A method and system for cross-layer optimization of resource scheduling and multi-hop forwarding in a vehicle cluster

Abstract

The present application relates to the technical field of wireless communication, and especially relates to a kind of vehicle cluster resource scheduling and multi-hop forwarding cross-layer optimization method and system, comprising: cluster member periodically senses wireless channel resource and reports resource occupation information to cluster head;In multiple vehicle cluster networks, each cluster contains cluster head and multiple members that meet preset similarity condition;Cluster head is based on the reporting result, assesses resource availability by distance weighted average calculation and dynamically allocates transmission resource;Combined with resource allocation state and topological feature, multi-dimensional decision model is used to select intra-cluster or inter-cluster forwarding node to realize multi-hop message forwarding.The present application significantly reduces the resource conflict rate of wireless channel through intra-cluster collaborative resource sensing and dynamic distance weighted evaluation mechanism, and optimizes the selection of forwarding node through multi-dimensional decision model combined with topological feature, which simultaneously improves the multi-hop transmission efficiency and communication reliability, thereby guaranteeing the low-latency demand of safety-related services.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a method and system for cross-layer optimization of resource scheduling and multi-hop forwarding within a vehicle cluster. Background Technology

[0002] Traditional multi-hop communication solutions for vehicle-to-everything (V2X) rely on base station forwarding or the IEEE WAVE standard resource scheduling mechanism, which suffers from high latency in secure service transmission and frequent resource conflicts. In particular, in vehicle clustering architectures, cluster heads lack the ability to dynamically and collaboratively analyze the perception results periodically reported by cluster members. They cannot accurately assess the availability of resources within a cluster through distance-weighted average calculations, resulting in a disconnect between resource allocation and the selection process of multi-hop forwarding nodes. This makes it difficult to meet the low-latency, high-reliability security service requirements in C-V2X scenarios and limits the overall improvement of V2X communication efficiency.

[0003] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and system for optimizing resource scheduling and multi-hop forwarding across layers within a vehicle cluster.

[0005] Firstly, the present invention provides a method for cross-layer optimization of resource scheduling and multi-hop forwarding within a vehicle cluster, the technical solution of which is as follows:

[0006] The cluster member nodes periodically sense wireless channel resources, obtain sensing results containing resource occupancy information, and report the sensing results to the cluster head; in the network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members.

[0007] Based on the perception results reported by all cluster members, the cluster head collaboratively assesses the availability of resources within the cluster through distance-weighted average calculation, obtains the assessment results, and dynamically allocates transmission resources to cluster members according to the assessment results;

[0008] The cluster head combines the resource allocation status and the topological characteristics of the cluster member nodes, and selects intra-cluster forwarding nodes or inter-cluster forwarding nodes through a pre-built multi-dimensional decision model to achieve multi-hop message forwarding.

[0009] The beneficial effects of the vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization method of the present invention are as follows:

[0010] The method of this invention significantly reduces the resource conflict rate of wireless channels through intra-cluster collaborative resource awareness and dynamic distance weighted evaluation mechanism. By optimizing the selection of forwarding nodes by combining a multi-dimensional decision model with topological features, it simultaneously improves multi-hop transmission efficiency and communication reliability, thereby ensuring the low latency requirements of security services.

[0011] Based on the above scheme, the vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization method of the present invention can be further improved as follows.

[0012] In one alternative approach, the step of obtaining a sensing result containing resource occupancy information by periodically sensing wireless channel resources through cluster member nodes includes:

[0013] Cluster member nodes perceive the occupancy status of resource blocks in the wireless channel and record resource occupancy information including the subframe number, subchannel number, and corresponding reference signal received power of the occupied resource block.

[0014] The recorded resource occupancy information is encapsulated into a sensing result, which includes: cluster ID, cluster member ID, resource-aware subframe, resource-aware subchannel, and reference signal received power.

[0015] Among the above-mentioned optional methods, the structured reporting of perception results is achieved by standardizing and encapsulating resource occupancy information, providing a unified data foundation for cluster head collaborative evaluation and significantly improving the efficiency and accuracy of resource status perception.

[0016] In an alternative approach, a message processing step is also included:

[0017] The cluster member nodes receive messages from the wireless channel, extract the cluster ID field from the message, and compare it with the local stored cluster ID.

