Automated driving scene-based communication optimization-based internet of vehicles resource allocation method and system

Abstract

This application relates to the field of autonomous driving technology, specifically to a method and system for vehicle-to-everything (V2X) resource allocation based on communication optimization in autonomous driving scenarios. The method includes: multiple clusters formed by vehicle nodes within the coverage area of ​​the same roadside unit; when communication is required, generating multiple candidate multi-hop relay paths from the source cluster head node to the target node; selecting the relay path with the best transmission efficiency as the optimal relay path; constructing and solving a joint resource allocation model with the goal of maximizing the total network throughput, generating resource allocation decisions for each hop link on the optimal relay path; each cluster head node receiving and confirming the resource allocation decision, and performing secondary allocation and local configuration of the obtained resources; after receiving configuration confirmations from all relevant cluster head nodes, establishing a multi-hop relay communication connection from the source cluster head node to the target node along the optimal relay path. This improves the stability, reliability, and efficiency of V2X communication.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, specifically to a method and system for allocating vehicle network resources based on communication optimization in autonomous driving scenarios. Background Technology

[0002] In autonomous driving scenarios, vehicle-to-everything (V2X) communication, as a key technology for achieving efficient communication between vehicles and between vehicles and infrastructure, directly impacts communication quality and the safety of autonomous driving through the rationality of resource allocation. Currently, V2X communication mainly faces the following technical challenges:

[0003] On the one hand, vehicle nodes are in a state of rapid movement, and the network topology changes frequently. Traditional fixed cluster structures are difficult to adapt to this dynamism, which can easily lead to high cluster maintenance costs, frequent communication interruptions, and affect the stability and reliability of communication.

[0004] On the other hand, selecting appropriate relay paths and allocating resources rationally in multi-hop relay communication is a major challenge. Existing methods often only consider path length or the quality of a single link, failing to comprehensively evaluate transmission performance based on factors such as link capacity and stability. This can lead to suboptimal path selections, impacting communication efficiency. Furthermore, resource allocation lacks global optimization, often aiming for local optima, making it difficult to maximize total network throughput and thus unable to meet the demands of real-time, reliable transmission of large amounts of data in autonomous driving scenarios. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method and system for allocating vehicle network resources based on communication optimization in autonomous driving scenarios.

[0006] In a first aspect, the present invention provides a method for allocating vehicle network resources based on communication optimization in an autonomous driving scenario, comprising the following steps:

[0007] S1. Within the coverage area of ​​the same roadside unit, each vehicle node periodically broadcasts its own status information and dynamically forms multiple clusters based on the status information of its neighboring nodes using a weighted clustering algorithm. Each cluster elects a cluster head node. Each cluster head node reports its own identity and cluster member information to the roadside unit. The roadside unit constructs and maintains global cluster topology information based on the information reported by each cluster head node.

[0008] S2. When a cluster head node needs to communicate with a target node, it sends a communication request to the roadside unit. The roadside unit generates multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluates the end-to-end transmission performance of each path. The transmission performance is a weighted function of the link capacity and link stability of each hop of the relay path. The cluster head node that sends the communication request to the roadside unit is defined as the source cluster head node.

[0009] S3. Select the relay path with the best transmission efficiency as the optimal relay path;

[0010] S4. The roadside unit constructs and solves a joint resource allocation model with the goal of maximizing the total network throughput, and generates resource allocation decisions for each hop link on the optimal relay path.

[0011] S5. The roadside unit sends the resource allocation decision to all cluster head nodes involved in the optimal relay path; each cluster head node receives and confirms the resource allocation decision, and performs secondary allocation and local configuration of the obtained resources according to the priority of the services of its cluster nodes.

[0012] S6. After receiving configuration confirmation from all relevant cluster head nodes, the roadside unit sends a communication start command to the source cluster head node, thereby establishing a multi-hop relay communication connection from the source cluster head node to the target node along the optimal relay path.

[0013] By periodically broadcasting status information from vehicle nodes and dynamically clustering them using a weighted clustering algorithm, the system can adapt to dynamic changes in network topology caused by rapid vehicle movement, reducing cluster maintenance costs, decreasing communication interruption frequency, and enhancing communication stability. Roadside units construct and maintain global cluster topology information, providing accurate and comprehensive network status data for subsequent communication path planning and resource allocation. During communication, multiple candidate multi-hop relay paths are generated based on global topology and channel information, and transmission efficiency is evaluated. The optimal path is selected, comprehensively considering factors such as link capacity and stability, thus improving communication efficiency. A joint resource allocation model is constructed and solved to allocate resources with the goal of maximizing total network throughput, meeting the requirements for real-time and reliable transmission of large amounts of data in autonomous driving scenarios and improving the performance of the entire vehicle-to-everything (V2X) system.

[0014] As a preferred embodiment of the technical solution of the present invention, the step of dynamically forming multiple clusters using a weighted clustering algorithm in S1, and electing a cluster head node for each cluster, includes:

[0015] S11. Each vehicle node calculates its own weight value to become the cluster head based on the received state information of its neighboring nodes.

[0016] S12. Determine if there are any vehicle nodes with weight values ​​less than a set threshold:

[0017] If it exists, then execute the first cluster formation process:

[0018] S12a1. Select the vehicle node with the smallest weight value as the cluster head node;

[0019] S12a2, Only neighboring nodes that meet preset conditions are allowed to join the cluster where the cluster head node is located as cluster member nodes; the preset conditions include at least one of the following: the distance between the neighboring node and the cluster head node is less than a first preset distance threshold, the channel signal-to-noise ratio between the neighboring node and the cluster head node is greater than a first preset signal-to-noise ratio threshold, the remaining energy of the neighboring node is greater than a first preset energy threshold, and the angle between the travel directions of the neighboring node and the cluster head node is less than a preset angle threshold.

[0020] S12a3. For the remaining vehicle nodes that have not joined the cluster, repeat the weight calculation and cluster head election among the remaining nodes until all vehicle nodes are included in a cluster.

[0021] If the weight values ​​of all vehicle nodes are greater than or equal to the set threshold, execute the second cluster formation process:

[0022] S12b1. Select N vehicle nodes with the smallest relative weight values ​​as the cluster head nodes of different clusters;

[0023] S12b2. For the remaining vehicle nodes that are not selected as cluster head nodes, select and join a cluster according to the preset joining rules.

[0024] By calculating the weight value of each vehicle node to become a cluster head and executing different clustering processes based on the relationship between the weight value and a set threshold, cluster head nodes can be determined more rationally. The first clustering process, for cases where nodes with small weight values ​​exist, selects the node with the smallest weight value as the cluster head and sets strict preset conditions to filter cluster members, ensuring the quality and stability of intra-cluster node communication. The second clustering process, when all node weight values ​​are large, selects multiple nodes with the relatively smallest weight values ​​as cluster heads, making clustering more flexible and adaptable to scenarios with different vehicle node distributions, thus improving the efficiency and rationality of clustering.

