Vehicle ad hoc network based on intelligent network connection vehicle WiFi and V2V interaction method

Abstract

The invention relates to the technical field of wireless communication networks, in particular to a vehicle ad hoc network based on intelligent networked vehicle WiFi and a V2V interaction method, a vehicle motion state and network load data are acquired through a vehicle-mounted sensor, AP / STA modes are dynamically switched by combining a deep Q network (DQN) model and an epsilon-greedy strategy, and an adaptive topological structure is constructed. And expanding OLSR protocol message embedding path delay fields, calculating and optimizing multi-hop routing selection in combination with pheromone gradient, and realizing message priority classification and differential privacy protection through a federated learning model. And dynamically adjusting an AP / STA proportion threshold based on a particle swarm algorithm, and triggering node role reconfiguration to balance network load. According to the method, the existing WiFi hardware is reused, the deployment cost is reduced, cross-layer routing optimization and a distributed decision mechanism are combined, and the network real-time performance, the expansibility and the group cooperation capability in a dynamic scene are improved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication network technology, and in particular to a vehicle self-organizing network and a V2V interaction method based on WiFi of an intelligent networked vehicle. Background Art

[0002] The vehicle ad hoc network (VANET) of the WiFi of intelligent connected vehicles is a distributed communication network built on wireless local area network technology. It realizes real-time data interaction between vehicles (V2V) and vehicles and roads (V2I) through the on-board communication unit and the nodes such as adjacent vehicles and roadside infrastructure. This technology breaks through the dependence of traditional cellular networks on central base stations, expands the communication range by using multi-hop relay transmission mechanism, and adapts to the changes in network structure caused by high-speed movement of vehicles through dynamic topology perception and self-organizing routing protocols. It can support low-latency communication requirements in scenarios such as collision warning, formation coordination, and road condition sharing. At the same time, it improves the utilization rate of wireless resources through spectrum reuse. It is a key supporting technology for improving driving safety and efficiency in intelligent transportation systems.

[0003] The vehicle-to-vehicle (V2V) interaction of intelligent connected vehicles is a distributed collaborative mechanism based on wireless communication technology. It uses dedicated short-range communication (DSRC) or cellular vehicle-to-everything (C-V2X) protocols through on-board terminals to achieve real-time interactive status data, driving intentions, and environmental perception information between vehicles. This technology relies on a self-organizing network architecture, supports multi-hop relay transmission and dynamic routing optimization, and can adapt to network topology fluctuations caused by high-speed vehicle movement under non-central node scheduling. It completes information synchronization in scenarios such as collision warning, collaborative obstacle avoidance, and platoon driving through low-latency, high-reliability communication links. At the same time, it combines channel anti-interference and data encryption technology to improve transmission stability and security. It is a core communication solution for improving the active safety and group collaborative efficiency of intelligent transportation systems.

[0004] The existing V2X technical solutions have high deployment costs due to their reliance on infrastructure such as roadside units and communication base stations, and dedicated communication modules such as DSRC and LTE-V have high hardware modification costs and limited penetration rates. At the same time, traditional V2V communication has delay bottlenecks in multi-hop routing efficiency and dynamic adaptability of network topology. The present invention reuses the existing WiFi modules of intelligent connected vehicles to build a dynamic self-organizing network, uses the AP / STA / bridge mode switching mechanism to achieve autonomous networking of vehicle nodes, and combines multi-hop routing optimization and bridging interconnection technology. Without the need to add dedicated hardware and infrastructure, the deployment cost of vehicle collaborative communication is reduced, and the real-time nature of message propagation and network scalability in complex traffic scenarios are improved. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention provides a vehicle self-organizing network and V2V interaction method based on the WiFi of intelligent connected vehicles, which solves the problems of high deployment cost and poor scalability of vehicle cooperative communication caused by the existing V2X technology due to its reliance on dedicated hardware and infrastructure, and the low real-time performance caused by the insufficient dynamic topology adaptability of the traditional networking mode.

[0006] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows: The present invention provides a vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi, including: The vehicle motion state data and network state data are collected through the vehicle-mounted sensors, and the speed, position coordinates and network load rate are normalized to generate a multi-dimensional feature vector; Input the multi-dimensional feature vector into a pre-trained deep Q network model, output an operation instruction set including AP / STA mode switching probability, and select the optimal instruction according to the ε-greedy strategy to build a dynamic self-organizing network topology; Based on the node distribution information in the dynamic ad hoc network topology, a path delay field is embedded in the OLSR protocol message to generate a pheromone data packet, and the routing table entries of each node are updated through periodic broadcasting; Receive V2V messages from the vehicle application layer, input the message type and propagation hop count into the federated learning classification model, output the priority tag and embed it into the message header, and trigger the MAC layer channel preemption mechanism; The shared position data of neighboring vehicles is aligned with the local IMU sensor data in time and space, and the vehicle pose estimation is generated through the particle filter algorithm. The collaborative control instructions are calculated based on the Nash equilibrium strategy matrix and distributed to the target vehicle through the optimized multi-hop path. The AP node load rate and the number of channel conflicts are collected, and the AP / STA ratio threshold is iteratively calculated through the particle swarm optimization algorithm. The node role reconfiguration is triggered according to the calculation results, and the load status is fed back to the self-organizing network topology construction module.

[0007] Furthermore, the vehicle ad hoc network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the construction of dynamic ad hoc network topology includes: The vehicle speed value is discretized into three states: low speed, medium speed, and high speed, and the current AP node density and channel occupancy are associated to construct a three-dimensional state space; Input the three-dimensional state space into a pre-trained deep Q network model, output the value function evaluation value of each candidate operation, and generate an AP / STA mode switching instruction based on the ε-greedy strategy; The switching instruction is written into the hardware register through the WIFI chipset driver interface, triggering the generation of an ad hoc network topology structure including bridge nodes, and transmitting the topology node distribution data to the routing layer.

[0008] Furthermore, the vehicle ad hoc network and V2V interaction method based on smart connected vehicle WiFi of the present invention, based on the node distribution information in the dynamic ad hoc network topology, embeds the path delay field in the OLSR protocol message to generate a pheromone data packet, and updates the routing table items of each node through periodic broadcasting, including: In the OLSR protocol Hello message header, the delay field is extended to periodically generate pheromone packets containing the number of hops, RSSI value, and current queue length; The intermediate node calculates the pheromone concentration gradient based on the delay field value and the RSSI attenuation coefficient, and updates the routing table entries in descending order of the gradient value; The routing table is transmitted across layers to the priority queue manager of the message transmission layer through the Socket interface, driving high-priority messages to seize low-latency paths. Furthermore, the vehicle self-organizing network and V2V interaction method based on smart connected vehicle WiFi described in the present invention, the V2V message received from the vehicle application layer, the message type and the number of propagation hops are input into the federated learning classification model, the priority label is output and embedded in the message header, and the MAC layer channel preemption mechanism is triggered, including: extracting the message type, propagation hops and end-to-end delay locally in the vehicle to construct a ternary feature vector, adding Laplace noise to implement differential privacy encryption; Upload the encrypted feature vector to the federated learning classification model of the edge server, aggregate multi-node data and update the message priority classification rules; The classification rules are digitally signed and written into the vehicle communication protocol stack via OTA, and the emergency braking message is labeled with the highest priority and the MAC layer CSMA / CA backoff mechanism is triggered to bypass. Furthermore, the vehicle self-organizing network and V2V interaction method based on the WiFi of intelligent connected vehicles described in the present invention, the shared position data of neighboring vehicles and the local IMU sensor data are aligned in time and space, the vehicle posture estimation is generated by the particle filter algorithm, the collaborative control instructions are calculated based on the Nash equilibrium strategy matrix, and the optimized multi-hop path is distributed to the target vehicle, including: Implement UTM projection transformation on the neighboring vehicle position data in the GPS coordinate system, perform sliding window alignment with the local IMU angular velocity data, and generate vehicle pose estimation through the particle filter algorithm; Calculate the safety distance threshold between the vehicle and the neighboring vehicle based on the Nash equilibrium strategy matrix, and generate a control instruction including the coordinated deceleration gradient value; The control instructions are distributed to the target vehicle group via the IPv6 multicast address, and the decision delay data is written into the routing layer path redundancy calculation module.

