent node connection states to construct a network state dataset with timestamps, thereby providing a space-time feature basis for topology prediction. A link state routing algorithm is combined with a Kalman filter prediction model to predict node motion trajectories. A time validity factor is embedded in path calculation and local routing update is triggered, thereby effectively reducing routing convergence delay. A deep neural network model is used to analyze real-time channel state and path stability. An optimal encoding scheme is dynamically selected by combining parallel simulation verification and historical transmission records, thereby significantly reducing bit error rate and retransmission frequency. A digital twin simulation environment is used to run a reinforcement learning resource allocation algorithm and an improved genetic scheduling algorithm in parallel. According to multi-objective optimization evaluation results, computing tasks are allocated to achieve balanced optimization of computing resources and energy consumption. Finally, a cross-layer optimization interface is used to cooperatively apply a routing path, an encoding scheme and a resource allocation strategy. System parameters are dynamically adjusted according to end-to-end delay and packet loss rate indicators, thereby systematically solving the problems of data transmission delay, high bit error rate and low energy efficiency caused by dynamic topology changes, fast channel time variation and resource limitations in a wind power generation scenario, and improving the reliability and real-time performance of multi-hop data transmission. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed in the embodiments. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on the drawings.

[0060] Figure 1 A flowchart of a method for routing addressing of a high-speed mobile terminal is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the technical solutions of the present application clearer, the present application will be described in detail below with reference to specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application. The present application provided by each embodiment of the present application will be described in detail below with reference to the drawings. In order to better understand the purpose of the present application, the present application will be further described in detail below.

[0062] In a first aspect, referring to Figure 1 The present application provides a method for routing addressing of a high-speed mobile terminal, comprising:

[0063] Step 1, collecting three-dimensional space coordinates, channel state information and adjacent node connection state, generating network state data set;

[0064] Step 2, based on the node position and connection state in the network state data set, using routing algorithm combined with topological prediction model to predict node motion trajectory, calculating the optimal forwarding path, and using incremental update mechanism;

[0065] Step 3, based on the quality index of the optimal forwarding path and the channel state information in the network state data set, selecting an encoding scheme from a forward error correction encoding scheme library, analyzing the path stability and channel conditions through a deep neural network model, dynamically selecting an encoding scheme, and verifying the selected encoding scheme through parallel simulation;

[0066] Step 4, according to the routing calculation requirement and the encoding processing requirement, using resource allocation algorithm and task scheduling algorithm to evaluate the energy efficiency performance in the simulation environment, and selecting resource allocation strategy to allocate calculation resources;

[0067] Step 5, the optimal forwarding path, the selected encoding scheme and the resource allocation strategy are applied cooperatively, through cross-layer optimization interface to share path quality information and allocate calculation resources, for multi-hop data transmission.

[0068] The mobile terminal first obtains real-time three-dimensional spatial coordinate data through an integrated high-precision positioning module, and simultaneously measures a current channel state indicator parameter using a wireless transceiver. The connection state and link quality of neighboring nodes are periodically scanned and recorded by an optimized neighbor discovery protocol. The three-dimensional spatial coordinate data, channel state indicator information, and neighboring node connection state are integrated, and all collected data are marked with accurate time stamps. Time-space correlation analysis is performed to generate a network state data set with time series characteristics.

[0069] Based on the node position vector and connection state table in the network state data set, the mobile terminal processes the data set using a link state routing algorithm. Historical position data in the data set is input into a topology prediction model based on a filtering algorithm to predict future position changes of the nodes. According to the prediction results and the current network topology state, a time validity factor is introduced in the path calculation process to assign a dynamic validity period to each optimal forwarding path. When a network topology change is detected, the system triggers local routing updates according to the change impact range, reduces the full-network flooding broadcast through an incremental update mechanism, and effectively reduces the routing convergence delay.

[0070] For forward error correction coding selection, the mobile terminal constructs a coding scheme library containing convolutional codes, Turbo codes, and LDPC codes. Real-time channel state parameters are extracted from the network state data set, and the signal-to-noise ratio, Doppler shift, and data priority parameters are input into a deep neural network model in combination with data service type characteristics. The neural network model analyzes the path stability and channel conditions and outputs a coding scheme decision. A parallel simulation module is also started to simulate the transmission error rate and throughput of multiple coding schemes under the current network state. By comparing the neural network decision results with the parallel simulation results and integrating historical transmission records, the final coding scheme with the best overall performance is selected.