[0025] As a preferred embodiment of the technical solution of the present invention, the preset joining rules include at least one of the following: selecting the cluster head node that is closest to the node, selecting the cluster head node with the best channel quality, or selecting the cluster head node with the highest remaining energy.

[0026] Joining the nearest cluster head node reduces communication distance, signal attenuation, and interference; joining the cluster head node with the best channel quality ensures communication reliability and data transmission rate; joining the cluster head node with the highest remaining energy helps extend its operating time and reduces the frequency of cluster reconfiguration due to energy depletion. Vehicle nodes can flexibly choose joining rules based on actual conditions, improving the success rate of joining a cluster and the stability of the cluster.

[0027] As a preferred embodiment of the technical solution of the present invention, the step of adding the remaining vehicle nodes that were not selected as cluster head nodes to the cluster includes:

[0028] For each remaining vehicle node that was not selected as a cluster head node, calculate its joining utility value with all selected cluster head nodes;

[0029] The added utility value The calculation formula is:

[0030]

[0031] in, For vehicle nodes With cluster head node distance, For vehicle nodes With cluster head node The signal-to-noise ratio between them Cluster head node The remaining energy, These are the normalized weighting coefficients;

[0032] Selection of each remaining vehicle node that was not selected as a cluster head node Find the cluster head node with the largest value and send a join request to it;

[0033] The cluster head node receives a join request. If the requesting node meets the preset conditions, it approves the join and updates the cluster member list; if the requesting node does not meet the preset conditions, it rejects the join and the requesting node then requests a different cluster member. The cluster head node with the second largest value sends a join request until it successfully joins a cluster or all candidate cluster head nodes fail to join.

[0034] By comprehensively considering multiple key factors such as distance, signal-to-noise ratio, and the remaining energy of the cluster head node, the decision-making process for node joining a cluster becomes more scientific and rational. The normalized weighting coefficients can be adjusted to reflect the relative importance of each factor based on actual needs, enhancing the method's flexibility and adaptability. Nodes select a cluster head node based on their joining utility value and send a joining request. The cluster head node then approves the request based on preset conditions. This interactive joining mechanism ensures the quality and communication performance of nodes within the cluster, improving the stability and reliability of the entire vehicular network.

[0035] As a preferred embodiment of the technical solution of the present invention, the method further includes:

[0036] If a vehicle node sends a join request to all cluster head nodes and is rejected by all of them, the vehicle node will bootstrap itself into an isolated cluster head node and broadcast that it has become the new cluster head node; the remaining isolated vehicle nodes that are also unable to join any cluster will, upon detecting the announcement, attempt to join the cluster formed by the new cluster head node according to the preset join rules or join utility value.

[0037] After the roadside unit detects a new cluster formed by isolated vehicle nodes in the network, it incorporates the new cluster into the global cluster topology management and treats it as an ordinary cluster for resource allocation in the subsequent resource allocation process.

[0038] This provides a self-bootstrapping clustering solution for isolated vehicle nodes that cannot join existing clusters, avoiding the problem of isolated nodes being unable to participate in communication and improving network connectivity. The roadside unit incorporates the new cluster into the global cluster topology management and treats it as a regular cluster in subsequent resource allocation, ensuring that the new cluster can participate in resource allocation fairly with other clusters. This makes the resource allocation of the entire vehicle network system more reasonable and comprehensive, further improving the overall performance and stability of the system.

[0039] As a preferred embodiment of the technical solution of the present invention, step S2 includes:

[0040] S21. The roadside unit uses the maintained global cluster topology graph as the network graph. Each cluster head node in the network graph is a vertex. If cluster head nodes can communicate directly, there is an edge between them. Using the K shortest path algorithm, K shortest paths from the source cluster head node to the target node are calculated in the network graph to generate a candidate path set. K is a natural number greater than 1.

[0041] S22. For each candidate path in the candidate path set, calculate the end-to-end transmission performance value P. The calculation formula is as follows:

[0042]

[0043] Where H is the total number of hops for the path. Let be the spectral efficiency of the h-th hop link on the path. , The expected lifetime of the h-th hop link on the path is calculated based on the position and velocity vectors of the adjacent cluster head nodes. and These are the normalized weighting coefficients, and + = 1; The channel quality indication value reported by the h-th hop link; These are calibration coefficients preset according to the mapping relationship between modulation and coding strategies;

[0044] S23. Sort all candidate paths in the candidate path set according to the transmission performance value.

[0045] The K-shortest path algorithm is used to calculate a set of candidate paths, generating multiple different paths for selection, increasing the likelihood of finding the optimal path. When calculating the end-to-end transmission performance value of a path, multiple factors are comprehensively considered, including the total number of hops, spectral efficiency, expected link lifetime, and channel quality indicator. Normalized weighting coefficients are used to adjust the influence of each factor, ensuring that the transmission performance value accurately reflects the communication performance of the path. Ranking the candidate paths based on their transmission performance values ​​provides a reliable basis for subsequent selection of the optimal path, contributing to improved communication efficiency and reliability.

[0046] As a preferred embodiment of the technical solution of the present invention, step S4 includes:

[0047] S41. Construct a joint resource allocation optimization problem:

[0048] Maximizing total network throughput is the optimization objective.

[0049] Where L is the set of all resource links to be allocated, including all hop links on the optimal relay path and other communication links; To be allocated to the link bandwidth; For link The signal-to-interference-plus-noise ratio; the optimization variable is each link in set L. Allocate bandwidth ;

[0050] The constraints include:

[0051] (a) Resource constraints: The total bandwidth allocated to all links cannot exceed the total available bandwidth of the system. ,Right now ;

[0052] (b) Flow conservation constraint: For any intermediate relay cluster head node on the optimal relay path, the sum of the rates of received data is equal to the sum of the rates of its forwarded data;

[0053] (c) QoS constraints: The bandwidth allocated to the optimal relay path must guarantee that the end-to-end throughput is not less than the minimum threshold required by its service. ;

[0054] S42. The optimization problem is transformed into a convex optimization problem, and a distributed solution is obtained using the Lagrange duality method to obtain the globally optimal bandwidth allocation scheme. ;

[0055] S43. Based on the bandwidth allocation scheme obtained from the solution. It generates specific resource allocation decisions, specifying the bandwidth to be allocated to each hop link h on the optimal relay path. .