[0009] Furthermore, the vehicle ad hoc network and V2V interaction method based on smart connected vehicle WiFi described in the present invention, the collection of AP node load rate and channel conflict times, iterative calculation of AP / STA ratio threshold by particle swarm optimization algorithm, triggering node role reconfiguration according to the calculation result, and feeding back the load status to the ad hoc network topology construction module include: The AP node density, STA connection number and channel conflict rate are quantified into a three-dimensional vector space, and the optimal AP / STA ratio threshold is iteratively calculated using the particle swarm algorithm; In areas where the AP density exceeds the threshold, a downgrade command is triggered, and the edge AP nodes are switched to STA mode and establish an 802.11s bridge link with the central AP; The role allocation result is written into the configuration parameter library of the ad hoc network topology construction module, triggering the topology structure reconstruction based on the current load status.

[0010] Furthermore, the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention also includes: The vehicle speed is divided into three intervals: [0, 20), [20, 60), and [60, ∞) km / h to define low-speed, medium-speed, and high-speed states. The current channel occupancy level is associated with the AP node density interval to construct a three-dimensional discrete state space. Input the three-dimensional discrete state space into the deep Q network model, and output the value function evaluation values ​​of three operations: AP mode maintenance, STA mode switching, and bridge mode startup; The mode switching probability distribution is calculated according to the value evaluation value of the cost function, and the WIFI chipset is driven to switch to the STA mode when the channel occupancy rate exceeds a threshold.

[0011] Furthermore, in the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the calculation of the pheromone concentration gradient includes: The path delay value is recorded in the TCL field of the OLSR protocol, the initial pheromone concentration is calculated based on the RSSI signal strength, and the delay weight factor is set to the inverse of the current network load rate; When the intermediate node forwards the data packet, it updates the pheromone concentration value according to the formula: concentration attenuation = number of hops × preset attenuation coefficient, and generates gradient distribution data; The gradient distribution data is written into the priority flag of the routing table entry, and the path with the highest flag value is preferentially selected to establish a TCP long connection. Furthermore, in the vehicle self-organizing network and V2V interaction method based on smart connected vehicle WiFi described in the present invention, the differential privacy encryption includes: Add Laplace distribution noise data to the feature vector of message type and propagation hop number, so that the correlation entropy value between single-hop transmission record and message type exceeds a preset threshold; Aggregating encrypted data through edge servers to update the Softmax output layer weights of the federated learning classification model; The classification rules are sent to the vehicle terminal after being attached with a SHA-256 hash value, and integrity verification is performed in the communication protocol stack parsing module.

[0012] Furthermore, in the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the bridge link includes: When subnet isolation caused by AP node degradation is detected, a dual-mode node supporting the 802.11s bridging protocol is selected as a relay; The mapping relationship between the source subnet ID, destination subnet ID and next-hop MAC address is maintained in the routing table of the bridge node, and the control message is forwarded through EtherType 0x88CC; The throughput and delay data of the bridge link are written into the particle swarm algorithm iteration parameter constraint library to limit the calculation range of the AP / STA ratio.

[0013] Beneficial effects of the present invention: The present invention realizes dynamic self-organizing networking by reusing the existing WiFi modules of the vehicle, significantly reducing the deployment cost of collaborative communication. Based on the AP / STA mode switching mechanism of the deep Q network (DQN) model and the three-dimensional discrete state space, the network topology is dynamically generated in combination with the ε-greedy strategy, so that the vehicle nodes can autonomously adapt to different mobile speeds and network load scenarios. For example, in high-speed movement or high-channel conflict areas, DQN triggers the STA mode switching instruction and combines the 802.11s bridging protocol to maintain network connectivity, avoiding the dependence of traditional V2X technology on dedicated hardware and infrastructure. At the same time, the particle swarm optimization algorithm dynamically adjusts the node role allocation by iteratively calculating the AP / STA ratio threshold, reducing local channel congestion, and improving network scalability and resource utilization.

[0014] The present invention improves real-time performance and reliability through cross-layer collaborative optimization and distributed decision-making mechanisms. The OLSR protocol extends the delay field and combines it with the pheromone gradient calculation to achieve dynamic optimization of multi-hop paths, so that high-priority messages can reduce end-to-end delays through the MAC layer preemption mechanism. The federated learning model aggregates encrypted feature data to generate global classification rules, and realizes rapid identification and transmission of emergency messages while protecting privacy. In addition, the particle filter fuses multi-source sensor data to generate vehicle posture estimates, combines the Nash equilibrium strategy matrix to output collaborative control instructions, and improves group collaboration efficiency through the IPv6 multicast distribution mechanism. The closed-loop load balancing and routing redundancy feedback mechanism further enhances the network's autonomy and adapts to the dynamic needs of complex traffic scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0016] Figure 1 A timing diagram of a vehicle ad hoc network and a V2V interaction method based on smart connected vehicle WiFi provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings. In order to better understand the purpose of the present invention, the present invention is further described in detail below.

[0018] See also Figure 1 The present invention provides a vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi, including: The vehicle motion state data and network state data are collected through the vehicle-mounted sensors, and the speed, position coordinates and network load rate are normalized to generate a multi-dimensional feature vector; The vehicle motion state data (including speed, position coordinates) and network state data (such as network load rate) are collected in real time through on-board sensors, and these heterogeneous data are normalized to eliminate dimensional differences and generate a multi-dimensional feature vector of unified dimension. The normalization process uses the minimum-maximum value scaling method to map each parameter to a preset numerical range to provide standardized input for subsequent model processing. This step provides basic data support for subsequent dynamic networking decisions.

[0019] Input the multi-dimensional feature vector into a pre-trained deep Q network model, output an operation instruction set including AP / STA mode switching probability, and select the optimal instruction according to the ε-greedy strategy to build a dynamic self-organizing network topology; The normalized multi-dimensional feature vector is input into the pre-trained deep Q network (DQN) model, and the output is a set of operation instructions containing the switching probability of AP (access point) mode and STA (station) mode. The DQN model is trained through historical vehicle networking scenario data to learn the optimal mode switching strategy under different network conditions. Based on the ε-greedy strategy, the model selects the operation instruction with the highest current value function evaluation value with a certain probability, dynamically adjusts the AP / STA role of the vehicle node, and generates a self-organizing network topology structure containing bridge nodes. The switching instruction is written into the hardware register through the WiFi chipset driver interface, the network topology configuration is updated in real time, and the topology distribution data is passed to the routing layer to achieve adaptive optimization of the network structure.

[0020] Based on the node distribution information in the dynamic ad hoc network topology, a path delay field is embedded in the OLSR protocol message to generate a pheromone data packet, and the routing table entries of each node are updated through periodic broadcasting; Based on the dynamic ad hoc network topology, the Hello message header of the OLSR (Optimized Link State Routing) protocol is extended, and the path delay field is embedded to generate a pheromone data packet. The intermediate node calculates the path pheromone concentration gradient based on the received pheromone data packet and the RSSI (Received Signal Strength Indicator) attenuation coefficient, and updates the local routing table items in descending order of the gradient value. The routing table is transmitted across layers to the priority queue manager of the message transmission layer through the Socket interface, so that high-priority messages can preferentially occupy low-latency paths and reduce end-to-end transmission delays.

[0021] Receive V2V messages from the vehicle application layer, input the message type and propagation hop count into the federated learning classification model, output the priority tag and embed it into the message header, and trigger the MAC layer channel preemption mechanism; The type, propagation hop count, and end-to-end delay characteristics of the V2V message are extracted on the vehicle terminal, a ternary feature vector is constructed, and Laplace noise is added to implement differential privacy encryption to prevent sensitive information leakage. The encrypted feature vector is uploaded to the federated learning classification model of the edge server, and the message priority classification rules are updated by aggregating multi-node data. The classification rules are digitally signed and sent to the vehicle terminal via OTA (over-the-air download technology) and embedded in the communication protocol stack. Emergency braking messages are marked as the highest priority, triggering the MAC layer CSMA / CA backoff mechanism to bypass the regular channel competition process and achieve low-latency transmission.