[0071] To optimize the allocation of computing resources, the mobile terminal runs a reinforcement learning resource allocation algorithm and an improved genetic task scheduling algorithm in parallel. A digital twin simulation environment is constructed to load actual hardware resource state data, and the two algorithms are executed in parallel in the simulation environment. Real-time CPU utilization, memory occupancy, and energy consumption indicators are collected, and the collected indicator data are analyzed by a multi-objective optimization evaluation function to select the resource allocation scheme with the highest computing efficiency and optimal energy consumption. According to the selected scheme, the topology update tasks are dynamically allocated to different processing cores to achieve efficient deployment of computing resources.

[0072] Finally, the optimized routing path, selected encoding scheme and resource allocation strategy are coordinated through the cross-layer optimization interface. The routing module transmits the path quality information to the encoding processing unit, and the resource allocation strategy provides calculation resource guarantee for routing calculation and encoding processing. The end-to-end delay and packet loss rate indicators are continuously collected during data transmission, and the routing parameters, encoding parameters and resource allocation parameters are dynamically adjusted according to the real-time monitoring indicators. Through the closed-loop control of the above technical chain, high-reliability low-latency multi-hop data transmission of high-speed mobile terminals in wind power scenarios is realized.

[0073] Specifically, the method for routing and addressing of high-speed mobile terminals provided by the application comprises the following steps:

[0074] acquiring three-dimensional spatial coordinate data through a positioning module;

[0075] measuring channel state indication information through a wireless transceiver;

[0076] scanning and recording the connection state of adjacent nodes through a neighbor discovery protocol;

[0077] integrating the three-dimensional spatial coordinate data, the channel state indication information and the connection state of adjacent nodes;

[0078] appending a time stamp to all collected data and performing spatio-temporal correlation analysis to generate a network state data set.

[0079] The mobile terminal acquires accurate three-dimensional spatial coordinate data through an integrated satellite positioning module, and measures channel quality indication parameters using a wireless transceiver hardware. An optimized neighbor discovery protocol is used to periodically scan adjacent nodes within the communication range, and the connection state and real-time link quality values of each node are recorded. The collected three-dimensional spatial coordinate data, channel state indication information and adjacent node connection state data are integrated and processed, and high-precision time stamps are appended to all data. A spatio-temporal correlation analysis algorithm is executed to correlate and calculate the network state data at different time points, and a network state data set with spatio-temporal continuity is generated.

[0080] Specifically, the method for routing and addressing of high-speed mobile terminals provided by the application comprises the following steps:

[0081] processing the network state data set using a link state routing algorithm;

[0082] inputting historical position data in the network state data set into a topology prediction model to predict the future position change trend of the node;

[0083] Based on the prediction results and the current network state, a time validity factor is introduced in the path calculation, and a dynamic validity period identifier is given to each path;

[0084] When a topology change is detected, only local route update is triggered according to the change range.

[0085] The link state routing algorithm processes the node position vector and the connection relationship table in the network state data set. The historical time sequence position data is extracted from the network state data set and input into the topology prediction model based on Kalman filtering, and the model outputs the probability distribution of the future position of the node. The routing algorithm combines the position prediction results and the current topology snapshot, introduces a time validity factor in the path calculation, and adds a dynamic validity period identifier to each forwarding path generated. When the node displacement exceeds the threshold or the link is interrupted, the system triggers local route table update according to the change influence range, and only broadcasts update information to the affected area.

[0086] Specifically, the method for high-speed mobile terminal routing and addressing provided by the application selects an encoding scheme from a forward error correction encoding scheme library based on the quality index of the optimal forwarding path and the channel state information in the network state data set, analyzes the path stability and the channel condition through a deep neural network model, dynamically selects the encoding scheme, and verifies the selected encoding scheme through parallel simulation, including:

[0087] The forward error correction encoding scheme library includes convolutional codes, Turbo codes and LDPC codes;

[0088] The real-time channel state parameters and the data service characteristics are input into the deep neural network model, and the encoding scheme selection decision is output;

[0089] The parallel simulation module is started, and the transmission error rate and the throughput of multiple encoding schemes are simulated based on the current network state;

[0090] The neural network decision result and the simulation result are compared, and the encoding scheme is selected in combination with the historical transmission record.

[0091] The forward error correction encoding scheme library stores the encoding rules and parameter configurations of convolutional codes, Turbo codes and LDPC codes. The deep neural network model receives the real-time signal-to-noise ratio, Doppler shift characteristics and data service priority parameters from the network state data set as input layer data. The neural network model analyzes the path stability index and the channel attenuation characteristics, and generates an encoding scheme selection decision vector in the output layer. The parallel simulation module loads the current network state, performs Monte Carlo transmission simulation on the three encoding schemes, and outputs the error rate and throughput prediction data. The system compares the neural network decision vector and the simulation prediction data, and finally selects the encoding scheme with the highest comprehensive score in combination with the actual performance data in the historical transmission record.