[0018] When the cluster ID in the message does not match the cluster ID to which it belongs, the resource-aware subframe, resource-aware subchannel, and reference signal received power occupied by the message are recorded and updated in the locally stored resource occupancy information.

[0019] When the cluster ID in the message matches the cluster ID to which it belongs and the message contains a resource allocation field, the subframe and subchannel information allocated to itself is extracted based on the cluster member ID in the message and updated in the local transmission resource record.

[0020] Among the above optional methods, messages are processed by comparing and splitting based on cluster IDs, and the local resource occupancy information database and transmission resource records are dynamically updated to achieve real-time synchronization of resource occupancy status and allocation results, avoid resource conflicts and ensure the continuity of data transmission.

[0021] In one alternative approach, the cluster head collaboratively assesses intra-cluster resource availability based on the perception results reported by all cluster members, using a distance-weighted average calculation, and includes the following steps:

[0022] For the target cluster member whose resources are currently to be allocated, the cluster head assigns dynamic weights to other cluster members based on the Euclidean distance between other cluster members and the target cluster member;

[0023] By assigning dynamic weights, the reference signal received power of other cluster members on each resource block is weighted and averaged to generate the resource block interference assessment matrix corresponding to the target cluster member.

[0024] Resource blocks with evaluation values ​​lower than a preset evaluation value threshold in the interference evaluation matrix are identified as available transmission resources for the target cluster member, and the resource block with the lowest evaluation value is selected and assigned to the target cluster member.

[0025] Among the above-mentioned optional methods, Euclidean distance is used to dynamically allocate weights, accurately quantify the interference impact of neighboring nodes, generate an interference evaluation matrix specific to the target member, optimize the selection of low-interference resources, and reduce the intra-cluster communication conflict rate.

[0026] In one alternative approach, a resource adjustment step is also included:

[0027] The cluster member nodes add a resource adjustment request field to the perception results to declare the required number of sub-channels.

[0028] By responding to the resource adjustment request by the cluster head, the distance-weighted average calculation is re-executed to generate an updated interference assessment matrix for the target cluster members, and transmission resources are reallocated to the cluster members based on the updated interference assessment matrix.

[0029] Among the above optional methods, cluster members can actively declare their resource requirements, triggering the cluster head to recalculate the interference assessment matrix in real time and reallocate resources, thereby improving resource utilization and business adaptability and meeting the dynamic needs of rapid vehicle movement scenarios.

[0030] In an alternative approach, a path maintenance step is also included:

[0031] At least one reachable path to the cluster head is dynamically maintained through cluster member nodes, and the maximum number of hops of the reachable path is ensured not to exceed a preset hop count threshold.

[0032] The connectivity of reachable paths is verified by periodically performing path probing by cluster member nodes.

[0033] Among the above-mentioned optional methods, the number of hops on reachable paths is forcibly limited, and path connectivity is ensured through periodic probing. This maintains network topology stability while controlling end-to-end latency, thus meeting the hard requirements for low latency in security services.

[0034] In one alternative approach, the multi-dimensional decision model includes: an intra-cluster forwarding decision model and an inter-cluster forwarding decision model;

[0035] The steps for constructing the intra-cluster forwarding decision model include:

[0036] The distribution dispersion factor is calculated using the density of common neighbor nodes of cluster members and cluster head;

[0037] The coverage contribution value is calculated using the ratio of the cumulative cluster coverage area of ​​cluster members to that of the cluster head;

[0038] Calculate link stability metrics using historical transmission reliability data from cluster members;

[0039] The time slot synchronization degree is calculated using the absolute value of the delay difference between the time slot allocated to cluster members and the time slot transmitted by the cluster head;

[0040] Multi-dimensional weighted fusion of distribution dispersion factor, coverage contribution value, link stability index and time slot synchronization is performed to obtain the evaluation value of intra-cluster forwarding node, so as to complete the construction of intra-cluster forwarding decision model;

[0041] The steps for constructing the inter-cluster forwarding decision model include:

[0042] The position priority is calculated using the ratio of the distance from the candidate node to the destination node to the distance from the cluster head to the destination node;

[0043] Forwarding timeliness is calculated using the absolute value of the delay difference between the candidate node's allocated time slot and the cluster head's transmitted time slot;

[0044] Calculate link stability metrics using historical transmission reliability data from candidate nodes;

[0045] The location priority, forwarding timeliness, and link stability indicators are weighted and fused from multiple dimensions to obtain the inter-cluster forwarding node evaluation value, thereby completing the construction of the inter-cluster forwarding decision model.