[0056] With the goal of maximizing total network throughput, multiple constraints, including resource constraints, flow conservation constraints, and QoS constraints, were set to comprehensively consider all aspects of vehicular network resource allocation. The optimization problem was transformed into a convex optimization problem and solved in a distributed manner using the Lagrange duality method, resulting in a globally optimal bandwidth allocation scheme that effectively improves network resource utilization. Based on the solution results, specific resource allocation decisions were generated to rationally allocate bandwidth to each hop on the optimal relay path, ensuring communication quality and efficiency.

[0057] As a preferred embodiment of the technical solution of the present invention, step S5 includes:

[0058] S51. The roadside unit encapsulates the generated resource allocation decision into a resource allocation signaling message. The message contains a resource allocation list, each item in the list corresponds to a one-hop link on the optimal relay path, and records the link identifier, the allocated spectrum resource block index, and the effective time window. The roadside unit sends the signaling message to all cluster head nodes involved on the optimal relay path through the control channel.

[0059] S52. After receiving the resource allocation signaling message, each cluster head node parses out the resource allocation information of the transmission and reception links related to itself, and immediately reserves resources in the local communication protocol stack; after successful reservation, it sends a resource allocation confirmation message to the roadside unit.

[0060] S53. Each cluster head node runs a local scheduling algorithm based on the service queue status and service priority reported by its cluster member nodes to generate a secondary allocation scheme for cluster resources; the scheduling algorithm ensures that high-priority service data obtains transmission resources first.

[0061] S54. The cluster head node configures its media access control layer scheduler according to the intra-cluster resource secondary allocation scheme; at the same time, it notifies the corresponding member nodes of the specific resource block authorization through intra-cluster signaling to complete the local configuration.

[0062] This ensures the accuracy and timeliness of information transmission. Cluster head nodes parse resource allocation information and reserve resources in their local communication protocol stack. Upon successful reservation, they send an acknowledgment message, guaranteeing the effective execution of resource allocation. Each cluster head node performs secondary resource allocation using a local scheduling algorithm based on the service queue status and priority of member nodes within the cluster. This ensures priority transmission of high-priority service data, meeting the real-time and reliability requirements of different services in autonomous driving scenarios. Resource block authorization is communicated to the corresponding member nodes via intra-cluster signaling to complete local configuration, ensuring that resource allocation is truly implemented at each node and improving the actual effectiveness of resource allocation.

[0063] As a preferred embodiment of the technical solution of the present invention, step S6 includes:

[0064] S61. The roadside unit maintains a confirmation status table to record the resource allocation confirmation status of each cluster head node on the optimal relay path; the communication start process is triggered only when the confirmation status table shows that all nodes have been confirmed.

[0065] S62. The roadside unit generates a communication initiation command message, which includes a unified global communication start timestamp. Subsequently, the roadside unit first sends the start command to the source cluster head node, and simultaneously or sequentially sends synchronization signaling to all relay cluster head nodes on the path;

[0066] S63, the source cluster head node and each relay cluster head node, upon receiving the start or synchronization signaling, according to... Synchronize the local clock, and At the indicated time, the assigned transmit / receive link is activated, and the system enters the data forwarding ready state.

[0067] S64, at that moment Upon arrival, the source cluster head node begins sending data packets to the first-hop relay node on its allocated spectrum resources. The packet header contains a path identifier. Each relay node receives and forwards the packets based on the path identifier, thereby formally establishing an end-to-end multi-hop relay communication connection and initiating data transmission.

[0068] The communication startup process is triggered only after all relevant cluster head nodes confirm resource allocation, ensuring the synchronization and accuracy of communication startup. The generated communication startup command message contains a unified global communication start timestamp and sends startup or synchronization signaling to the source cluster head node and relay cluster head node, enabling each node to synchronize its local clock according to the timestamp and simultaneously activate the transmit / receive link at the specified time, entering the data forwarding ready state. The data packet header contains a path identifier, and each relay node receives and forwards data based on the identifier, formally establishing an end-to-end multi-hop relay communication connection and starting data transmission, ensuring smooth communication and accurate data transmission, and improving the reliability and stability of vehicle-to-everything (V2X) communication.

[0069] Secondly, the technical solution of the present invention also provides a vehicle network resource allocation system based on communication optimization in an autonomous driving scenario, including a central control module deployed on a roadside unit and an on-board communication control module deployed on each vehicle node;

[0070] The central control module includes:

[0071] The global topology management unit is used to receive and process the identity and cluster membership information reported by all cluster head nodes within its coverage area, in order to build and maintain global cluster topology information;

[0072] The path calculation unit is used to respond to the communication request of the source cluster head node, generate multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluate and select the one with the best transmission performance as the optimal relay path.

[0073] The resource allocation unit is used to construct and solve a joint resource allocation model with the goal of maximizing the total network throughput, and generate resource allocation decisions for each hop link on the optimal relay path.

[0074] The signaling scheduling unit is used to send the resource allocation decision to all cluster head nodes involved in the optimal relay path, and after receiving the configuration confirmation from all relevant cluster head nodes, send a communication start command to the source cluster head node.

[0075] The vehicle-mounted communication control module includes:

[0076] Clustering and cluster management units are used to execute weighted clustering algorithms, participate in cluster head election and cluster formation, and manage the relationships between members within a cluster;

[0077] The local resource scheduling unit is used to perform secondary allocation and local configuration of resources based on the priority of services of nodes within the cluster, according to the received resource allocation decision.

[0078] The data routing and forwarding unit is used to perform multi-hop relay transmission of data along the optimal relay path after receiving the communication start command.

[0079] As a preferred embodiment of the technical solution of the present invention, the path calculation unit is specifically used for:

[0080] Using the maintained global cluster topology graph as the network graph, each cluster head node in the network graph is a vertex, and there is an edge between cluster head nodes if they can communicate directly; using the K shortest path algorithm, K shortest paths from the source cluster head node to the target node are calculated in the network graph, and a candidate path set is generated; where K is a natural number greater than 1.

[0081] For each candidate path in the candidate path set, calculate the end-to-end transmission performance value P of the path, using the following formula:

[0082]

[0083] Where H is the total number of hops for the path. Let be the spectral efficiency of the h-th hop link on the path. , The expected lifetime of the h-th hop link on the path is calculated based on the position and velocity vectors of the adjacent cluster head nodes. and These are the normalized weighting coefficients, and + = 1; The channel quality indication value reported by the h-th hop link; These are calibration coefficients preset according to the mapping relationship between modulation and coding strategies;

[0084] Sort all candidate paths in the candidate path set according to their transmission performance values.

[0085] As a preferred embodiment of the technical solution of the present invention, the resource allocation signaling message issued by the signaling scheduling unit includes a spectrum resource block index and an effective time window; the local resource scheduling unit of the vehicle communication control module reserves resources in the local communication protocol stack accordingly.