[0092] Specifically, the method for routing addressing of high-speed mobile terminals according to the application adopts a resource allocation algorithm and a task scheduling algorithm according to the requirements of route calculation and encoding processing, evaluates the energy efficiency performance in a simulation environment, and selects a resource allocation strategy to allocate computing resources, including:

[0093] The reinforcement learning resource allocation algorithm and the improved genetic task scheduling algorithm are adopted.

[0094] A digital twin simulation environment is constructed, and actual hardware resource state data is loaded.

[0095] The two algorithms are run in parallel in the simulation environment, and CPU utilization, memory occupation and energy consumption indicators are collected.

[0096] The collected indicators are analyzed by a multi-objective optimization evaluation function, and a resource allocation scheme is selected.

[0097] According to the selected scheme, the computing tasks are dynamically allocated to different processing cores.

[0098] The reinforcement learning resource allocation algorithm and the improved genetic task scheduling algorithm are executed in parallel in the digital twin simulation environment. The digital twin system loads actual hardware parameters such as the number of processor cores, memory capacity and energy consumption constraints. The simulation environment collects real-time CPU utilization, memory occupation and energy consumption rate indicators during algorithm execution. The multi-objective optimization evaluation function analyzes the collected indicator dataset, and calculates the weighted scores of the two algorithms in terms of computing delay and energy consumption. The system selects the scheme with the optimal weighted score, and dynamically allocates the topology update task to a specific processing core according to the core allocation strategy in the scheme.

[0099] Specifically, the method for routing addressing of high-speed mobile terminals according to the application adopts the selected encoding scheme and resource allocation strategy in cooperation, including:

[0100] The route path quality information is transmitted to the encoding processing unit through a cross-layer optimization interface.

[0101] The resource allocation strategy is used to allocate computing resources for route calculation and encoding processing.

[0102] During data transmission, the end-to-end delay and packet loss rate indicators are collected.

[0103] According to the collected delay and packet loss rate indicators, the routing parameters, encoding parameters and resource allocation parameters are adjusted.

[0104] The cross-layer optimization interface transmits the optimal path stability index generated by the routing module to the encoding processing unit. The resource allocation strategy allocates differentiated processor time slices and memory resources to the routing calculation thread and the encoding processing thread. During data transmission, the system continuously collects end-to-end transmission latency and packet loss rate measurement data. Based on the changing trends of latency and packet loss rate, the system dynamically adjusts the probing frequency of the routing algorithm, the redundancy parameters of the encoding scheme, and the resource allocation weights of the task scheduling.

[0105] Specifically, the high-speed mobile terminal routing and addressing method of the present invention includes, in the step of scanning and recording the connection status of adjacent nodes through a neighbor discovery protocol, the following steps:

[0106] An optimized link-state routing protocol is used to periodically scan neighboring nodes;

[0107] Record the connection status and link quality of adjacent nodes;

[0108] The recorded connection status and link quality data are stored in the network status dataset.

[0109] The optimized link-state routing protocol broadcasts probe messages at preset intervals to scan neighboring nodes within the communication radius. The system records the response status, signal strength, and link quality indicator values ​​of each neighboring node. The recorded connection state table and link quality matrix are stored in the neighbor relationship storage area of ​​the network state dataset.

[0110] Specifically, the high-speed mobile terminal routing and addressing method of the present invention, wherein the step of using historical location data from the network state dataset to input the topology prediction model to predict the future location change trend of nodes includes:

[0111] Extract historical location data from the network status dataset;

[0112] The extracted historical location data is input into a prediction model based on a filtering algorithm;

[0113] By analyzing historical motion trajectory data using a predictive model, the system outputs predictions of the future locations of nodes.

[0114] Temporal location coordinate sequences are extracted from the location history storage of the network state dataset. The extracted historical location data is then input into a node motion prediction model based on the Kalman filter algorithm. The prediction model analyzes the motion trajectory characteristics of the location sequences and outputs prediction results containing future location coordinates and confidence levels.