[0022] The shared position data of neighboring vehicles is aligned with the local IMU sensor data in time and space, and the vehicle pose estimation is generated through the particle filter algorithm. The collaborative control instructions are calculated based on the Nash equilibrium strategy matrix and distributed to the target vehicle through the optimized multi-hop path. The GPS location data shared by neighboring vehicles is transformed into a UTM (Universal Transverse Mercator) projection, and the sliding window is aligned with the angular velocity data of the local IMU (Inertial Measurement Unit) to eliminate the spatial and temporal deviation. The multi-source sensor data is fused through the particle filter algorithm to generate a high-precision vehicle pose estimate. The safe distance threshold between the vehicle and the neighboring vehicle is calculated based on the Nash equilibrium strategy matrix, and a control instruction containing a coordinated deceleration gradient value is generated. The instruction is distributed to the target vehicle group along the optimized multi-hop path through the IPv6 multicast address, and the decision delay data is recorded to optimize the routing redundancy calculation.

[0023] The AP node load rate and the number of channel conflicts are collected, and the AP / STA ratio threshold is iteratively calculated through the particle swarm optimization algorithm. The node role reconfiguration is triggered according to the calculation results, and the load status is fed back to the self-organizing network topology construction module.

[0024] The AP node density, STA connection number and channel conflict rate are quantified into a three-dimensional vector space, and the particle swarm optimization (PSO) algorithm is input to iteratively calculate the optimal AP / STA ratio threshold. When the AP node density exceeds the threshold, the edge AP node downgrade instruction is triggered, switching it to STA mode and establishing a bridge link based on the 802.11s protocol with the central AP. The role allocation result is fed back to the self-organizing network topology construction module to drive the network topology reconstruction, realize dynamic load balancing and efficient use of channel resources.

[0025] The above technical contents constitute a closed-loop collaborative system: data collection and feature processing provide input for the DQN model, driving AP / STA mode switching and dynamic networking; networking topology information supports routing protocol optimization to ensure low-latency path selection; message priority classification and collaborative decision-making modules rely on routing optimization results to achieve efficient instruction distribution; load balancing mechanism monitors network status in real time, and feeds back to the topology construction module through node role reconfiguration to form an adaptive adjustment loop. Each module jointly improves the real-time, reliability and scalability of vehicle self-organizing networks through cross-layer data interaction and status sharing.

[0026] Specifically, the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the construction of dynamic self-organizing network topology includes: The vehicle speed value is discretized into three states: low speed, medium speed, and high speed, and the current AP node density and channel occupancy are associated to construct a three-dimensional state space; Input the three-dimensional state space into a pre-trained deep Q network model, output the value function evaluation value of each candidate operation, and generate an AP / STA mode switching instruction based on the ε-greedy strategy; The switching instruction is written into the hardware register through the WIFI chipset driver interface, triggering the generation of an ad hoc network topology structure including bridge nodes, and transmitting the topology node distribution data to the routing layer.

[0027] The vehicle speed values ​​collected in real time are divided into three discrete states: low speed, medium speed, and high speed according to the preset intervals. The interval division is based on the speed limit standards of urban roads and the characteristics of vehicle movement, and the model processing efficiency is improved by reducing the dimension of the state space. For example, the low speed state corresponds to congestion or parking scenarios, the medium speed corresponds to regular driving, and the high speed corresponds to expressway or highway scenarios. This discretization process provides structured input for subsequent association analysis.

[0028] The discretized speed state is associated with the current AP (access point) node density level and channel occupancy level to form a three-dimensional discrete state space. The AP node density is obtained by scanning the number of surrounding SSIDs and signal strength statistics, and the channel occupancy is quantified by monitoring the proportion of channel idle time. The three-dimensional state space represents the comprehensive state of vehicle mobility, network coverage and channel load, and provides multi-dimensional environmental perception data for mode switching decisions.

[0029] The constructed three-dimensional state space is input into the pre-trained deep Q network (DQN) model, and the value function evaluation values ​​of the three operations of AP mode maintenance, STA (station) mode switching and bridge mode startup are output. The DQN model is trained with historical networking scenario data to learn the long-term benefits of each operation on network throughput, latency and other indicators under different states. The model uses a convolutional neural network structure to extract state space features and output the value probability distribution of each operation.

[0030] Based on the value function evaluation value output by the DQN model, the ε-greedy strategy is used to generate mode switching instructions. This strategy selects the operation with the highest value evaluation value (utilization) with a preset probability, or selects other operations (exploration) with a random probability, balancing network optimization and adaptability to unknown states. For example, in a high-load channel state, it is preferred to switch to STA mode to reduce interference, while retaining the bridge mode in a low-load scenario to expand network coverage.

[0031] The generated mode switching instruction is written into the hardware register through the WiFi chipset driver interface to trigger the wireless communication module working mode switching. The AP mode node creates an independent SSID and broadcasts a beacon frame, the STA mode node scans and associates with the neighboring AP, and the bridge mode node establishes a wireless distributed system (WDS) under the 802.11s protocol. This generates an ad hoc network topology containing multi-hop bridge nodes to achieve dynamic interconnection between vehicle nodes.

[0032] The reconstructed topology node distribution data is passed to the routing layer through shared memory or inter-process communication mechanism. The routing layer updates the neighbor node list based on the topology information, calculates the optimal path weight, and provides the topology change event trigger signal for the OLSR protocol. This data flow realizes the coordinated optimization of the network layer and the MAC layer, and supports the dynamic update of routing table entries.

[0033] The above technical contents form a closed-loop control link: vehicle state discretization and three-dimensional space construction provide structured input for the DQN model; model reasoning combined with the ε-greedy strategy generates switching instructions that take into account stability and exploration; hardware instruction execution directly drives network topology reconstruction and affects subsequent state space parameters through data feedback mechanism; topology data is transmitted across layers to achieve the linkage between routing optimization and resource scheduling. Each step realizes the adaptive adjustment of the vehicle self-organizing network topology to dynamic traffic scenarios through the progressive relationship of state perception, decision execution, and effect feedback.

[0034] Specifically, the vehicle ad hoc network and V2V interaction method based on smart connected vehicle WiFi of the present invention, based on the node distribution information in the dynamic ad hoc network topology, embeds the path delay field in the OLSR protocol message to generate a pheromone data packet, and updates the routing table items of each node through periodic broadcasting, including: In the OLSR protocol Hello message header, the delay field is extended to periodically generate pheromone packets containing the number of hops, RSSI value, and current queue length; The intermediate node calculates the pheromone concentration gradient based on the delay field value and the RSSI attenuation coefficient, and updates the routing table entries in descending order of the gradient value; The routing table is transmitted across layers to the priority queue manager of the message transmission layer through the Socket interface, driving high-priority messages to preempt low-latency paths sorted in descending order based on pheromone concentration gradient values. The pheromone concentration gradient is obtained by weighted calculation of path delay, RSSI attenuation coefficient and network load rate. The multi-hop routing path generation method described in the present invention has the following specific implementation steps and technical associations: A custom delay field is extended in the Hello message header of the OLSR (Optimized Link State Routing) protocol to record the transmission delay data from the current node to the neighboring node. The Hello message is broadcast at a preset period, carrying the number of hops, RSSI (received signal strength indication) value and network layer queue length information to form a dynamic pheromone data packet. The queue length is obtained by counting the number of data packets to be sent on the network interface, reflecting the instantaneous load status of the node. This extension mechanism realizes the real-time collection and flooding transmission of routing information, providing a data basis for path optimization.

[0035] After receiving the pheromone data packet, the intermediate node parses the delay field value and associates it with the RSSI attenuation coefficient to calculate the path pheromone concentration gradient. The attenuation coefficient is dynamically adjusted according to the wireless channel fading model to characterize the attenuation degree of signal strength with distance and obstacles. The node sorts the adjacent paths in descending order by gradient value, and gives priority to high-concentration gradient paths to update local routing table entries. This calculation process integrates link quality and network load indicators to avoid path congestion or signal attenuation problems caused by a single metric.