[0046] Among the above-mentioned optional methods, by integrating four-dimensional parameters such as topology dispersion, coverage contribution, link stability, and time slot synchronization, it is possible to construct a quantitative evaluation value for forwarding nodes, realize global optimization of forwarding node selection, and balance coverage completeness and resource efficiency.

[0047] In one alternative approach, the step of selecting an intra-cluster forwarding node or an inter-cluster forwarding node includes:

[0048] When selecting intra-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the intra-cluster forwarding decision model. Candidate nodes are selected as intra-cluster forwarding nodes in descending order of evaluation value. The sum of the cumulative cluster coverage area ratios of all selected intra-cluster forwarding nodes and the cluster head is calculated in real time, and the sum of the cumulative cluster coverage area ratios is always kept greater than or equal to 1.

[0049] When selecting inter-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the inter-cluster forwarding decision model, and the candidate node with the highest evaluation value is selected as the inter-cluster forwarding node.

[0050] When the delay difference between the allocation time slot of the selected candidate node and the transmission time slot of the cluster head is greater than the delay difference between the allocation time slot of the idle resource block and the transmission time slot of the cluster head, the dynamic resource reallocation process is triggered. The cluster head adds a forwarding node field to the message to be forwarded. The forwarding node field contains the selected cluster member ID. When the cluster member node resolves that the forwarding node field contains its own cluster member ID, it performs the message forwarding operation.

[0051] Among the above-mentioned optional methods, intra-cluster forwarding is sorted by evaluation value until the cumulative coverage ratio is greater than or equal to 1, which can achieve full coverage of the cluster range with the minimum number of nodes; inter-cluster forwarding directly selects the node with the highest evaluation value, and combines the forwarding field parsing mechanism to achieve precise directional control, simultaneously improving transmission efficiency and reducing channel conflicts.

[0052] In one alternative approach, the dynamic resource reallocation step includes:

[0053] The signal strength weighted average of the idle resource blocks is calculated by distance weighted average, and it is determined whether it is lower than the preset average threshold.

[0054] If the value is below a preset average threshold, the original allocated time slot of the selected candidate node will be replaced with the allocated time slot of an idle resource block, and the cluster head will send a forwarding instruction containing the newly allocated time slot to the selected candidate node; otherwise, the original allocated time slot of the selected candidate node will be maintained.

[0055] Among the above-mentioned optional methods, the candidate node time slot resources are dynamically replaced based on the comparison of time delay difference and the determination of signal strength threshold, which shortens the forwarding waiting time and significantly optimizes the end-to-end latency performance of inter-cluster services.

[0056] Secondly, this invention provides a vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization system, the technical solution of which is as follows:

[0057] The vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization system includes: a resource awareness module, a resource allocation module, and a multi-hop forwarding module;

[0058] The resource awareness module is used to: periodically sense wireless channel resources through cluster member nodes, obtain a sensing result containing resource occupancy information, and report the sensing result to the cluster head; wherein, in a network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members.

[0059] The resource allocation module is used for: the cluster head to collaboratively evaluate the availability of resources within the cluster based on the perception results reported by all cluster members, through distance-weighted average calculation, to obtain the evaluation results, and to dynamically allocate transmission resources to cluster members according to the evaluation results;

[0060] The multi-hop forwarding module is used to: combine the resource allocation status and the topological feature information of the cluster member nodes, and select intra-cluster forwarding nodes or inter-cluster forwarding nodes through a pre-built multi-dimensional decision model to achieve multi-hop message forwarding.

[0061] The beneficial effects of the vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization system of the present invention are as follows:

[0062] The system of this invention significantly reduces the resource conflict rate of wireless channels through intra-cluster collaborative resource awareness and dynamic distance-weighted evaluation mechanism. By optimizing the selection of forwarding nodes by combining a multi-dimensional decision model with topological features, it simultaneously improves multi-hop transmission efficiency and communication reliability, thereby ensuring the low latency requirements of security services.

[0063] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0064] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0065] Figure 1 This is a flowchart illustrating a cross-layer optimization method for resource scheduling and multi-hop forwarding within a vehicle cluster.