[0115] Specifically, the high-speed mobile terminal routing and addressing method of the present invention, wherein the step of inputting real-time channel state parameters and data service characteristics into a deep neural network model and outputting a coding scheme selection decision includes:

[0116] Extract signal-to-noise ratio, Doppler shift, and data priority parameters from channel state information and data service characteristics;

[0117] The extracted parameters are input into the input layer of the deep neural network model;

[0118] The input parameters are processed by a deep neural network model, and the output layer generates a decision on the coding scheme selection.

[0119] Signal-to-noise ratio (SNR) features and Doppler shift are extracted from real-time channel state information, and service priority parameters are parsed from the data packet header. The extracted SNR, Doppler shift, and priority parameters are then input into the input layer of a deep neural network model. The deep neural network model processes the input features through multiple convolutional layers, generating a selection decision vector containing the coding type and parameter configuration at the output layer.

[0120] Secondly, the present invention provides a communication device for high-speed mobile terminal routing addressing, applied to the aforementioned method for high-speed mobile terminal routing addressing, comprising:

[0121] The data acquisition module is used to collect three-dimensional spatial coordinates, channel status indication information, and connection status of adjacent nodes to generate a network status dataset.

[0122] The path calculation module is used to predict the movement trajectory of nodes based on the node locations and connection status in the network status dataset, using a routing algorithm combined with a topology prediction model, to calculate the optimal forwarding path, and employs an incremental update mechanism.

[0123] The coding selection module is used to select a coding scheme from the forward error correction coding scheme library based on the quality index of the optimal forwarding path and the channel state information in the network state dataset. It analyzes the path stability and channel conditions through a deep neural network model, dynamically selects the coding scheme, and verifies the selected coding scheme through parallel simulation.

[0124] The resource allocation module is used to evaluate energy efficiency performance in a simulation environment and select a resource allocation strategy to allocate computing resources based on routing calculation requirements and encoding processing requirements, using resource allocation algorithms and task scheduling algorithms.

[0125] The strategy execution module is used to coordinate the application of the optimal forwarding path, the selected encoding scheme and the resource allocation strategy, and to share path quality information and allocate computing resources through the cross-layer optimization interface for multi-hop data transmission.

[0126] The data acquisition module integrates a satellite positioning unit, a radio frequency (RF) unit, and a neighbor discovery protocol processor. The satellite positioning unit acquires the mobile terminal's three-dimensional spatial coordinate data in real time, while the RF unit collects channel status indication parameters through a signal strength measurement circuit. The neighbor discovery protocol processor periodically sends probe frames to scan neighboring nodes, recording node connection status and link quality indication values. The data integration unit appends millisecond-level timestamps to these three types of data and generates a network state dataset containing a node location matrix, a channel quality vector, and a neighbor relationship table using a spatiotemporal correlation algorithm.

[0127] The path calculation module comprises a route processor and a topology prediction engine. The route processor uses a link-state routing algorithm to parse the node location matrix and neighbor relationship table from the network state dataset. The topology prediction engine loads historical location sequences from the dataset and outputs predicted node trajectories using a Kalman filter model. The path generation unit combines the predicted trajectories with the current topology state, embedding a time validity factor into the path calculation to add a dynamic validity period identifier to the optimal forwarding path. The incremental update controller triggers a local routing table update based on the node displacement difference when a topology change is detected.

[0128] The coding selection module deploys a forward error correction coding library, a deep neural network processor, and a parallel simulator. The coding library stores coding rules and parameter configuration sets for convolutional codes, Turbo codes, and LDPC codes. The input layer of the deep neural network processor receives the path stability index output by the path calculation module, the signal-to-noise ratio / Doppler shift parameters from the network state dataset, and data service priority labels. The neural network analyzes path stability and channel conditions through convolutional layers and generates coding scheme decision vectors at the output layer. The parallel simulator performs transmission simulations on the three coding schemes based on the current network state, outputting predicted bit error rate and throughput values. The decision arbitration unit compares the neural network decision with the simulation results and selects the final coding scheme based on historical transmission records.

[0129] The resource allocation module includes a reinforcement learning decision-maker, a genetic algorithm scheduler, and a digital twin environment. The digital twin environment mirrors hardware parameters such as the number of processor cores, memory capacity, and energy consumption constraints. The reinforcement learning decision-maker generates resource allocation strategies, and the genetic algorithm scheduler outputs task scheduling sequences. The simulation executor runs both algorithms in parallel within the digital twin environment, collecting real-time measurements of CPU utilization, memory usage, and power consumption. The evaluation unit analyzes the collected data using a multi-objective optimization function, outputting a weighted score for computational efficiency and energy consumption. The strategy selector selects a resource allocation scheme based on the score results, and the task distribution unit assigns routing computation tasks and encoding processing tasks to designated processing cores according to the scheme.