[0036] The updated routing table is transmitted across layers to the priority queue manager of the message transport layer via the Socket interface. The Socket interface is bound to a specific port number and uses the UDP protocol to achieve low-overhead data transmission. The priority queue manager parses the latency and hop count data in the routing table entries, dynamically allocates low-latency path resources for high-priority messages, and triggers the MAC layer channel preemption mechanism. For example, emergency safety messages are directly mapped to the preset high-priority queue, bypassing the conventional scheduling algorithm to achieve fast forwarding.

[0037] Based on the path delay information provided by the routing table, the priority queue manager implements a differentiated scheduling strategy. High-priority messages preempt low-latency paths by reserving time slots or increasing retransmission priorities, while low-priority messages are distributed to multiple paths using a load balancing strategy. The preemption mechanism is synchronized with the routing maintenance cycle of the OLSR protocol to avoid message disorder or packet loss caused by frequent path switching.

[0038] The above technical contents constitute the adaptive routing optimization link: protocol extension realizes the periodic collection of topology and load data; pheromone gradient calculation associates the signal strength of the physical layer with the status of the network layer to generate a comprehensive path evaluation result; the cross-layer transmission mechanism opens up the data interaction between the routing layer and the transmission layer to support priority scheduling decisions; the path preemption strategy dynamically adjusts resource allocation based on real-time routing information to form a closed-loop control of "data collection-path evaluation-resource scheduling". Each step improves the stability and real-time performance of multi-hop routing in high-speed vehicle movement scenarios through protocol extension, cross-layer collaboration and dynamic scheduling linkage.

[0039] Specifically, the vehicle self-organizing network and V2V interaction method based on WiFi of intelligent connected vehicles described in the present invention, the V2V message of the vehicle application layer is received, the message type and the number of propagation hops are input into the federated learning classification model, the priority label is output and embedded in the message header, and the MAC layer channel preemption mechanism is triggered, including: extracting the message type, the number of propagation hops and the end-to-end delay in the vehicle to construct a ternary feature vector, adding Laplace noise to implement differential privacy encryption; Upload the encrypted feature vector to the federated learning classification model of the edge server, aggregate multi-node data and update the message priority classification rules; The classification rules are digitally signed and written into the vehicle communication protocol stack via OTA, adding the highest priority tag to the emergency braking message and triggering the MAC layer CSMA / CA backoff mechanism to bypass.

[0040] In the vehicle local communication protocol stack, the type identifier, number of transmitted hops and end-to-end delay measurement of the V2V message are extracted in real time to construct a three-dimensional feature vector. The message types are defined according to the SAE J2735 standard and are divided into categories such as safety warnings, collaborative control, and infotainment. To protect data privacy, Laplace noise that complies with the differential privacy mechanism is added to the feature vector so that the transmission record of a single vehicle cannot be reversed through feature association. The noise injection process is completed in the on-board processing unit, and the original sensitive information is deleted immediately after the encrypted feature data is generated.

[0041] The encrypted feature vector is uploaded to the federated learning classification model of the edge server through the edge computing node of the Internet of Vehicles. The server uses a secure multi-party computing protocol to aggregate multi-node data and update the decision boundary of the message priority classification rule. The federated learning model adopts a convolutional neural network structure and completes the model parameter update in the encrypted data domain through a distributed gradient descent algorithm, preventing the original data from leaving the vehicle terminal, and achieving a balance between data privacy and model accuracy.

[0042] The updated classification rules are digitally signed by the edge server using the RSA algorithm to generate a rule file containing the version number and effective timestamp. The signature file is distributed to the vehicle terminal through the OTA (Over-the-Air Download Technology) encrypted channel, and the signature verification and integrity check are completed in the rule parsing module of the communication protocol stack. The verified rule file is loaded into the QoS (Quality of Service) policy engine of the protocol stack to form a message priority mapping table.

[0043] When an emergency braking message is identified, the classification engine embeds the highest priority tag in the message header. This tag triggers the CSMA / CA backoff mechanism optimization process of the MAC layer: priority messages directly enter the zero backoff window sending phase and skip the conventional random backoff counting process. At the same time, the physical layer modulation and coding scheme automatically switches to the more robust QPSK modulation to improve the transmission reliability of high-priority messages in complex channel environments.

[0044] The above technical contents form a collaborative optimization link: feature vector construction and encryption realize the unity of local data validity and privacy; the federated learning mechanism completes the classification model iteration under the premise of protecting data privacy; the secure delivery process ensures the integrity and legality of the rule file; priority tags trigger cross-layer transmission optimization to achieve low-latency transmission of emergency messages. Through the progressive relationship of data encryption, distributed learning, secure transmission, and protocol optimization, each link builds a V2V message classification system that takes into account both privacy protection and transmission efficiency.

[0045] Specifically, the vehicle self-organizing network and V2V interaction method based on WiFi of intelligent connected vehicles described in the present invention, the temporal and spatial alignment of the shared position data of neighboring vehicles and the local IMU sensor data, the generation of vehicle posture estimation through a particle filter algorithm, the calculation of collaborative control instructions based on the Nash equilibrium strategy matrix, and the distribution to the target vehicle through the optimized multi-hop path include: Implement UTM projection transformation on the neighboring vehicle position data in the GPS coordinate system, perform sliding window alignment with the local IMU angular velocity data, and generate vehicle pose estimation through the particle filter algorithm; Calculate the safety distance threshold between the vehicle and the neighboring vehicle based on the Nash equilibrium strategy matrix, and generate a control instruction including the coordinated deceleration gradient value; The control instructions are distributed to the target vehicle group via the IPv6 multicast address, and the decision delay data is written into the routing layer path redundancy calculation module.

[0046] The GPS location data shared by neighboring vehicles is converted into the UTM (Universal Transverse Mercator) plane coordinate system to eliminate the influence of the earth's curvature on the calculation of the relative position of close-range vehicles. At the same time, the angular velocity data of the local IMU (Inertial Measurement Unit) is collected, and the sliding window mechanism is used to align the timestamps to compensate for the difference in sensor sampling frequency. The length of the sliding window is dynamically adjusted according to the vehicle's motion state. The window is shortened in high-speed scenarios to improve real-time performance, and the window is extended in low-speed scenarios to improve data fusion accuracy. This step realizes the unification of the spatiotemporal benchmarks of multi-source heterogeneous data.

[0047] The UTM coordinate data and IMU angular velocity data after time and space alignment are input into the particle filter algorithm, and the particle weights are iteratively updated through the importance sampling and resampling mechanism. The particle state space contains the vehicle position, heading angle and velocity components, and the kinematic model prediction is matched with the multi-sensor observation value to generate a high-confidence vehicle pose estimation result. This fusion process effectively suppresses the positioning drift problem caused by GPS signal loss or IMU cumulative error.

[0048] A multi-vehicle cooperative game model is established based on Nash equilibrium theory, and the vehicle posture estimation results, road friction coefficient and braking response time are quantified as strategy matrix parameters. The Pareto optimal solution of the vehicle and the neighboring vehicle in the longitudinal following and lateral lane changing scenarios is calculated to generate a dynamic safety distance threshold. The threshold changes nonlinearly with the square value of the relative speed to adapt to the safety margin requirements under different traffic flow densities.

[0049] According to the safety distance threshold deviation, a PID control algorithm is used to generate a collaborative control instruction including a target deceleration gradient value and an action time window. The gradient value is distributed according to an exponential law, triggering a steep deceleration instruction in an emergency collision avoidance scenario and a gentle speed adjustment instruction in a conventional following scenario. The instruction data is encapsulated in the ASN.1 encoding format, with a timestamp and a vehicle group identifier added.

[0050] The control command is sent to the target vehicle group through the IPv6 multicast address, and the dynamic group membership is maintained using the MLD (Multicast Listening Discovery) protocol. The routing layer gives priority to multicast paths with a delay jitter below the threshold and the least number of hops based on the decision delay data stored in the path redundancy calculation module. The delay data is obtained by timestamp difference calculation and is used to evaluate path stability and trigger routing table updates.