[0066] Figure 2 This is a schematic diagram of the cluster head collaborative sensing computing and resource allocation process;

[0067] Figure 3 A schematic diagram of collaborative sensing of cluster resource sets;

[0068] Figure 4 This is a schematic diagram of cluster forwarding combined with intra-cluster wireless resource scheduling;

[0069] Figure 5 A diagram illustrating the reallocation of resources for forwarding nodes;

[0070] Figure 6 This is a schematic diagram of the structure of a cross-layer optimization system for resource scheduling and multi-hop forwarding within a vehicle cluster. Detailed Implementation

[0071] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0072] Figure 1 The following is a flowchart illustrating an embodiment of a cross-layer optimization method for intra-cluster resource scheduling and multi-hop forwarding provided by the present invention: Figure 1 As shown, it includes the following steps:

[0073] S1. By periodically sensing wireless channel resources through cluster member nodes, a sensing result containing resource occupancy information is obtained, and the sensing result is reported to the cluster head; wherein, in a network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members, in S1:

[0074] 1) Wireless channel resources refer to the wireless resource blocks used in vehicle communication, including subframes (time domain) and subchannels (frequency domain), which are used for data transmission in C-V2X technology. They are used to provide physical layer transmission channels for vehicle nodes, ensuring reliable message delivery.

[0075] 2) Resource occupancy information refers to the perceived situation where a specific resource block is occupied by other vehicle nodes, including the Reference Signal Received Power (RSRP). It is used to indicate signal strength, assess the risk of resource conflicts, and avoid allocating highly interfering resource blocks that are already occupied.

[0076] 3) The sensing results are structured information periodically reported by cluster members, including cluster ID, cluster member ID, resource-aware subframe, resource-aware subchannel, and reference signal received power. This provides a resource occupancy view for cluster heads to collaboratively assess resource availability.

[0077] 4) A cluster is a logical group of vehicle nodes that meet preset similarity conditions. It is used to improve communication efficiency through hierarchical topology management of resource scheduling and forwarding.

[0078] The cluster head is the management node of each cluster, responsible for receiving sensing results, allocating resources, and deciding which nodes to forward data to. It coordinates intra-cluster operations to ensure resource optimization and multi-hop reliability.

[0079] Cluster members are vehicle nodes within the cluster that perform perception tasks and data forwarding. They are used for distributed collection of resource occupancy information and participation in communication.

[0080] 5) Preset similarity criteria define the standards for cluster membership, including geographical proximity, consistent travel direction, and similar speed. These are used to ensure the reliability of communication and topology stability within the cluster.

[0081] 6) Periodic sensing refers to cluster members scanning wireless channel resources at fixed time intervals (e.g., every sensing cycle). This is used to achieve dynamic resource monitoring and adapt to changes in the vehicle-to-everything (V2X) environment.

[0082] Specifically, in a network architecture composed of multiple vehicle clusters, each cluster is initially formed based on preset similarity conditions, comprising a cluster head and multiple cluster members. After initialization, cluster member nodes periodically perform wireless channel resource awareness tasks: each cluster member independently scans available subframes and subchannels, measures the reference signal received power of each resource block, and records resource occupancy information (including occupied subframes, subchannels, and RSRP values). After awareness is completed, the cluster member encapsulates the structured awareness results (including cluster ID, its own cluster member ID, resource-aware subframes, resource-aware subchannels, and RSRP) into a message and reports it to the cluster head. Upon receiving the message, the cluster head verifies whether the message source belongs to its own cluster using the cluster ID field. If it belongs to its own cluster, the cluster head updates its local resource records (e.g., storing the awareness results for subsequent evaluation); otherwise, the message is discarded. This process ensures that the cluster head obtains a comprehensive view of the resource occupancy of the entire cluster, providing a data foundation for dynamic resource scheduling, while avoiding irrelevant messages interfering with intra-cluster operations.

[0083] S2: Based on the perception results reported by all cluster members, the cluster head collaboratively assesses the availability of resources within the cluster through distance-weighted averaging, obtains the assessment result, and dynamically allocates transmission resources to cluster members according to the assessment result. In S2:

[0084] 1) Distance-weighted average calculation is a collaborative calculation method in which the cluster head assigns weights based on the distance between cluster members (the closer the distance, the greater the weight) to calculate the weighted average reference signal received power for each resource block. This is used to quantify the interference level of the resource block; a low value indicates high resource availability.