[0130] The strategy execution module includes a cross-layer optimization interface, a resource allocation executor, and a parameter adjustment unit. The cross-layer optimization interface transmits the path quality metrics generated by the path calculation module to the encoding processing unit, while simultaneously distributing the resource allocation strategy to each calculation unit. The resource allocation executor allocates processor time slices and memory blocks to the routing calculation thread and configures dedicated hardware acceleration resources for the encoding processing thread. The data transmission monitoring unit collects end-to-end latency data and packet loss rate measurements in real time. The parameter adjustment unit dynamically modifies the routing algorithm's probe frequency parameters, the encoding scheme's redundancy coefficient, and the task scheduler's resource allocation weights based on the monitoring data.

[0131] This invention utilizes a mobile terminal to periodically collect three-dimensional spatial coordinates, channel state indication parameters, and the connection status of adjacent nodes to construct a timestamped network state dataset. This dataset comprehensively records the dynamic characteristics of topology changes, providing a real-time data foundation for subsequent processing. Based on the node location vectors and connection status tables in the dataset, a link-state routing algorithm combined with a Kalman filter prediction model is used to predict node movement trajectories and calculate the optimal forwarding path. A time validity factor is introduced during path calculation, attaching a dynamic validity period identifier to each path. When a topology change is detected, a local route update is triggered based on the displacement difference, significantly reducing the number of network-wide flooding broadcasts and lowering routing convergence delay.

[0132] To address the rapidly changing time-varying characteristics of wireless channels, a coding scheme is dynamically selected from a forward error correction coding scheme library. A deep neural network model analyzes real-time signal-to-noise ratio, Doppler shift characteristics, and data service priorities, outputting a coding decision vector. A parallel simulation module performs transmission simulations on convolutional codes, Turbo codes, and LDPC codes based on the current network state, generating bit error rate prediction data. The system integrates neural network decisions, simulation results, and historical transmission records to select the optimal coding scheme, thereby reducing the bit error rate. This mechanism reduces the probability of retransmission triggering, improving the efficiency of the retransmission mechanism.

[0133] To address the issue of limited computing resources, a reinforcement learning resource allocation algorithm and an improved genetic task scheduling algorithm were run in parallel in a digital twin simulation environment. The simulation environment was configured with hardware parameters such as the number of processor cores and memory capacity, and real-time data was collected on CPU utilization, memory usage, and energy consumption rates during algorithm execution. A multi-objective optimization evaluation function was used to analyze the data, selecting the scheme with the lowest computational latency and optimal energy consumption. Topology update tasks were dynamically allocated to designated processing cores according to a core allocation strategy, achieving a computing resource utilization rate of over 92%.

[0134] The strategy execution module transmits routing path quality information to the encoding processing unit through a cross-layer optimization interface, and simultaneously allocates differentiated resources to each computing thread according to the resource allocation strategy. During data transmission, end-to-end latency and packet loss rate metrics are continuously collected, and the route detection frequency, encoding redundancy coefficient, and task scheduling weights are dynamically adjusted based on the monitoring data. This closed-loop control system stabilizes multi-hop data transmission latency within 50ms, improves system energy efficiency, and systematically solves the transmission bottleneck problem of high-speed mobile terminals in wind farm scenarios.

[0135] In wind power plant applications, mobile terminals are mounted on inspection drones or automated equipment to collect high-precision three-dimensional spatial coordinate data via satellite positioning modules at a sampling frequency of 10Hz. Wireless transceivers measure received signal strength and channel quality indicators in real time, while simultaneously running an optimized link-state routing protocol and periodically broadcasting probe frames to scan neighboring nodes. The system records the response latency, link stability, and signal attenuation characteristics of each node. This data, combined with positioning information and appended with millisecond-level timestamps, is then used to generate a network state dataset containing a node location matrix, channel quality vector, and neighbor relationship table through spatiotemporal correlation analysis.

[0136] The path calculation module loads node location vectors and connectivity tables from the network state dataset and uses an improved Dijkstra algorithm to calculate the initial path. The topology prediction engine extracts historical location sequences from the dataset and predicts the probability distribution of future node locations using a Kalman filter model. The routing processor combines the predicted trajectory with the current topology snapshot, embedding a time validity factor into the path calculation and attaching a dynamic validity period identifier to each forwarding path. When a node displacement exceeds a preset threshold or the link quality falls below a critical value, the incremental update controller only broadcasts route update instructions to the affected area, avoiding network-wide flooding.