[0051] The above technical contents form a closed-loop decision-making link: spatiotemporal alignment and posture estimation provide high-precision environmental perception data for collaborative decision-making; the Nash equilibrium model transforms the motion relationship between vehicles into a mathematical optimization problem to generate a safety strategy; control instruction encoding and multicast distribution realize the reliable transmission of decision results; delay data is fed back to the routing layer to form a basis for path optimization. Each link realizes the real-time and safety unification of vehicle collaborative decision-making through the two-way interaction of data flow and control flow. The accuracy of posture estimation directly affects the convergence speed of the strategy matrix, and the routing optimization effect determines the end-to-end transmission reliability of control instructions.

[0052] Specifically, the vehicle ad hoc network and V2V interaction method based on smart connected vehicle WiFi of the present invention collects AP node load rate and channel conflict times, iteratively calculates AP / STA ratio threshold through particle swarm optimization algorithm, triggers node role reconfiguration according to the calculation result, and feeds back the load status to the ad hoc network topology construction module, including: The AP node density, STA connection number and channel conflict rate are quantified into a three-dimensional vector space, and the optimal AP / STA ratio threshold is iteratively calculated using the particle swarm algorithm; In areas where the AP density exceeds the threshold, a downgrade command is triggered, and the edge AP nodes are switched to STA mode and establish an 802.11s bridge link with the central AP; The role allocation result is written into the configuration parameter library of the ad hoc network topology construction module, triggering the topology structure reconstruction based on the current load status.

[0053] The AP (access point) node density, STA (station) connection number and channel conflict rate data are collected in real time through the vehicle communication module and mapped to the preset normalized intervals respectively. The AP node density is counted by scanning the surrounding SSID broadcast frequency, the STA connection number is obtained based on the association request frame count, and the channel conflict rate is calculated by the physical layer carrier sense failure rate. The three types of parameters are constructed into a three-dimensional vector space to characterize the distribution characteristics of the regional network load. The quantitative model provides standardized input for dynamic threshold calculation.

[0054] Initialize the particle swarm parameters and set the AP / STA ratio threshold as the optimization objective function. Each particle position corresponds to a threshold candidate value, and the fitness function is defined as the weighted inverse of network throughput and transmission delay. The algorithm adjusts the particle speed by inertia weight and iteratively searches for the Pareto optimal solution in the solution space. When the fitness fluctuation of three consecutive iterations is less than the preset tolerance, the algorithm is judged to have converged and the current optimal threshold is output.

[0055] In areas where the AP density exceeds the threshold, edge AP nodes are screened based on the node location entropy value. After the downgrade instruction is triggered, the target node stops SSID broadcasting and releases the IP address pool, and establishes a Mesh bridge link with the central AP through the IEEE 802.11s protocol. The bridge link enables the Airtime fair scheduling algorithm to allocate wireless channel occupancy time according to the node load ratio, eliminating the problem of channel overload in local areas.

[0056] The node role assignment results are written into the configuration parameter library of the self-organizing network topology construction module, and the AP / STA mode switching rule weights are updated. After receiving the update instruction, the topology construction module rescans the network status and calls the deep Q network model to generate a topology reconstruction plan. The reconstructed topology data is synchronized to the routing layer, driving the OLSR protocol to update the multi-hop routing path priority list, forming a closed-loop control of load status perception-threshold calculation-topology adjustment.

[0057] The above technical contents constitute an adaptive load balancing loop: the quantitative modeling of network status provides input features for the particle swarm algorithm; the output threshold of the optimization algorithm drives the node role switching decision; the establishment of bridge links alleviates local congestion and maintains network connectivity; the role allocation results are fed back to the topology construction module to trigger reconstruction, and the reconstructed topology state affects the network parameter collection again. Each link is linked through parameter transmission and event triggering mechanism to achieve load balancing and resource optimization allocation of vehicle self-organizing networks in dynamic traffic scenarios. Among them, the particle swarm algorithm solves multi-objective optimization problems, the bridging protocol ensures the smooth transition of the topology reconstruction process, and the deep Q network model ensures the matching degree between the reconstruction strategy and the real-time scenario.

[0058] Specifically, the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention also includes: The vehicle speed is divided into three intervals: [0, 20), [20, 60), and [60, ∞) km / h to define low-speed, medium-speed, and high-speed states. The current channel occupancy level is associated with the AP node density interval to construct a three-dimensional discrete state space. Input the three-dimensional discrete state space into the deep Q network model, and output the value function evaluation values ​​of three operations: AP mode maintenance, STA mode switching, and bridge mode startup; The mode switching probability distribution is calculated according to the value evaluation value of the cost function, and the WIFI chipset is driven to switch to the STA mode when the channel occupancy rate exceeds a threshold.

[0059] According to the typical working conditions of urban roads, the vehicle speed is divided into three discrete intervals: [0,20)km / h, [20,60)km / h, and [60,∞)km / h, corresponding to low speed (congestion or parking), medium speed (regular urban driving), and high speed (expressway or high-speed driving). The interval threshold refers to the definition of the association between vehicle speed and road type in the SAE J3016 autonomous driving classification standard, and the model processing efficiency is improved by reducing the dimension of continuous speed data.

[0060] The channel occupancy level and AP node density level are collected in real time and quantified into three discrete values: low, medium, and high. The channel occupancy rate is calculated by monitoring the busy-idle ratio of the target channel CCA (free channel assessment), and the AP node density is based on the number of SSID broadcasts obtained by WiFi spectrum scanning. The speed state, channel occupancy level, and AP node density level are constructed into a three-dimensional discrete state space. Each dimension contains 3 discrete values, forming 27 combination states, which fully characterizes the coupling relationship between vehicle mobility and network load.

[0061] The three-dimensional discrete state space is input into the pre-trained deep Q network (DQN) model, and the value function evaluation values ​​of the three operations of AP mode maintenance, STA mode switching and bridge mode startup are output. The DQN model adopts a dual network architecture, including an evaluation network and a target network, and obtains the state-action mapping relationship through historical networking data training. The model output layer adopts a linear activation function to directly map the expected network benefits of each operation in a specific state.

[0062] Based on the value function evaluation value output by the DQN model, the Boltzmann distribution is used to calculate the mode switching probability distribution. High-value operations obtain higher selection probabilities, while retaining the low-probability random selection mechanism to avoid local optimal solutions. When it is detected that the channel occupancy rate exceeds the preset threshold, the probability calculation module sends a mode switching instruction to the WiFi chipset to trigger the STA mode switching process.

[0063] The mode switching instruction is written into the hardware control register through the PCIe interface of the WiFi chipset. When the AP mode is switched to the STA mode, the SSID broadcast is stopped and the DHCP server resources are released, and the passive scanning process is started to associate with the neighboring AP nodes; when the bridge mode is activated, the Mesh networking function of the 802.11s protocol is enabled to establish a multi-hop wireless distributed system. After the switch is completed, the new network status data is transmitted back to the state space construction module to form a closed-loop feedback.

[0064] The discrete state space construction provides structured input for deep reinforcement learning, converting continuous environmental parameters into processable discrete states; the DQN model generates operation value evaluation based on state space reasoning to drive the probabilistic decision-making mechanism; the mode switching instruction directly changes the network topology configuration through the hardware interface, and the topology change in turn affects the channel occupancy rate and AP density parameters, forming a closed-loop control link of "state perception-decision execution-environmental feedback". Each step realizes the real-time adaptive adjustment of the vehicle self-organizing network topology to dynamic traffic scenarios through the progressive relationship of parameter discretization, model reasoning, probabilistic decision-making and hardware control.

[0065] Specifically, in the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi of the present invention, the calculation of pheromone concentration gradient includes: The TCL (Delay Control List) field is extended in the TC (Topology Control) message of the OLSR protocol to record the path delay value, calculate the initial pheromone concentration based on the RSSI signal strength, and set the delay weight factor to the inverse of the current network load rate; When the intermediate node forwards the data packet, the pheromone concentration value is updated according to the formula: concentration attenuation = number of hops × preset attenuation coefficient. The attenuation coefficient is dynamically adjusted according to the stability of the network topology. Specifically, when the average vehicle speed exceeds 60km / h, the attenuation coefficient increases to 0.8; when the network load rate is higher than 70%, the attenuation coefficient decreases to 0.5. Generate gradient distribution data; The gradient distribution data is written into the priority flag of the routing table entry, and the path with the highest flag value is preferentially selected to establish a TCP long connection.