[0085] 2) Intra-cluster resource availability refers to the assessment of whether resource blocks are suitable for allocation to cluster members through collaborative sensing, avoiding high collision probabilities. This is used to ensure maximum spectrum utilization and reduce the risk of transmission failure.

[0086] 3) The evaluation result is a calculated weighted average RSRP matrix, representing the interference intensity of each resource block on each cluster member. It is used to provide a quantitative basis for resource allocation decisions.

[0087] Interference assessment matrix is ​​a two-dimensional assessment table generated for a specific target cluster member. The rows of the matrix represent resource block indices, and the column values ​​are weighted average reference signal received power. The smaller the value, the higher the resource availability.

[0088] 4) The evaluation threshold is a preset interference intensity threshold, used to filter low-interference resource blocks.

[0089] 5) Transmission resources are radio resource blocks dynamically allocated to cluster members, including subframes and subchannels. They are used to provide dedicated transmission channels for cluster members and support efficient data transmission.

[0090] Specifically, after collecting the perception results reported by all cluster members, the cluster head initiates a resource collaborative evaluation and allocation process. First, for each cluster member, the cluster head assigns weights based on the distance between it and other cluster members (closer distances result in higher weights to reflect stronger neighbor interference). Then, it calculates the weighted average reference signal received power for each resource block (lower values ​​indicate less interference and are more suitable for allocation). Next, the cluster head evaluates these weighted results and selects the resource blocks with the lowest interference (such as subframes and subchannels) as transmission resources to allocate to that cluster member. If a cluster member includes a resource request in its message (such as requesting a specific number of subchannels), the cluster head prioritizes these requests; otherwise, according to the adopted C-V2X technology standard, the pre-configured resource allocation mechanism is used by default: a semi-persistent scheduling (SPS) scheme is enabled in LTE-V2X mode, and the Mode2 resource reservation mechanism is used in NR-V2X mode. Finally, the cluster head sends an allocation message to notify cluster members to update their local resource records, ensuring dynamic resource optimization to reduce conflicts.

[0091] S3, the cluster head, combining resource allocation status and topological characteristics of cluster member nodes, selects intra-cluster or inter-cluster forwarding nodes through a pre-built multi-dimensional decision model to achieve multi-hop message forwarding. In S3:

[0092] 1) Resource allocation status refers to the current radio resource allocation of cluster members, including time slot location and sub-channel. It is used to provide time-domain constraints for forwarding decisions and ensure low latency.

[0093] 2) Topological feature information includes the number of common neighbor nodes of cluster members and cluster head, cumulative cluster coverage area, distance to the target, and transmission reliability. This information is used to reflect node location and connectivity, and to optimize forwarding paths.

[0094] 3) The multi-dimensional decision model is a pre-built optimization model that selects the optimal forwarding node by weightedly combining multiple parameters (such as neighbor density, coverage area, time slot location, and reliability). It is used to balance reliability and efficiency to achieve efficient multi-hop communication.

[0095] 4) Intra-cluster forwarding nodes are nodes used to forward messages within the cluster, selected based on a multi-parameter model to maximize priority. Their function is to extend intra-cluster coverage and support restricted multi-hop (e.g., hop count ≤ 2) transmission.

[0096] 5) Inter-cluster forwarding nodes are used to forward messages to other clusters or free nodes, and their selection is based on a multi-parameter model to maximize priority. Their function is to enable cross-cluster communication and reduce end-to-end latency.

[0097] Specifically, the cluster head combines resource allocation status (such as the time slot location of each cluster member) and topological characteristic information (such as the number of common neighbors, cumulative cluster coverage area, distance from the target, and transmission reliability) to select forwarding nodes using a multi-dimensional decision model. For intra-cluster forwarding, the model prioritizes nodes with fewer common neighbors to the cluster head, larger coverage area, closer time slots to the cluster head, and high reliability to ensure that the forwarding path covers the entire cluster and minimizes the number of forwarding nodes.

[0098] The technical solution in this embodiment significantly reduces the resource conflict rate of the wireless channel through intra-cluster collaborative resource awareness and dynamic distance weighted evaluation mechanism. By optimizing the selection of forwarding nodes by combining a multi-dimensional decision model with topological features, it simultaneously improves multi-hop transmission efficiency and communication reliability, thereby ensuring the low latency requirements of security services.