[0137] The coding selection module retrieves configuration parameters for convolutional codes, Turbo codes, and LDPC codes from the forward error correction coding library. The deep neural network processor receives the path hop count and stability coefficient output by the path calculation module, as well as the real-time signal-to-noise ratio, Doppler shift, and data service priority tags from the dataset. The neural network analyzes the time-varying characteristics of the channel and path reliability through three layers of convolution, outputting a coding decision vector. The parallel simulator loads the current network state model and performs Monte Carlo transmission simulations on the three coding schemes, generating bit error rate and throughput prediction data. The decision arbitration unit integrates the neural network output, simulation results, and actual performance data from historical transmission records to select the coding scheme with the lowest bit error rate.

[0138] The resource allocation module constructs a digital twin environment mirror of a quad-core processor architecture, loading processor clock speed, memory bandwidth, and power consumption constraints. A reinforcement learning decision-maker generates a core allocation strategy, and an improved genetic algorithm scheduler outputs the task execution sequence. The simulation environment runs both algorithms in parallel, collecting real-time data on CPU core utilization, peak memory usage, and power consumption. The evaluation unit uses a multi-objective optimization function to analyze a weighted score of latency and energy consumption, selecting the scheme with the optimal overall score. The task distribution unit allocates topology computation tasks to cores 1-2 and encoding processing tasks to cores 3-4 based on the selected scheme.

[0139] The strategy execution module transmits path quality information to the encoding processor through a cross-layer optimization interface. Simultaneously, it allocates 60% of the processor time slice to the path calculation thread according to the resource allocation strategy, reserving dedicated hardware acceleration resources for the encoding thread. During data transmission, the monitoring unit collects end-to-end latency and packet loss rate metrics. The parameter adjustment unit dynamically increases the route probing frequency based on latency trends, adjusts the encoding redundancy coefficient according to packet loss rate fluctuations, and reallocates the core load weights. This closed-loop control system achieves stable multi-hop data transmission, systematically solving the problems of route convergence delay, high bit error rate, and low energy efficiency in the dynamic topology environment of wind farms.

[0140] The topology prediction model employs the Kalman filter algorithm to process historical location sequences in the network state dataset. The Kalman filter model recursively predicts node trajectories using state and observation equations, establishes a motion state transition matrix based on temporal features from historical location data, corrects prediction errors using current measurements, and outputs a probability distribution of future location coordinates. This model effectively reduces positioning noise interference from high-speed mobile terminals in wind power generation scenarios and improves trajectory prediction accuracy.

[0141] The deep neural network model comprises an input layer, convolutional layers, and an output layer. The input layer receives path stability metrics, real-time signal-to-noise ratio, Doppler shift, and data service priority parameters. The convolutional layers analyze the nonlinear relationship between channel conditions and path reliability through multi-layer feature extraction. The output layer generates a decision vector containing the coding type and parameter configuration. The neural network model optimizes the weight parameters using a backpropagation algorithm, enabling adaptive selection of coding schemes in complex channel environments.

[0142] The digital twin simulation environment constructs a virtual image of hardware resources, loading physical parameters such as the number of processor cores, memory capacity, cache size, and power consumption thresholds. The simulation environment creates logical operation units that perfectly correspond to the real hardware, mapping real-time operating states such as CPU utilization, memory bandwidth usage, and power consumption rate. This environment supports the parallel execution of various resource scheduling algorithms, providing an accurate computing platform for energy efficiency assessment.

[0143] The Monte Carlo simulation module performs random sampling simulations during the forward error correction coding verification process. Based on the current network state parameters, the module establishes a transmission channel model and conducts transmission tests on convolutional codes, Turbo codes, and LDPC codes with hundreds of thousands of data packets. Each simulation randomly generates channel noise, interference attenuation, and multipath fading parameters, and statistically analyzes the bit error rate and throughput data of different coding schemes under dynamic channel conditions.

[0144] The multi-objective optimization evaluation function defines a weighted scoring rule for computational latency and energy consumption. The function takes into account real-time metrics such as CPU utilization, peak memory usage, and power consumption, and converts these multi-dimensional parameters into a single evaluation value through linear weighting. The evaluation function sets the weighting coefficient for resource allocation algorithm latency to 0.6 and the weighting coefficient for energy consumption to 0.4, and selects the Pareto optimal solution based on the scoring results.

[0145] The improved Dijkstra algorithm introduces a time validity factor as an additional constraint in path calculation. The algorithm traverses the node location vectors and connectivity tables in the network state dataset, considering not only hop count and bandwidth but also the path's survival probability within a future time window. A dynamic validity identifier is attached to each candidate path, automatically discarding paths when predicted trajectories indicate nodes will leave communication range, thus reducing invalid route broadcasts.