[0086] As can be seen from the above technical solutions, this application has the following advantages: by periodically broadcasting status information by vehicle nodes and dynamically forming clusters using a weighted clustering algorithm, roadside units construct and maintain global cluster topology information, generate multiple candidate multi-hop relay paths based on global topology and channel information and evaluate transmission efficiency, construct a joint resource allocation model to solve after selecting the optimal path, generate resource allocation decisions and perform secondary allocation and local configuration, and finally establish a multi-hop relay communication connection, thereby improving the stability, reliability and efficiency of vehicle network communication. Attached Figure Description

[0087] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0088] Figure 1 This is a flowchart illustrating the method provided in an embodiment of the present invention.

[0089] Figure 2 A block diagram of a system provided in an embodiment of the present invention. Detailed Implementation

[0090] To make the purpose, features, and advantages of this application more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solution protected by this application. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0091] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0092] This invention takes a highway autonomous driving scenario as its background. In this scenario, the roadside unit (RSU) has a coverage range of 1000 meters, and there are 30 autonomous vehicles (node ​​numbers V1-V30) within the coverage area. The vehicle speed is between 60-120 km / h, the total available bandwidth of the system is 100MHz, and the minimum end-to-end throughput threshold of the service requirements is... It is 5Mbps. For example... Figure 1 As shown, this embodiment of the invention provides a vehicle-to-everything (V2X) resource allocation method based on communication optimization in an autonomous driving scenario, comprising the following steps:

[0093] S1. Within the coverage area of ​​the same roadside unit, each vehicle node periodically broadcasts its own status information and dynamically forms multiple clusters based on the status information of its neighboring nodes using a weighted clustering algorithm. Each cluster elects a cluster head node. Each cluster head node reports its own identity and cluster member information to the roadside unit. The roadside unit constructs and maintains global cluster topology information based on the information reported by each cluster head node.

[0094] Each vehicle node broadcasts its own status information via Dedicated Short Range Communication (DSRC) technology at 100ms intervals. This information includes the node number, current location (latitude and longitude, e.g., V1: 30°12′N, 120°08′E), speed (e.g., V1: 80km / h), direction of travel (e.g., V1: east to west), remaining energy (e.g., V1: 85%), and channel quality indicator (CQI, e.g., V1: 7). Simultaneously, each vehicle node receives status information broadcast by its neighbors (other vehicles within communication range) and stores it in its local neighbor information list.

[0095] In this step, multiple clusters are dynamically formed using a weighted clustering algorithm, and a cluster head node is elected for each cluster. The steps include:

[0096] S11. Each vehicle node calculates its own weight value to become the cluster head based on the received state information of its neighboring nodes.

[0097] Each vehicle node calculates its own weight value to become the cluster head based on the received state information of its neighboring nodes. The weight value calculation comprehensively considers factors such as the node's remaining energy, the average distance to neighboring nodes, and channel quality. The calculation formula is as follows:

[0098] in E The remaining energy of the vehicle node (normalized to between 0 and 1). This represents the average distance (in km) between a node and all its neighboring nodes. This is a channel quality indicator value (normalized to between 0 and 1). For example, the remaining energy of node V1. E =0.85, with distances to its 5 neighboring nodes of 0.1km, 0.15km, 0.2km, 0.08km, and 0.12km respectively, and an average distance of... =0.13 km , ≈7.69 (0.769 after normalization). After normalization, the value is 0.7. Therefore, the weight value of V1 is:

[0099] =0.4×0.85+0.3×0.769+0.3×0.7≈0.77. The weight threshold is set to 0.6.

[0100] S12. Determine if there are any vehicle nodes with weight values ​​less than a set threshold:

[0101] If it exists, then execute the first cluster formation process:

[0102] S12a1. Select the vehicle node with the smallest weight value as the cluster head node;

[0103] S12a2, Only neighboring nodes that meet preset conditions are allowed to join the cluster where the cluster head node is located as cluster member nodes; the preset conditions include at least one of the following: the distance between the neighboring node and the cluster head node is less than a first preset distance threshold, the channel signal-to-noise ratio between the neighboring node and the cluster head node is greater than a first preset signal-to-noise ratio threshold, the remaining energy of the neighboring node is greater than a first preset energy threshold, and the angle between the driving direction of the neighboring node and the cluster head node is less than a preset angle threshold; that is, in this application, cluster members are selected from vehicles traveling in the same direction or with similar driving directions.

[0104] S12a3. For the remaining vehicle nodes that have not joined the cluster, repeat the weight calculation and cluster head election among the remaining nodes until all vehicle nodes are included in a cluster.

[0105] If the weight values ​​of all vehicle nodes are greater than or equal to the set threshold, execute the second cluster formation process:

[0106] S12b1. Select N vehicle nodes with the smallest relative weight values ​​as the cluster head nodes of different clusters;

[0107] S12b2. For the remaining vehicle nodes that are not selected as cluster head nodes, select and join a cluster according to the preset joining rules.

[0108] Calculations show that 25 out of the 30 vehicle nodes have a weight value less than 0.6, and 5 nodes have a weight value greater than or equal to 0.6. Therefore, the first cluster formation process (S12a1-S12a3) is executed.

[0109] Initial cluster head and cluster member selection (S12a1-S12a2): Among nodes with a weight value less than 0.6, select the V5 node with the smallest weight value (weight value of 0.42) as the first cluster head node. Preset conditions are: distance to the cluster head node less than 0.2km (first preset distance threshold), channel signal-to-noise ratio (SNR) greater than 15dB (first preset SNR threshold), and remaining energy greater than 60% (first preset energy threshold). Among the neighboring nodes of node V5, V3 (distance 0.12km, SNR 18dB, remaining energy 72%), V4 (distance 0.15km, SNR 16dB, remaining energy 68%), V6 (distance 0.18km, SNR 17dB, remaining energy 75%), and V7 (distance 0.1km, SNR 20dB, remaining energy 65%) meet the preset conditions and are allowed to join the cluster where V5 is located, forming cluster 1 (cluster head V5, members V3, V4, V6, V7).

[0110] Remaining node cluster formation (S12a3): For the 25 remaining vehicle nodes that have not joined cluster 1, recalculate the weight value of each node, select the node with the smallest weight value (V12, weight value 0.45) as the new cluster head, and select cluster members according to the same preset conditions to form cluster 2. Repeat this process until all 30 vehicle nodes are included in a cluster, finally forming 5 clusters (cluster 1-cluster 5), with the cluster heads of each cluster being V5, V12, V18, V23, and V28, respectively.