[0099] In one alternative approach, the step of obtaining a sensing result containing resource occupancy information by periodically sensing wireless channel resources through cluster member nodes includes:

[0100] Cluster member nodes perceive the occupancy status of resource blocks in the wireless channel and record resource occupancy information including the subframe number, subchannel number, and corresponding reference signal received power of the occupied resource block.

[0101] The recorded resource occupancy information is encapsulated into a sensing result, which includes: cluster ID, cluster member ID, resource-aware subframe, resource-aware subchannel, and reference signal received power.

[0102] In an alternative approach, a message processing step is also included:

[0103] The cluster member nodes receive messages from the wireless channel, extract the cluster ID field from the message, and compare it with the local stored cluster ID.

[0104] When the cluster ID in the message does not match the cluster ID to which it belongs, the resource-aware subframe, resource-aware subchannel, and reference signal received power occupied by the message are recorded and updated in the locally stored resource occupancy information.

[0105] When the cluster ID in the message matches the cluster ID to which it belongs and the message contains a resource allocation field, the subframe and subchannel information allocated to itself is extracted based on the cluster member ID in the message and updated in the local transmission resource record.

[0106] In this embodiment, after the cluster architecture initialization is complete, cluster member nodes (such as vehicle terminals) initiate periodic resource awareness tasks. A fixed awareness period is set according to the C-V2X technical standard. Each cluster member (denoted as CM)... i The operation of scanning wireless channel resource blocks and detecting resource occupancy status according to a preset cycle.

[0107] Furthermore, within the preset frequency domain range N F (Total number of sub-channels) and time domain range N T Within (total number of subframes), monitor all available resource blocks R. m,n Where m is the subframe number and n is the subchannel number.

[0108] By analyzing physical layer signals, it is determined whether a resource block is occupied by another node. If an occupation signal is detected, the reference received signal power (RSRP) of the resource block is recorded, denoted as p(j). Rm,n (Unit: dBm), where j represents the identifier of the occupying node. If the resource block is idle, it is not recorded.

[0109] Cluster members record resource occupancy information in the following structured format: Resource-aware subframe: Subframe number m of the occupied resource block (1≤m≤N) T Resource-aware subchannel: The subchannel number n (1≤n≤N) of the occupied resource block. F Reference signal received power: measured value p(j) Rm,n It is used to quantify signal strength and interference level.

[0110] Cluster members encapsulate the resource occupancy information recorded in this cycle into a standardized sensing result message. For different fields: Cluster ID (a unique identifier for the cluster to which this node belongs, used to identify the cluster and avoid cross-cluster interference); Cluster Member ID (a unique identifier for this node, used to uniquely identify the sending node); Resource-aware subframe (the recorded occupancy subframe number m, representing the time-domain location of the occupied resource); Resource-aware subchannel (the recorded occupancy subchannel number n, representing the frequency-domain location of the occupied resource); Reference Signal Received Power (RSRP value p(j)R). m,n This is used to quantify signal strength. Finally, at the end of each sensing cycle, a message is generated containing information on all detected occupied resource blocks, and this information is reported to the cluster head via a direct connection, such as a PC5 interface.

[0111] When a cluster member node receives a message from any wireless channel, it extracts the cluster ID field (denoted as ID) from the message. msg And compare it with the cluster ID stored locally (denoted as ID). local Compare them.

[0112] When ID msg ≠ID local When a message originates from an external cluster or a free node, resource occupancy information is recorded. Specifically, this involves extracting the resource-aware subframe m occupied by the message. occ Extract the resource-aware subchannel n occupied by the message occ ;Measurement message reference signal received power p occThen, update the local awareness window, that is, update the resource usage information in the local storage, and set (m occ ,n occ ,p occ Add to resource usage records.

[0113] When ID msg =ID local When the message contains a resource allocation field, the resource allocation information is parsed. Specifically, the cluster member ID field in the message is extracted to determine whether it matches the current CM. i If a match is found, read the allocated subframe T. slot (i) and sub-channel n alloc And update the local transmission resource record, and transfer subframe T slot (i) and sub-channel n alloc Write to the local resource allocation table, that is, update the local transmission resource record, and then record the resource usage information (m) of the message. occ ,n occ ,p occ ).