[0146] The Kalman filter algorithm recursively processes historical position sequences in the network state dataset. The algorithm predicts the next node position using state transition equations and updates the covariance matrix by combining real-time positioning measurements. The motion model uses node velocity and acceleration as state variables, outputting a probability distribution cloud map of future positions. The filter dynamically adjusts process noise parameters to adapt to the variable-speed motion characteristics of the drone in windy conditions.

[0147] The deep neural network algorithm employs a three-layer convolutional structure to process channel state parameters and path metrics. The input layer receives the signal-to-noise ratio, Doppler shift, path stability coefficient, and data priority label; the convolutional layers extract the nonlinear features of channel attenuation and path reliability using the ReLU activation function; and the fully connected layers output an encoded decision vector. The network weights are optimized using stochastic gradient descent, and the loss function integrates cross-entropy and mean squared error.

[0148] The Monte Carlo simulation algorithm constructs a transmission channel based on the Rayleigh fading model. The algorithm randomly generates channel noise figure, multipath attenuation factor, and interference intensity parameters that conform to the current network conditions. Tens of thousands of independent data packet transmission tests are performed on each coding scheme, and the frequency of bit error events and effective throughput are statistically analyzed. The simulation results generate a bit error rate distribution function and throughput confidence intervals.

[0149] The reinforcement learning resource allocation algorithm's state space includes processor core load rate, memory usage ratio, and real-time power consumption. The action space is defined as the mapping relationship between each computational task and different cores. The reward function combines the reduction in computational latency and the energy savings, and iteratively optimizes the resource allocation scheme through a Q-learning strategy. The exploration mechanism employs an ε-greedy strategy to balance development and exploration.

[0150] The improved genetic task scheduling algorithm employs a binary chromosome-encoded task allocation scheme. The fitness function evaluates the computational latency and energy consumption weighted average of the task scheduling sequence. The selection operation uses a roulette wheel algorithm to retain high-performing individuals, the crossover operation performs two-point crossover to exchange task blocks, and the mutation operation randomly flips the core allocation flag. An elite retention strategy ensures that the best individuals from each generation advance to the next.

Claims

1. A method for routing and addressing in a high-speed mobile terminal, characterized in that, include: Step 1: Collect three-dimensional spatial coordinates, channel status indication information, and connection status of adjacent nodes to generate a network status dataset; Step 2: Based on the node locations and connection states in the network state dataset, a routing algorithm combined with a topology prediction model is used to predict the node movement trajectory, calculate the optimal forwarding path, and adopt an incremental update mechanism. Step 3: Based on the quality index of the optimal forwarding path and the channel state information in the network state dataset, select a coding scheme from the forward error correction coding scheme library, analyze the path stability and channel conditions through a deep neural network model, dynamically select the coding scheme, and verify the selected coding scheme through parallel simulation. Step 4: Based on the routing calculation requirements and encoding processing requirements, resource allocation algorithms and task scheduling algorithms are used to evaluate energy efficiency performance in a simulation environment, and resource allocation strategies are selected to allocate computing resources. Step 5: The optimal forwarding path, the selected encoding scheme, and the resource allocation strategy are applied in a coordinated manner. Path quality information is shared and computing resources are allocated through the cross-layer optimization interface for multi-hop data transmission.

2. The high-speed mobile terminal routing and addressing method according to claim 1, characterized in that, The process of collecting three-dimensional spatial coordinates, channel state indication information, and connection status of adjacent nodes to generate a network state dataset includes: Three-dimensional spatial coordinate data is obtained through the positioning module; Channel status indication information is measured using a wireless transceiver; The neighbor discovery protocol is used to scan and record the connection status of adjacent nodes; The three-dimensional spatial coordinate data, the channel status indication information, and the connection status of adjacent nodes are integrated; All collected data are timestamped, and spatiotemporal correlation analysis is performed to generate a network state dataset.

3. The high-speed mobile terminal routing and addressing method according to claim 1, characterized in that, Based on the node locations and connection states in the network state dataset, a routing algorithm combined with a topology prediction model is used to predict node movement trajectories, calculate the optimal forwarding path, and employ an incremental update mechanism, including: The network state dataset is processed using a link-state routing algorithm. The historical location data from the network state dataset is used to input the topology prediction model to predict the future location change trend of nodes; Based on the prediction results and the current network status, a time validity factor is introduced when calculating the path, and each path is assigned a dynamic validity period identifier; When a topology change is detected, only a local route update is triggered based on the scope of the change.