[0111] Each cluster head node (V5, V12, V18, V23, V28) reports its own identity (node ​​number, cluster head identifier) ​​and cluster member information (list of member node numbers) to the roadside unit. Based on the information reported by each cluster head node, the roadside unit constructs a global cluster topology graph. Each cluster head node in the graph is a vertex, and vertex attributes include the cluster head node number, the number of members within the cluster, and the cluster coverage area. If two cluster head nodes can communicate directly (through channel detection, a communication signal-to-noise ratio greater than 12dB is considered as direct communication capability), an edge is established between the corresponding vertices. Edge attributes include communication distance and channel signal-to-noise ratio. The roadside unit maintains this global cluster topology information in real time, updating the topology graph promptly when there are changes in cluster members or significant changes in the location or status of cluster head nodes (e.g., a location change exceeding 50 meters or remaining energy below 30%).

[0112] S2. When a cluster head node needs to communicate with a target node, it sends a communication request to the roadside unit. The roadside unit generates multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluates the end-to-end transmission performance of each path. The transmission performance is a weighted function of the link capacity and link stability of each hop of the relay path. The cluster head node that sends the communication request to the roadside unit is defined as the source cluster head node.

[0113] Suppose that a vehicle node in cluster 3 (cluster head V18) needs to transmit high-definition map data for autonomous driving to cluster 5 (cluster head V28) (target node is V28). Cluster head V18, as the source cluster head node, sends a communication request to the roadside unit. The request includes information such as the source cluster head node number V18, the target node number V28, the service type (high-definition map data transmission), the service bandwidth requirement (8Mbps), and the service priority (high).

[0114] This step specifically includes:

[0115] S21. The roadside unit uses the maintained global cluster topology graph as the network graph. Each cluster head node in the network graph is a vertex. If cluster head nodes can communicate directly, there is an edge between them. Using the K shortest path algorithm, K shortest paths from the source cluster head node to the target node are calculated in the network graph to generate a candidate path set. K is a natural number greater than 1.

[0116] The roadside unit uses the maintained global cluster topology graph as the network graph and uses the K-shortest path algorithm (setting K=3) to calculate the three shortest paths from the source cluster head node V18 to the target node V28, generating a candidate path set.

[0117] Path 1: V18→V23→V28 (2 jumps)

[0118] Path 2: V18→V12→V23→V28 (3 jumps)

[0119] Path 3: V18→V5→V12→V23→V28 (4 jumps).

[0120] S22. For each candidate path in the candidate path set, calculate the end-to-end transmission performance value P. The calculation formula is as follows:

[0121]

[0122] Where H is the total number of hops for the path. Let be the spectral efficiency of the h-th hop link on the path. , The expected lifetime of the h-th hop link on the path is calculated based on the position and velocity vectors of the adjacent cluster head nodes. and These are the normalized weighting coefficients, and + = 1; The channel quality indication value reported by the h-th hop link; These are calibration coefficients preset according to the mapping relationship between modulation and coding strategies;

[0123] S23. Sort all candidate paths in the candidate path set according to the transmission performance value.

[0124] Assume the h-th hop link connects two adjacent cluster head nodes. A and B ,node A The position coordinates are ( ), the velocity vector is ( ),in For speed magnitude, The direction angle of motion; node B The position coordinates are ( ), the velocity vector is ( ).

[0125] Calculate the relative velocity vector of node A with respect to node B. ):

[0126]

[0127]

[0128] compute nodes A With nodes B The relative position vector ( ):

[0129]

[0130]

[0131] Link Expected Lifetime The calculation formula is:

[0132]

[0133] in, For nodes A With nodes B The maximum distance at which effective communication can be maintained between them.

[0134] For each candidate path, obtain the relevant parameters of each hop link on the path and calculate the departure transmission performance value for each path. Let's assume that are the transmission performance values ​​for path 1. The transmission efficiency value for path 2 is 52.012. The transmission efficiency value for path 3 is 74.416. The value is 98.636. Based on the transmission performance value, all candidate paths in the candidate path set are sorted, and the sorting result is: Path 3 > Path 2 > Path 1.

[0135] S3. Select the relay path with the best transmission efficiency as the optimal relay path;

[0136] Here, the path 3 (V18→V5→V12→V23→V28) with the best transmission performance is selected as the optimal relay path.

[0137] S4. The roadside unit, aiming to maximize the total network throughput, constructs and solves a joint resource allocation model to generate resource allocation decisions for each hop link on the optimal relay path; the steps of S4 include:

[0138] S41. Construct a joint resource allocation optimization problem:

[0139] Maximizing total network throughput is the optimization objective.

[0140] Where L is the set of all resource links to be allocated, including the 4-hop links on the optimal relay path (V18→V5, V5→V12, V12→V23, V23→V28) and the other 5 links that are currently communicating (such as the internal communication links V5→V3, V5→V4, etc. within cluster 1). To be allocated to the link bandwidth; For link The signal-to-interference-plus-noise ratio (SIR) is obtained through channel measurements; for example, the SIR of link V18→V5 is 25. dB The optimization variable is each link in set L. Allocate bandwidth ;

[0141] The constraints include:

[0142] (a) Resource constraints: The total bandwidth allocated to all links cannot exceed the total available bandwidth of the system. ,Right now Total available bandwidth Here it is 100MHz.

[0143] (b) Flow conservation constraint: For any intermediate relay cluster head node on the optimal relay path, the sum of the rates of received data is equal to the sum of the rates of its forwarded data;

[0144] Taking V5 as an example, the receiving rate (V18→V5 link rate, ) equals the forwarding rate (V5→V12 link rate, ).

[0145] (c) QoS constraints: The bandwidth allocated to the optimal relay path must guarantee that the end-to-end throughput is not less than the minimum threshold required by its service. =5MHz;

[0146] S42. The optimization problem is transformed into a convex optimization problem, and a distributed solution is obtained using the Lagrange duality method to obtain the globally optimal bandwidth allocation scheme. ;

[0147] Constructing the Lagrange function: ,in, Let be the Lagrange multiplier. Taking the partial derivative of the Lagrange function and setting it equal to 0 yields the necessary condition for optimal bandwidth allocation:

[0148] ,Right now

[0149] Based on the above conditions and constraints, the globally optimal bandwidth allocation scheme is finally obtained by iteratively calculating and adjusting the Lagrange multiplier λ. .

[0150] For example, the bandwidth allocation result for each hop link on the optimal relay path is as follows: =12MHz, =10MHz, =11MHz, =9MHz, and the remaining bandwidth for other links is allocated according to the signal-to-interference-plus-noise ratio and service requirements.

[0151] S43. Based on the bandwidth allocation scheme obtained from the solution. It generates specific resource allocation decisions, specifying the bandwidth to be allocated to each hop link h on the optimal relay path. .