[0066] The TCL (Delay Control List) field is extended in the TC (Topology Control) message header of the OLSR (Optimized Link State Routing) protocol to record the end-to-end transmission delay data of the path. The delay value is calculated by the timestamp difference method to record the cumulative transmission delay of the data packet from the source node to the current node. At the same time, the initial pheromone concentration is calculated based on the RSSI (Received Signal Strength Indicator) value of the received data packet. The RSSI strength is associated with the transmission power and antenna gain parameters, and is mapped to the initial concentration value through a table lookup method. The TCL field adopts the TLV (Type-Length-Value) encoding format and supports dynamic field expansion and parsing.

[0067] The current network load rate is used as the basis for calculating the delay weight factor, and the weight factor is set to the inverse of the network load rate. The network load rate is obtained by counting the proportion of channel busy time per unit time. When the load rate increases, the weight proportion of the delay parameter is reduced to avoid over-selection of high-load paths. The weight factor configuration module periodically collects MAC layer statistics, generates dynamic adjustment parameters and writes them into the configuration register of the routing protocol stack.

[0068] When the intermediate node forwards the data packet, it parses the path delay value in the TCL field and calculates the pheromone concentration attenuation in combination with the preset attenuation coefficient. The attenuation coefficient is dynamically adjusted according to the stability of the network topology. In the scenario of high-speed vehicle movement, the attenuation coefficient is increased to accelerate the elimination of old paths, and in the stable scenario, the coefficient is reduced to maintain routing consistency. The attenuation calculation result is updated to the local pheromone concentration table to form the path quality gradient distribution data.

[0069] The updated pheromone gradient data is written into the priority flag of the routing table entry, and the flag value is mapped to the range of 0-255 through normalization. The routing engine preferentially selects the path with the highest flag value to establish a TCP long connection and reserves bandwidth resources for the path. The TCP connection enables the fast retransmission mechanism, dynamically adjusts the congestion window growth factor according to the pheromone gradient value, and realizes the matching optimization of transmission rate and path quality.

[0070] The above technical contents constitute the adaptive routing optimization mechanism: protocol field extension realizes the standardized carrying of path quality parameters; dynamic weight factor configuration balances the decision weight of delay and load indicators; concentration decay calculation reflects the spatiotemporal variation characteristics of path quality; priority mapping mechanism converts abstract gradient data into operational routing strategies. Each link forms a closed-loop control link of "data collection-quality assessment-path optimization" through cross-layer interaction and parameter linkage of the protocol stack, in which pheromone gradient data not only guides single path selection, but also realizes transmission layer optimization through TCP connection parameter adjustment, improving the routing stability and transmission reliability of vehicle self-organizing networks in dynamic environments.

[0071] Specifically, in the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the differential privacy encryption includes: Add Laplace distribution noise data to the feature vector of message type and propagation hop number, so that the correlation entropy value between single-hop transmission record and message type exceeds a preset threshold; Aggregating encrypted data through edge servers and updating the Softmax output layer weights of the federated learning classification model; The classification rules are sent to the vehicle terminal after being attached with a SHA-256 hash value, and integrity verification is performed in the communication protocol stack parsing module.

[0072] In the feature extraction module of the vehicle communication protocol stack, a binary feature vector is constructed for the message type encoding (such as emergency braking, cooperative lane change, etc.) and the propagation hop value. A Laplace noise generator is used to inject noise data that conforms to the differential privacy mechanism into each dimension of the feature vector, and the noise level is dynamically adjusted according to the preset privacy budget parameters. After the noise is injected, the mutual information between the single-hop transmission record and the message type is suppressed to below the correlation entropy threshold, making it impossible for attackers to reverse engineer the original message source and propagation path. The noise injection process is completed in the vehicle's local security coprocessor, and the original data is encrypted and temporarily stored in the memory and then erased immediately.

[0073] The encrypted feature vector is uploaded to the edge server through the edge computing node of the Internet of Vehicles, and the server uses a secure multi-party computing protocol to implement the federated learning model update. The encrypted data uploaded by each node performs gradient calculation in the ciphertext domain, and the Softmax output layer weights of the federated learning classification model are updated through the homomorphic encryption algorithm. During the model update process, the local data of each vehicle terminal is always retained in the original device, and the server only obtains the encrypted gradient parameters to achieve a balance between privacy protection and model generalization ability.

[0074] The updated message priority classification rules are digitally signed using the RSA algorithm, with timestamp and version number information attached. The rule file generates a summary value using the SHA-256 hash algorithm, and the rule content and summary value are bound and encapsulated into a secure data package. The encapsulation process is completed in the trusted execution environment (TEE) of the edge server to prevent rule tampering and man-in-the-middle attacks.

[0075] The safety data packet is sent to the vehicle terminal through the cellular network or DSRC channel. The parsing module of the communication protocol stack separates the rule content and hash value, recalculates the summary value using the same SHA-256 algorithm and verifies the consistency. The verified rule file is loaded into the QoS policy engine of the protocol stack, and the message type is mapped to the MAC layer priority tag. After the emergency braking message triggers the priority tag, it bypasses the conventional CSMA / CA backoff counter and directly enters the channel preemption process.

[0076] The above steps form a collaborative mechanism for privacy protection and transmission optimization: noise injection confuses sensitive features and blocks the risk of privacy leakage; federated learning implements model iteration based on encrypted data to ensure classification accuracy; secure encapsulation and verification mechanisms maintain rule integrity; priority tags drive cross-layer transmission optimization. Each link builds a V2V message processing system that takes into account data privacy and communication efficiency through a closed-loop process of "local encryption-cloud aggregation-secure distribution-terminal execution". The combination of differential privacy mechanism and federated learning solves the contradiction between data sharing and privacy protection, and hash verification and digital signatures establish an end-to-end secure trust chain.

[0077] Specifically, the vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi described in the present invention, the bridge link includes: When subnet isolation caused by AP node degradation is detected, a dual-mode node supporting the 802.11s bridging protocol is selected as a relay; The mapping relationship between the source subnet ID, destination subnet ID and next-hop MAC address is maintained in the routing table of the bridge node, and the control message is forwarded through EtherType 0x88CC; The throughput and delay data of the bridge link are written into the particle swarm algorithm iteration parameter constraint library to limit the calculation range of the AP / STA ratio.

[0078] The subnet connectivity is detected by periodically sending ICMPv6 neighbor discovery messages. When no response is received from the target subnet AP node for three consecutive times, a subnet isolation event is determined. The candidate node screening process is started, and nodes that support the IEEE802.11s protocol and have dual-mode (AP / STA) working capabilities are preferentially selected as relays. The screening criteria include signal strength threshold, node load status, and remaining power level. The node with the highest comprehensive score is selected through a weighted scoring algorithm to activate the bridge mode.

[0079] The mapping relationship between the source subnet ID, the destination subnet ID and the next-hop MAC address is established in the routing table of the bridge node, and a hash table structure is used to store and establish a fast query index. The control message is encapsulated by the EtherType 0x88CC identifier and is transparently transmitted across subnets at the data link layer. After the bridge node parses the destination subnet ID of the message, it queries the mapping table to obtain the next-hop MAC address, re-encapsulates the frame header and forwards it to the target subnet, maintaining the communication continuity of the control plane between subnets.

[0080] The throughput, latency and packet loss rate indicators of the bridge link are monitored in real time, and the data is formatted into a JSON structure and written into the particle swarm algorithm iteration parameter constraint library. The constraint library defines the upper and lower boundaries of the AP / STA ratio threshold. When the bridge link latency exceeds the preset threshold, the algorithm search space is dynamically narrowed to limit the calculation range of the AP node density. The constraint parameters are synchronized to the topology construction module through a shared memory mechanism to avoid large-scale topology oscillations caused by the degradation of the bridge link performance.