[0114] In one alternative approach, the cluster head collaboratively assesses intra-cluster resource availability based on the perception results reported by all cluster members, using a distance-weighted average calculation, and includes the following steps:

[0115] For the target cluster member whose resources are currently to be allocated, the cluster head assigns dynamic weights to other cluster members based on the Euclidean distance between other cluster members and the target cluster member;

[0116] By assigning dynamic weights, the reference signal received power of other cluster members on each resource block is weighted and averaged to generate the resource block interference assessment matrix corresponding to the target cluster member.

[0117] Resource blocks with evaluation values ​​lower than a preset evaluation value threshold in the interference evaluation matrix are identified as available transmission resources for the target cluster member, and the resource block with the lowest evaluation value is selected and assigned to the target cluster member.

[0118] In one alternative approach, a resource adjustment step is also included:

[0119] The cluster member nodes add a resource adjustment request field to the perception results to declare the required number of sub-channels.

[0120] By responding to the resource adjustment request by the cluster head, the distance-weighted average calculation is re-executed to generate an updated interference assessment matrix for the target cluster members, and transmission resources are reallocated to the cluster members based on the updated interference assessment matrix.

[0121] It should be noted that, Figure 2The specific process of cluster head collaborative sensing calculation and resource allocation is shown. After receiving a message, the cluster head first extracts the cluster to which it belongs from the cluster ID in the message and determines whether it comes from its own cluster member. If it is a message from its own cluster member, the sensing result attached to the message is recorded locally. Otherwise, the message is discarded. If the message also includes a resource request, a weighted average matrix is ​​calculated and the resource with the smallest weighted average is selected and allocated to the cluster member.

[0122] In this embodiment, the cluster head will be based on the cluster member CM. i The individual resource awareness and the overall resource awareness of cluster K are based on the information of each other cluster member CM. j (CM j ∈K,CM j ≠CM i ) and the cluster member CM i Assign weights ξ(d) based on the Euclidean distance between them i,j ), and calculate the weighted average relative to cluster member CM. i Cluster-wide collaborative resource-aware reference signal received power weighted average matrix

[0123] Furthermore, let the target cluster member of the resource to be allocated be CM. i (e.g., CM1). After the cluster head obtains the perception results reported by all cluster members, it first calculates the Euclidean distance and measures CM1. i Other cluster members CM j The Euclidean distance d for (j≠i) i,j (Unit: meters)

[0124] For example: If cluster member CM2 and CM i Distance d i,2 =50m, CM3 and CM i Distance d i,3 =150m.

[0125] Then, dynamic weights are assigned based on the distance d. i,j Calculate the weight ξ(d) i,j ), where the closer the distance, the greater the weight, and the weighting rule is as follows: (α is the attenuation factor, default α = 1).

[0126] For example: ξ(d i,2 )=1 / 50=0.02,ξ(d i,3 = 1 / 150 ≈ 0.0067.

[0127] Next, weighted average calculation and interference assessment matrix generation are performed. First, the resource blocks are traversed, and for each resource block R... m,n (m∈[1,N T ], n∈[1,NF The calculation is performed using the following formula:

[0128]

[0129] Where, N k The number of resources in a cluster resource set (e.g., N) k =100); P(j)R m,n Indicates other cluster members CM j For resource block R m,n The reference signal received power (e.g., -85dBm); ξ(d i,j (Based on CM) j With cluster member CM i The weights assigned based on the distance between them; A smaller value means that for cluster members, a certain resource block R... m,n The weaker the received power of the reference signal, the lower the probability of it being occupied by neighboring vehicle nodes, making it more suitable for allocation to a member of that cluster. After calculation, an interference evaluation matrix is ​​generated, and the output matrix is... (N T ×N F dimension);

[0130] Then, available resources are filtered, and the evaluation value is compared with a preset evaluation value threshold. If compared, The resource block is then marked as a low-interference available resource, and the set of available resources for the target cluster members is output:

[0131]

[0132] After completing independent resource assessments for all cluster members, a global resource scheduling optimization model is established for the cluster head to optimize overall cluster spectrum utilization and resolve resource allocation conflicts that may arise from local selection. This model aims to minimize the total weighted interference across the entire cluster, and its mathematical expression is as follows:

[0133]

[0134] 1≤m≤N T ,1≤n≤N F ;

[0135]

[0136] in, Indicates cluster member CM i For resource block R m,n The reference signal received power evaluation value; The threshold representing the received power of the reference signal, i.e., the preset evaluation threshold; N T The range of the time domain; N FThe range is defined in the frequency domain. This optimization problem, while ensuring that the interference of a single resource block does not exceed a threshold, can coordinate the allocation of all cluster resources to minimize the total interference.