[0152] In this step, the roadside unit creates a resource allocation instruction message for the optimal relay path; the message contains at least a resource allocation information list, and each item in the list explicitly records the following information: link identifier, used to uniquely identify the h-th hop link on the optimal relay path;

[0153] allocated bandwidth value ;

[0154] The start time and duration of the assignment.

[0155] S5. The roadside unit distributes the resource allocation decision to all cluster head nodes involved in the optimal relay path; each cluster head node receives and confirms the resource allocation decision, and performs secondary allocation and local configuration of the acquired resources according to the priority of the services of its cluster nodes; this step specifically includes:

[0156] S51. The roadside unit encapsulates the generated resource allocation decision into a resource allocation signaling message. The message contains a resource allocation list, each item in the list corresponds to a one-hop link on the optimal relay path, and records the link identifier, the allocated spectrum resource block index, and the effective time window. The roadside unit sends the signaling message to all cluster head nodes involved on the optimal relay path through the control channel.

[0157] S52. After receiving the resource allocation signaling message, each cluster head node parses out the resource allocation information of the transmission and reception links related to itself, and immediately reserves resources in the local communication protocol stack; after successful reservation, it sends a resource allocation confirmation message to the roadside unit.

[0158] S53. Each cluster head node runs a local scheduling algorithm based on the service queue status and service priority reported by its cluster member nodes to generate a secondary allocation scheme for cluster resources; the scheduling algorithm ensures that high-priority service data obtains transmission resources first.

[0159] S54. The cluster head node configures its media access control layer scheduler according to the intra-cluster resource secondary allocation scheme; at the same time, it notifies the corresponding member nodes of the specific resource block authorization through intra-cluster signaling to complete the local configuration.

[0160] Each cluster head node maintains an intra-cluster service information table, recording the priority weight of the service to which the service data packets to be transmitted belong for each member node. ;

[0161] The cluster head node uses a weighted fair queue scheduling algorithm for secondary resource allocation; for the current scheduling slot, the service flow of each member node is... Calculate a virtual completion time The update rules are as follows:

[0162]

[0163] in, This is the current system virtual time. The length of the current data packet to be scheduled. The total bandwidth resources allocated to this cluster by the cluster head node from the roadside unit; priority is given to scheduling the node with the shortest virtual completion time. Data packets of the business flow;

[0164] The cluster head node translates resource allocation decisions into specific resource block allocation schemes and sends them to the corresponding member nodes through the intra-cluster control channel; member nodes send data to the cluster head node on the specified resources;

[0165] For data that needs to be forwarded, the cluster head node performs the following:

[0166] If this node is the source cluster head node, it will encapsulate and send the raw data from members within the cluster or the data it generates, according to the next-hop address of the optimal relay path.

[0167] If this node is a relay cluster head node, it receives the data from the previous hop and forwards the data to the next hop node according to the forwarding link bandwidth specified in the resource allocation decision.

[0168] If this node is the last cluster head node on the path, then the data will be delivered to the target node.

[0169] S6. After receiving configuration confirmations from all relevant cluster head nodes, the roadside unit sends a communication start command to the source cluster head node, thereby establishing a multi-hop relay communication connection from the source cluster head node to the target node along the optimal relay path. This step specifically includes:

[0170] S61. The roadside unit maintains a confirmation status table to record the resource allocation confirmation status of each cluster head node on the optimal relay path; the communication start process is triggered only when the confirmation status table shows that all nodes have been confirmed.

[0171] S62. The roadside unit generates a communication initiation command message, which includes a unified global communication start timestamp. Subsequently, the roadside unit first sends the start command to the source cluster head node, and simultaneously or sequentially sends synchronization signaling to all relay cluster head nodes on the path;

[0172] S63, the source cluster head node and each relay cluster head node, upon receiving the start or synchronization signaling, according to... Synchronize the local clock, and At the indicated time, the assigned transmit / receive link is activated, and the system enters the data forwarding ready state.

[0173] S64, at that moment Upon arrival, the source cluster head node begins sending data packets to the first-hop relay node on its allocated spectrum resources. The packet header contains a path identifier. Each relay node receives and forwards the packets based on the path identifier, thereby formally establishing an end-to-end multi-hop relay communication connection and initiating data transmission.

[0174] In this embodiment of the invention, the preset joining rules include at least one of the following: joining by selecting the nearest cluster head node, joining by selecting the cluster head node with the best channel quality, or joining by selecting the cluster head node with the highest remaining energy.

[0175] In some embodiments, the step of adding the remaining vehicle nodes that were not selected as cluster head nodes to the cluster includes:

[0176] For each remaining vehicle node that was not selected as a cluster head node, calculate its joining utility value with all selected cluster head nodes;

[0177] The added utility value The calculation formula is:

[0178]

[0179] in, For vehicle nodes With cluster head node distance, For vehicle nodes With cluster head node The signal-to-noise ratio between them Cluster head node The remaining energy, These are the normalized weighting coefficients;

[0180] Selection of each remaining vehicle node that was not selected as a cluster head node Find the cluster head node with the largest value and send a join request to it;

[0181] The cluster head node receives a join request. If the requesting node meets the preset conditions, it approves the join and updates the cluster member list; if the requesting node does not meet the preset conditions, it rejects the join and the requesting node then requests a different cluster member. The cluster head node with the second largest value sends a join request until it successfully joins a cluster or all candidate cluster head nodes fail to join.

[0182] In some embodiments, the method further includes:

[0183] If a vehicle node sends a join request to all cluster head nodes and is rejected by all of them, the vehicle node will bootstrap itself into an isolated cluster head node and broadcast that it has become the new cluster head node; the remaining isolated vehicle nodes that are also unable to join any cluster will, upon detecting the announcement, attempt to join the cluster formed by the new cluster head node according to the preset join rules or join utility value.

[0184] After the roadside unit detects a new cluster formed by isolated vehicle nodes in the network, it incorporates the new cluster into the global cluster topology management and treats it as an ordinary cluster for resource allocation in the subsequent resource allocation process.

[0185] The specific functions of the roadside unit in this embodiment of the invention include:

[0186] Coverage and Access Point: As a cellular network (such as 5G / 6G) base station or dedicated short-range communication site, it provides network access services to vehicles within its wireless signal coverage area.

[0187] The global information aggregation center receives and processes the periodic status information (such as location, speed, cluster member list, channel quality, etc.) reported by all vehicle nodes (especially cluster head nodes) within its coverage area, thereby constructing and maintaining a real-time global network view. This is the data foundation for achieving centralized optimization.

[0188] Centralized decision-making brain: Based on the aforementioned global network view, it executes the core algorithm flow of this invention. This includes:

[0189] Path calculation: In response to the communication request from the source cluster head, calculate and evaluate multiple candidate relay paths to the target node.