[0081] The above steps form a closed-loop maintenance mechanism: subnet isolation detection triggers relay node selection to ensure network connectivity; routing table maintenance enables efficient forwarding of cross-subnet control messages; performance data is fed back to the particle swarm algorithm constraint library to reversely optimize network parameter decisions. Each link is linked to the data flow through event triggering, in which the bridge link performance data is used for real-time routing optimization and as historical experience data to guide subsequent AP / STA ratio threshold calculations, achieving a dynamic balance between network load balancing and topology stability.

[0082] The technical solution of the present invention is specifically described in the following embodiments: Embodiment 1: Dynamic self-organizing network topology construction and dynamic switching of node roles; This embodiment realizes the autonomous switching of vehicle node AP / STA mode and network topology reconstruction. First, the vehicle-mounted sensor collects vehicle speed, position and network load rate data, and generates a multi-dimensional feature vector through normalization processing; the speed is discretized into three states: low speed (0-20 km / h), medium speed (20-60 km / h), and high speed (>60 km / h), and the channel occupancy rate and AP node density are associated to construct a three-dimensional discrete state space. The deep Q network (DQN) model outputs the evaluation values ​​of AP mode maintenance, STA switching and bridge mode startup based on the state space, and uses the ε-greedy strategy to generate switching instructions. For example, when a high-speed moving vehicle enters an AP high-density area, DQN triggers the STA mode switching instruction, writes the register through the WiFi chipset driver interface, forms a dynamic topology containing 802.11s bridge nodes, and passes the topology data to the routing layer. This mechanism reduces dependence on infrastructure through distributed decision-making and improves network connectivity in dynamic scenarios.

[0083] Embodiment 2: Multi-hop routing optimization and message priority classification; The Hello message delay field is extended in the OLSR protocol, and pheromone packets containing hop counts, RSSI values, and queue lengths are periodically generated. The intermediate node calculates the pheromone gradient based on the delay weight factor (the inverse of the network load rate) and the RSSI attenuation coefficient, and updates the routing table entries in descending order of the gradient value. At the same time, the vehicle terminal extracts the message type and the number of propagation hops to construct a ternary feature vector, adds Laplace noise to implement differential privacy encryption, and then uploads it to the federated learning model of the edge server. The model aggregates multi-node data to update the classification rules, which are then sent to the vehicle terminal with a digital signature. For example, the emergency braking message is marked as the highest priority, triggering the MAC layer CSMA / CA backoff mechanism to bypass and give priority to the low-latency path. This solution achieves end-to-end low-latency transmission of high-priority messages through cross-layer routing optimization and distributed learning.

[0084] Example 3: Closed-loop load balancing and collaborative decision control; The AP / STA ratio threshold is iteratively calculated through the particle swarm optimization algorithm. When the AP density exceeds the threshold, the edge node downgrade instruction is triggered, and a bridge link is established to alleviate local congestion. At the same time, the UTM projection transformation is implemented on the GPS data of the neighboring vehicle, and after alignment with the local IMU angular velocity data, the particle filter algorithm is used to generate the vehicle pose estimate. The collaborative deceleration gradient instructions are calculated based on the Nash equilibrium strategy matrix and distributed to the target vehicle group via IPv6 multicast. For example, in a vehicle formation scenario, the bridge node dynamically diverts the communication load, and the collaborative control instructions are distributed along the optimized path to reduce the multi-hop transmission delay. This embodiment improves resource utilization and group collaboration efficiency in complex scenarios through the closed-loop linkage of network state feedback and collaborative decision-making.

[0085] Explanation of the technical features of the present invention: Dynamic self-organizing network topology construction and AP / STA mode switching; AP / STA mode: AP (access point) mode means that the vehicle broadcasts SSID and manages connections as a wireless network center node; STA (station) mode means that the vehicle associates to a nearby AP as a terminal node. Mode switching implements hardware layer configuration changes through the WiFi chipset driver interface.

[0086] Deep Q Network (DQN) model: A reinforcement learning model that drives dynamic decision-making by evaluating the long-term benefits of different operations (such as AP mode maintenance and STA switching) through a value function based on state data such as vehicle speed and network load.

[0087] ε-greedy strategy: A decision-making mechanism that balances exploration and exploitation, randomly selecting actions with probability ε (exploring new states), and selecting the current optimal action with probability 1-ε (utilizing known benefits) to avoid local optimal solutions.

[0088] Multi-hop routing optimization and pheromone mechanism; OLSR protocol extension: Add a delay field to the Hello message header of the Optimized Link State Routing Protocol (OLSR) to record the path transmission delay for dynamic calculation of route quality.

[0089] Pheromone data packet: A broadcast message containing the number of hops, RSSI value, and queue length. The intermediate node updates the pheromone concentration gradient based on the delay weight factor (the inverse of the network load rate) to guide the priority sorting of the routing table.

[0090] Federated learning classification model: a distributed machine learning framework that uploads local encrypted feature data (message type, number of hops) from the vehicle to the edge server for aggregation, generates and distributes global classification rules, and implements dynamic labeling of message priorities under privacy protection.

[0091] Load balancing and collaborative decision making; Particle swarm optimization algorithm: It iteratively calculates the three-dimensional vector space of AP node density, STA connection number and channel conflict rate, outputs the optimal AP / STA ratio threshold, and triggers node role reconfiguration to balance the load.

[0092] Nash equilibrium strategy matrix: A multi-vehicle coordination model based on game theory maps vehicle posture estimation and safety distance threshold into a strategy matrix to generate Pareto optimal coordinated deceleration instructions.

[0093] 802.11s bridge link: A mesh network protocol that complies with the IEEE 802.11s standard. It establishes a multi-hop wireless distribution system when subnets are isolated, forwards control messages through the EtherType 0x88CC identifier, and maintains network connectivity.

[0094] The above features form a closed-loop technology system: the DQN model and ε-greedy strategy realize dynamic topology reconstruction, providing a basis for routing optimization; OLSR expansion and pheromone mechanism improve path selection efficiency and support high-priority message transmission of federated learning models; particle swarm algorithm and bridge protocol dynamically adjust network load, and cooperate with Nash equilibrium strategy to achieve low-latency distribution of group control instructions. Each module optimizes linkage through state perception, decision execution and feedback, solving the defects of traditional V2X technology that relies on infrastructure and lacks real-time performance.

[0095] The present invention effectively solves the problems of high deployment cost and poor scalability of existing V2X technology due to reliance on dedicated hardware and infrastructure, as well as low real-time performance caused by insufficient dynamic topology adaptability of traditional networking mode through the following technical solutions: First, the vehicle motion status and network load data are collected in real time through on-board sensors, and the deep Q network (DQN) model and ε-greedy strategy are combined to realize the dynamic switching of the AP / STA role of the vehicle node. The DQN model generates mode switching instructions based on multi-dimensional feature vector reasoning, driving the WiFi chipset to reconstruct the network topology at the hardware layer. For example, the edge AP node is downgraded to STA mode in the high-load channel area, while maintaining network connectivity through the 802.11s bridging protocol. This mechanism does not rely on dedicated communication modules or roadside infrastructure, and can achieve distributed networking by reusing existing WiFi hardware, significantly reducing deployment costs and hardware modification costs. In addition, the three-dimensional discrete state space modeling and dynamic decision-making mechanism enable the network topology to adapt to the high-speed movement of vehicles and reduce communication interruptions caused by frequent topology changes.

[0096] Secondly, by extending the OLSR protocol message to embed the path delay field, combined with the pheromone gradient calculation, dynamic optimization of multi-hop routing is achieved. The intermediate node updates the routing table entry according to the delay and RSSI attenuation coefficient, and the priority queue manager schedules high-priority messages across layers to seize the low-latency path. At the same time, the federated learning classification model aggregates multi-node encrypted feature data at the edge server, generates message priority classification rules and sends them to the vehicle terminal. For example, emergency braking messages bypass the conventional channel competition process through the MAC layer CSMA / CA backoff mechanism and directly trigger low-latency transmission. Through distributed learning and cross-layer collaboration, this solution not only avoids the dependence of traditional V2V communication on centralized base stations, but also improves the real-time and reliability of multi-hop routing, and adapts to the dynamic expansion needs of large-scale vehicle nodes.