4. The high-speed mobile terminal routing and addressing method according to claim 1, characterized in that, The method involves selecting a coding scheme from a forward error correction coding scheme library based on the quality index of the optimal forwarding path and channel state information in the network state dataset. This selection is achieved by analyzing path stability and channel conditions using a deep neural network model, dynamically choosing the coding scheme, and verifying the selected scheme through parallel simulation. This includes: The forward error correction coding scheme library includes convolutional codes, Turbo codes, and LDPC codes; The real-time channel state parameters and data service characteristics are input into a deep neural network model, and the output is a coding scheme selection decision. Start the parallel simulation module to simulate the transmission error rate and throughput of various coding schemes based on the current network status; The coding scheme is selected by comparing the neural network decision results with the simulation results and combining them with historical transmission records.

5. The method for routing and addressing high-speed mobile terminals according to claim 1, characterized in that, The process involves evaluating energy efficiency in a simulation environment using resource allocation and task scheduling algorithms based on routing and encoding processing requirements, and selecting a resource allocation strategy to allocate computing resources. This includes: A reinforcement learning resource allocation algorithm and an improved genetic task scheduling algorithm are adopted. Construct a digital twin simulation environment and load actual hardware resource status data; Two algorithms were run in parallel in a simulation environment, and CPU utilization, memory usage, and energy consumption metrics were collected. By analyzing the collected indicators through multi-objective optimization evaluation functions, a resource allocation scheme is selected. The computing tasks are dynamically allocated to different processing cores according to the selected scheme.

6. The method for routing and addressing high-speed mobile terminals according to claim 1, characterized in that, The selected encoding scheme and resource allocation strategy are applied in tandem, including: The routing path quality information is transmitted to the encoding processing unit through the cross-layer optimization interface; Use resource allocation strategies to allocate computing resources for routing computation and encoding processing; During data transmission, end-to-end latency and packet loss rate metrics are collected. Adjust routing parameters, encoding parameters, and resource allocation parameters based on the collected latency and packet loss rate metrics.

7. The high-speed mobile terminal routing and addressing method according to claim 2, characterized in that, The process of scanning and recording the connection status of adjacent nodes through the neighbor discovery protocol includes: An optimized link-state routing protocol is used to periodically scan neighboring nodes; Record the connection status and link quality of adjacent nodes; The recorded connection status and link quality data are stored in the network status dataset.

8. The high-speed mobile terminal routing and addressing method according to claim 3, characterized in that, The method of using historical location data from the network state dataset to input the topology prediction model to predict the future location change trend of nodes includes: Extract historical location data from the network status dataset; The extracted historical location data is input into a prediction model based on a filtering algorithm; By analyzing historical motion trajectory data using a predictive model, the system outputs predictions of the future locations of nodes.

9. The high-speed mobile terminal routing and addressing method according to claim 4, characterized in that, The process of inputting real-time channel state parameters and data service characteristics into a deep neural network model and outputting a coding scheme selection decision includes: Extract signal-to-noise ratio, Doppler shift, and data priority parameters from channel state information and data service characteristics; The extracted parameters are input into the input layer of the deep neural network model; The input parameters are processed by a deep neural network model, and the output layer generates a decision on the coding scheme selection.

10. A communication apparatus for high-speed mobile terminal routing, applied to a high-speed mobile terminal routing method as described in any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to collect three-dimensional spatial coordinates, channel status indication information, and connection status of adjacent nodes to generate a network status dataset. The path calculation module is used to predict the movement trajectory of nodes based on the node locations and connection status in the network status dataset, using a routing algorithm combined with a topology prediction model, to calculate the optimal forwarding path, and employs an incremental update mechanism. The coding selection module is used to select a coding scheme from the forward error correction coding scheme library based on the quality index of the optimal forwarding path and the channel state information in the network state dataset. It analyzes the path stability and channel conditions through a deep neural network model, dynamically selects the coding scheme, and verifies the selected coding scheme through parallel simulation. The resource allocation module is used to evaluate energy efficiency performance in a simulation environment and select a resource allocation strategy to allocate computing resources based on routing calculation requirements and encoding processing requirements, using resource allocation algorithms and task scheduling algorithms. The strategy execution module is used to coordinate the application of the optimal forwarding path, the selected encoding scheme and the resource allocation strategy, and to share path quality information and allocate computing resources through the cross-layer optimization interface for multi-hop data transmission.

Images