[0190] Resource allocation: With the goal of maximizing system efficiency, a joint resource allocation optimization model is solved to allocate appropriate spectrum resources to each hop link on the optimal path.

[0191] Control command issuer: Reliably sends the calculated resource allocation decisions, communication initiation commands, and other control signaling to the relevant vehicle nodes (cluster head nodes) to coordinate the communication behavior of the entire area.

[0192] like Figure 2 As shown, this embodiment of the invention also provides a vehicle network resource allocation system based on communication optimization in an autonomous driving scenario, including a central control module deployed on a roadside unit and an on-board communication control module deployed on each vehicle node;

[0193] The central control module includes:

[0194] The global topology management unit is used to receive and process the identity and cluster membership information reported by all cluster head nodes within its coverage area, in order to build and maintain global cluster topology information;

[0195] The path calculation unit is used to respond to the communication request of the source cluster head node, generate multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluate and select the one with the best transmission performance as the optimal relay path.

[0196] The resource allocation unit is used to construct and solve a joint resource allocation model with the goal of maximizing the total network throughput, and generate resource allocation decisions for each hop link on the optimal relay path.

[0197] The signaling scheduling unit is used to send the resource allocation decision to all cluster head nodes involved in the optimal relay path, and after receiving the configuration confirmation from all relevant cluster head nodes, send a communication start command to the source cluster head node.

[0198] The vehicle-mounted communication control module includes:

[0199] Clustering and cluster management units are used to execute weighted clustering algorithms, participate in cluster head election and cluster formation, and manage the relationships between members within a cluster;

[0200] The local resource scheduling unit is used to perform secondary allocation and local configuration of resources based on the priority of services of nodes within the cluster, according to the received resource allocation decision.

[0201] The data routing and forwarding unit is used to perform multi-hop relay transmission of data along the optimal relay path after receiving the communication start command.

[0202] In some embodiments, the path calculation unit is specifically used for:

[0203] Using the maintained global cluster topology graph as the network graph, each cluster head node in the network graph is a vertex, and there is an edge between cluster head nodes if they can communicate directly; using the K shortest path algorithm, K shortest paths from the source cluster head node to the target node are calculated in the network graph, and a candidate path set is generated; where K is a natural number greater than 1.

[0204] For each candidate path in the candidate path set, calculate the end-to-end transmission performance value P of the path, using the following formula:

[0205]

[0206] Where H is the total number of hops for the path. Let be the spectral efficiency of the h-th hop link on the path. , The expected lifetime of the h-th hop link on the path is calculated based on the position and velocity vectors of the adjacent cluster head nodes. and These are the normalized weighting coefficients, and + = 1; The channel quality indication value reported by the h-th hop link; These are calibration coefficients preset according to the mapping relationship between modulation and coding strategies;

[0207] Sort all candidate paths in the candidate path set according to their transmission performance values.

[0208] The resource allocation signaling message issued by the signaling scheduling unit includes a spectrum resource block index and an effective time window; the local resource scheduling unit of the vehicle communication control module reserves resources in the local communication protocol stack accordingly.

[0209] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for allocating vehicle network resources based on communication optimization in an autonomous driving scenario, characterized in that, Includes the following steps: S1. Within the coverage area of ​​the same roadside unit, each vehicle node periodically broadcasts its own status information and dynamically forms multiple clusters based on the status information of its neighboring nodes using a weighted clustering algorithm. Each cluster elects a cluster head node. Each cluster head node reports its own identity and cluster member information to the roadside unit. The roadside unit constructs and maintains global cluster topology information based on the information reported by each cluster head node. S2. When a cluster head node needs to communicate with a target node, it sends a communication request to the roadside unit. The roadside unit generates multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluates the end-to-end transmission performance of each path. The transmission performance is a weighted function of the link capacity and link stability of each hop in the relay path. The cluster head node that sends the communication request to the roadside unit is defined as the source cluster head node. Specifically, this includes: S21. The roadside unit uses the maintained global cluster topology graph as the network graph. Each cluster head node in the network graph is a vertex. If cluster head nodes can communicate directly, there is an edge between them. Using the K shortest path algorithm, K shortest paths from the source cluster head node to the target node are calculated in the network graph to generate a candidate path set. K is a natural number greater than 1. S22. For each candidate path in the candidate path set, calculate the end-to-end transmission performance value P. The calculation formula is as follows: Where H is the total number of hops for the path. Let be the spectral efficiency of the h-th hop link on the path. , The expected lifetime of the h-th hop link on the path is calculated based on the position and velocity vectors of the adjacent cluster head nodes. and These are the normalized weighting coefficients, and + = 1; The channel quality indication value reported by the h-th hop link; These are calibration coefficients preset according to the mapping relationship between modulation and coding strategies; S23. Sort all candidate paths in the candidate path set according to their transmission performance values; S3. Select the relay path with the best transmission efficiency as the optimal relay path; S4. The roadside unit constructs and solves a joint resource allocation model with the goal of maximizing the total network throughput, and generates resource allocation decisions for each hop link on the optimal relay path. S5. The roadside unit sends the resource allocation decision to all cluster head nodes involved in the optimal relay path; each cluster head node receives and confirms the resource allocation decision, and performs secondary allocation and local configuration of the obtained resources according to the priority of the services of its cluster nodes. S6. After receiving configuration confirmation from all relevant cluster head nodes, the roadside unit sends a communication start command to the source cluster head node, thereby establishing a multi-hop relay communication connection from the source cluster head node to the target node along the optimal relay path.

2. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 1, characterized in that, In S1, a weighted clustering algorithm is used to dynamically form multiple clusters. The steps for electing a cluster head node for each cluster include: S11. Each vehicle node calculates its own weight value to become the cluster head based on the received state information of its neighboring nodes. S12. Determine if there are any vehicle nodes with weight values ​​less than a set threshold: If it exists, then execute the first cluster formation process: S12a1. Select the vehicle node with the smallest weight value as the cluster head node; S12a2, Only neighboring nodes that meet preset conditions are allowed to join the cluster where the cluster head node is located as cluster member nodes; the preset conditions include at least one of the following: the distance between the neighboring node and the cluster head node is less than a first preset distance threshold, the channel signal-to-noise ratio between the neighboring node and the cluster head node is greater than a first preset signal-to-noise ratio threshold, the remaining energy of the neighboring node is greater than a first preset energy threshold, and the angle between the travel directions of the neighboring node and the cluster head node is less than a preset angle threshold. S12a3. For the remaining vehicle nodes that have not joined the cluster, repeat the weight calculation and cluster head election among the remaining nodes until all vehicle nodes are included in a cluster. If the weight values ​​of all vehicle nodes are greater than or equal to the set threshold, execute the second cluster formation process: S12b1. Select N vehicle nodes with the smallest relative weight values ​​as the cluster head nodes of different clusters; S12b2. For the remaining vehicle nodes that are not selected as cluster head nodes, select and join a cluster according to the preset joining rules.

3. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 2, characterized in that, The preset joining rules include at least one of the following: joining by selecting the nearest cluster head node, joining by selecting the cluster head node with the best channel quality, or joining by selecting the cluster head node with the highest remaining energy.

4. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 3, characterized in that, The steps for adding the remaining vehicle nodes that were not selected as cluster head nodes to the cluster include: For each remaining vehicle node that was not selected as a cluster head node, calculate its joining utility value with all selected cluster head nodes; The added utility value The calculation formula is: in, For vehicle nodes With cluster head node distance, For vehicle nodes With cluster head node The signal-to-noise ratio between them Cluster head node The remaining energy, These are the normalized weighting coefficients; Selection of each remaining vehicle node that was not selected as a cluster head node Find the cluster head node with the largest value and send a join request to it; The cluster head node receives a join request. If the requesting node meets preset conditions, it approves its joining and updates the cluster member list; if the requesting node does not meet the preset conditions, it rejects its joining and the requesting node then requests a different cluster member. The cluster head node with the second largest value sends a join request until it successfully joins a cluster or all candidate cluster head nodes fail to join.

5. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 4, characterized in that, The method also includes: If a vehicle node sends a join request to all cluster head nodes and is rejected by all of them, the vehicle node will bootstrap itself into an isolated cluster head node and broadcast that it has become the new cluster head node; the remaining isolated vehicle nodes that are also unable to join any cluster will, upon detecting the announcement, attempt to join the cluster formed by the new cluster head node according to the preset join rules or join utility value. After the roadside unit detects a new cluster formed by isolated vehicle nodes in the network, it incorporates the new cluster into the global cluster topology management and treats it as an ordinary cluster for resource allocation in the subsequent resource allocation process.

6. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 5, characterized in that, The steps in S4 include: S41. Construct a joint resource allocation optimization problem: Maximizing total network throughput is the optimization objective. Where L is the set of all resource links to be allocated, including all hop links on the optimal relay path and other communication links; To be allocated to the link bandwidth; For link The signal-to-interference-to-noise ratio; the optimization variable is each link in set L. Allocate bandwidth ; The constraints include: (a) Resource constraints: The total bandwidth allocated to all links cannot exceed the total available bandwidth of the system. ,Right now ; (b) Flow conservation constraint: For any intermediate relay cluster head node on the optimal relay path, the sum of the rates of received data is equal to the sum of the rates of its forwarded data; (c) QoS constraints: The bandwidth allocated to the optimal relay path must guarantee that the end-to-end throughput is not less than the minimum threshold required by its service. ; S42. The optimization problem is transformed into a convex optimization problem, and a distributed solution is obtained using the Lagrange duality method to obtain the globally optimal bandwidth allocation scheme. ; S43. Based on the bandwidth allocation scheme obtained from the solution. It generates specific resource allocation decisions, specifying the bandwidth to be allocated to each hop link h on the optimal relay path. .

7. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 6, characterized in that, The steps in S5 include: S51. The roadside unit encapsulates the generated resource allocation decision into a resource allocation signaling message. The message contains a resource allocation list, each item in the list corresponds to a one-hop link on the optimal relay path, and records the link identifier, the allocated spectrum resource block index, and the effective time window. The roadside unit sends the signaling message to all cluster head nodes involved on the optimal relay path through the control channel. S52. After receiving the resource allocation signaling message, each cluster head node parses out the resource allocation information of the transmission and reception links related to itself, and immediately reserves resources in the local communication protocol stack; after successful reservation, it sends a resource allocation confirmation message to the roadside unit. S53. Each cluster head node runs a local scheduling algorithm based on the service queue status and service priority reported by its cluster member nodes to generate a secondary allocation scheme for cluster resources; the scheduling algorithm ensures that high-priority service data obtains transmission resources first. S54. The cluster head node configures its media access control layer scheduler according to the intra-cluster resource secondary allocation scheme; at the same time, it notifies the corresponding member nodes of the specific resource block authorization through intra-cluster signaling to complete the local configuration.

8. The vehicle network resource allocation method based on communication optimization in autonomous driving scenarios according to claim 7, characterized in that, The steps in S6 include: S61. The roadside unit maintains a confirmation status table to record the resource allocation confirmation status of each cluster head node on the optimal relay path; the communication start process is triggered only when the confirmation status table shows that all nodes have been confirmed. S62. The roadside unit generates a communication initiation command message, which includes a unified global communication start timestamp. Subsequently, the roadside unit first sends the start command to the source cluster head node, and simultaneously or sequentially sends synchronization signaling to all relay cluster head nodes on the path; S63, the source cluster head node and each relay cluster head node, upon receiving the start or synchronization signaling, according to... Synchronize the local clock, and At the indicated time, the assigned transmit / receive link is activated, and the system enters the data forwarding ready state. S64, at that moment Upon arrival, the source cluster head node begins sending data packets to the first-hop relay node on its allocated spectrum resources. The packet header contains a path identifier. Each relay node receives and forwards the packets based on the path identifier, thereby formally establishing an end-to-end multi-hop relay communication connection and initiating data transmission.

9. A vehicle-to-everything (V2X) resource allocation system based on communication optimization for implementing the method described in any one of claims 1-8 in an autonomous driving scenario, characterized in that, This includes a central control module deployed on the roadside unit and an onboard communication control module deployed on each vehicle node; The central control module includes: The global topology management unit is used to receive and process the identity and cluster membership information reported by all cluster head nodes within its coverage area, in order to build and maintain global cluster topology information; The path calculation unit is used to respond to the communication request of the source cluster head node, generate multiple candidate multi-hop relay paths from the source cluster head node to the target node based on the global cluster topology and channel information, and evaluate and select the one with the best transmission performance as the optimal relay path. The resource allocation unit is used to construct and solve a joint resource allocation model with the goal of maximizing the total network throughput, and to generate resource allocation decisions for each hop link on the optimal relay path. The signaling scheduling unit is used to send the resource allocation decision to all cluster head nodes involved in the optimal relay path, and after receiving the configuration confirmation from all relevant cluster head nodes, send a communication start command to the source cluster head node. The vehicle-mounted communication control module includes: The clustering and cluster management unit is used to execute weighted clustering algorithms, participate in cluster head election and cluster formation, and manage the relationships between members within a cluster; The local resource scheduling unit is used to perform secondary allocation and local configuration of resources based on the priority of services of nodes within the cluster, according to the received resource allocation decision. The data routing and forwarding unit is used to perform multi-hop relay transmission of data along the optimal relay path after receiving the communication start command.

Images