[0097] Finally, the AP / STA ratio threshold is iteratively calculated based on the particle swarm optimization algorithm, and the node role is dynamically adjusted to balance the network load. When the AP node density exceeds the threshold, the edge AP downgrade instruction is triggered and a bridge link is established to alleviate local channel congestion. In addition, the collaborative decision-making module generates vehicle pose estimation by fusing multi-source sensor data through the particle filter algorithm, calculates collaborative control instructions in combination with the Nash equilibrium strategy matrix, and distributes them to the target vehicle group through IPv6 multicast. This mechanism links network state perception, path optimization and collaborative control closed loops, while reducing the channel conflict rate, improving network scalability and group collaboration efficiency in complex traffic scenarios. In summary, through dynamic resource allocation and multi-protocol collaboration, this solution achieves efficient autonomy and low-latency interaction in vehicle self-organizing networks.

Claims

1. A vehicle self-organizing network and V2V interaction method based on WiFi of intelligent connected vehicles, characterized in that: include: The vehicle motion state data and network state data are collected through the vehicle-mounted sensors, and the speed, position coordinates and network load rate are normalized to generate a multi-dimensional feature vector; Input the multi-dimensional feature vector into a pre-trained deep Q network model, output an operation instruction set including AP / STA mode switching probability, and select the optimal instruction according to the ε-greedy strategy to build a dynamic self-organizing network topology; Based on the node distribution information in the dynamic ad hoc network topology, a path delay field is embedded in the OLSR protocol message to generate a pheromone data packet, and the routing table entries of each node are updated through periodic broadcasting; Receive V2V messages from the vehicle application layer, input the message type and propagation hop count into the federated learning classification model, output the priority tag and embed it into the message header, and trigger the MAC layer channel preemption mechanism; The shared position data of neighboring vehicles is aligned with the local IMU sensor data in time and space, and the vehicle pose estimation is generated through the particle filter algorithm. The collaborative control instructions are calculated based on the Nash equilibrium strategy matrix and distributed to the target vehicle through the optimized multi-hop path. The AP node load rate and the number of channel conflicts are collected, and the AP / STA ratio threshold is iteratively calculated through the particle swarm optimization algorithm. The node role reconfiguration is triggered according to the calculation results, and the load status is fed back to the self-organizing network topology construction module.

2. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 1 is characterized in that: The construction of dynamic ad hoc network topology includes: The vehicle speed value is discretized into three states: low speed, medium speed, and high speed, and the current AP node density and channel occupancy are associated to construct a three-dimensional state space; Input the three-dimensional state space into a pre-trained deep Q network model, output the value function evaluation value of each candidate operation, and generate an AP / STA mode switching instruction based on the ε-greedy strategy; The switching instruction is written into the hardware register through the WIFI chipset driver interface, triggering the generation of an ad hoc network topology structure including bridge nodes, and transmitting the topology node distribution data to the routing layer.

3. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 1, characterized in that: The method of embedding a path delay field in an OLSR protocol message based on the node distribution information in the dynamic ad hoc network topology to generate a pheromone data packet and updating the routing table entries of each node by periodic broadcasting includes: In the OLSR protocol Hello message header, the delay field is extended to periodically generate pheromone packets containing the number of hops, RSSI value, and current queue length; The intermediate node calculates the pheromone concentration gradient based on the delay field value and the RSSI attenuation coefficient, and updates the routing table entries in descending order of the gradient value; The routing table is transmitted across layers to the priority queue manager of the message transmission layer via the Socket interface, driving high-priority messages to seize the low-latency path.

4. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 1, characterized in that: The receiving V2V message of the vehicle application layer, inputting the message type and the number of propagation hops into the federated learning classification model, outputting the priority tag and embedding it into the message header, and triggering the MAC layer channel preemption mechanism include: extracting the message type, the number of propagation hops and the end-to-end delay in the vehicle to construct a ternary feature vector, adding Laplace noise to implement differential privacy encryption; Upload the encrypted feature vector to the federated learning classification model of the edge server, aggregate multi-node data and update the message priority classification rules; The classification rules are digitally signed and written into the vehicle communication protocol stack via OTA, adding the highest priority tag to the emergency braking message and triggering the MAC layer CSMA / CA backoff mechanism to bypass.

5. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 1, characterized in that: The method of performing spatiotemporal alignment of the shared position data of neighboring vehicles with the local IMU sensor data, generating vehicle pose estimation through a particle filter algorithm, calculating collaborative control instructions based on a Nash equilibrium strategy matrix, and distributing them to the target vehicle through an optimized multi-hop path includes: Implement UTM projection transformation on the neighboring vehicle position data in the GPS coordinate system, perform sliding window alignment with the local IMU angular velocity data, and generate vehicle pose estimation through the particle filter algorithm; Calculate the safety distance threshold between the vehicle and the neighboring vehicle based on the Nash equilibrium strategy matrix, and generate a control instruction including the coordinated deceleration gradient value; The control instructions are distributed to the target vehicle group via the IPv6 multicast address, and the decision delay data is written into the routing layer path redundancy calculation module.

6. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 1, characterized in that: The collecting of AP node load rate and channel conflict times, iterative calculation of AP / STA ratio threshold by particle swarm optimization algorithm, triggering node role reconfiguration according to the calculation result, and feeding back the load status to the self-organizing network topology construction module include: The AP node density, STA connection number and channel conflict rate are quantified into a three-dimensional vector space, and the optimal AP / STA ratio threshold is iteratively calculated using the particle swarm algorithm; In areas where the AP density exceeds the threshold, a downgrade command is triggered, and the edge AP nodes are switched to STA mode and establish an 802.11s bridge link with the central AP; The role allocation result is written into the configuration parameter library of the ad hoc network topology construction module, triggering the topology structure reconstruction based on the current load status.

7. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 2, characterized in that: Also includes: The vehicle speed is divided into three intervals: [0, 20), [20, 60), and [60, ∞) km / h to define low-speed, medium-speed, and high-speed states. The current channel occupancy level is associated with the AP node density interval to construct a three-dimensional discrete state space. Input the three-dimensional discrete state space into the deep Q network model, and output the value function evaluation values ​​of three operations: AP mode maintenance, STA mode switching, and bridge mode startup; The mode switching probability distribution is calculated according to the value evaluation value of the cost function, and the WIFI chipset is driven to switch to the STA mode when the channel occupancy rate exceeds a threshold.

8. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 3 is characterized in that: The calculation of the pheromone concentration gradient comprises: The path delay value is recorded in the TCL field of the OLSR protocol, the initial pheromone concentration is calculated based on the RSSI signal strength, and the delay weight factor is set to the inverse of the current network load rate; When the intermediate node forwards the data packet, it updates the pheromone concentration value according to the formula: concentration attenuation = number of hops × preset attenuation coefficient, and generates gradient distribution data; The gradient distribution data is written into the priority flag of the routing table entry, and the path with the highest flag value is preferentially selected to establish a TCP long connection.

9. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 4, characterized in that: The differential privacy encryption includes: Add Laplace distribution noise data to the feature vector of message type and propagation hop number, so that the correlation entropy value between single-hop transmission record and message type exceeds a preset threshold; Aggregating encrypted data through edge servers to update the Softmax output layer weights of the federated learning classification model; The classification rules are sent to the vehicle terminal after being attached with a SHA-256 hash value, and integrity verification is performed in the communication protocol stack parsing module.

10. The vehicle self-organizing network and V2V interaction method based on intelligent connected vehicle WiFi according to claim 6, characterized in that: The bridge link comprises: When subnet isolation caused by AP node degradation is detected, a dual-mode node supporting the 802.11s bridging protocol is selected as a relay; The mapping relationship between the source subnet ID, destination subnet ID and next-hop MAC address is maintained in the routing table of the bridge node, and the control message is forwarded through EtherType 0x88CC; The throughput and delay data of the bridge link are written into the particle swarm algorithm iteration parameter constraint library to limit the calculation range of the AP / STA ratio.

Images