[0137] Figure 3 The diagram illustrates collaborative sensing of cluster resource sets. For a cluster resource set K, cluster members CM1, CM2, and CM3 have different sensing results. The cluster head can perform collaborative calculations on the cluster resource set based on the sensing results reported by cluster members CM1, CM2, and CM3. The cluster head calculates the weighted average signal power and selects the resource with the smaller weighted average as the CM. i The transmission resources.

[0138] Furthermore, at the initial setup of the cluster structure, the cluster head directly allocates resources to cluster members based on predefined parameters (such as default time intervals and the number of sub-channels). Subsequently, when cluster members have resource adjustment needs, they attach resource adjustment requests to their periodically reported messages. If cluster members do not have additional resource reselection needs, the cluster head defaults to using the semi-persistent scheduling scheme to allocate radio resources to cluster members. When cluster members have additional resource reselection needs, they add a request message field to the message, which includes the number of sub-channels required. After receiving a message from a node, if the message also includes a resource request, the cluster head calculates a weighted average matrix and selects the resource with the smallest weighted average to allocate to the cluster member.

[0139] Furthermore, regarding the dynamic resource adjustment process, firstly, cluster members initiate a resource adjustment request, when CM... i When more resources are needed (e.g., for transmitting high-priority security messages), add a resource adjustment request field to the periodic reporting message. The field content is: Number of sub-channels required = k (e.g., k = 3 indicates CM). i The perception result message declares that three additional sub-channels are required.

[0140] Then, cluster head response and reallocation are performed. After detecting the resource request field, the cluster head assigns it to CM. i Re-execute distance weight assignment (ξ(d)) in sequence i,j Update, calculate weighted average, and generate an updated interference assessment matrix. Select from the updated interference assessment matrix The resource blocks are allocated by prioritizing the k lowest-valued resource blocks (those with the least interference). This allocation of k resource blocks satisfies the needs of the k sub-channels, and the cluster member resource allocation table is updated, including: cluster ID (K), cluster member ID (CM). i ), subframe (T) new ), sub-channel (n new The cluster head notifies the CM via resource allocation messages. i CM i Update local transfer resource records.

[0141] In an alternative approach, a path maintenance step is also included:

[0142] At least one reachable path to the cluster head is dynamically maintained through cluster member nodes, and the maximum number of hops of the reachable path is ensured not to exceed a preset hop count threshold.

[0143] The connectivity of reachable paths is verified by periodically performing path probing by cluster member nodes.

[0144] In this embodiment, each cluster member node needs to dynamically maintain at least one reachable path to the cluster head. This path can adopt a single-hop direct path or a limited multi-hop path (number of hops ≤ 2) transmission method, thereby effectively controlling end-to-end latency while ensuring network connectivity.

[0145] Furthermore, each cluster member CM i At least one reachable path to the cluster head must be maintained. Path types include:

[0146] Single hop path: CM i Direct communication with cluster head CH (distance d) i,CH ≤R com R com (for communication radius);

[0147] Two-hop path: via relay node CM relay Forward (CM) i →CM relay →CH).

[0148] The hop count constraint strictly satisfies: hop count ≤ 2, that is, the maximum number of hops in the path is 2.

[0149] For periodic path detection and verification, firstly, according to the preset period T... probe (Default 100ms) Execute path verification, where:

[0150] During single-hop path verification, CM i Send a path probe request message (HELLO message) directly to the cluster head CH. If the response timeout threshold t is reached... timeout If an acknowledgment message (ACK) is received from CH within 10ms, the path is marked as connected.

[0151] During double-hop path verification, CM i To relay node CM relay Send HELLO message, CM relay Upon receipt, forward to the cluster head CH and send an ACK back to the CM. i If CM i In 2×t timeout If an ACK is received internally, the path is marked as connected;

[0152] If a path fails to be verified k times consecutively (e.g., k=3), the path is removed from the reachable path table, and path reconstruction is triggered. The neighbor list is scanned, and a new path with a hop count ≤ 2 is selected, with priority given to nodes with fewer common neighbors.

[0153] In one alternative approach, the multi-dimensional decision model includes: an intra-cluster forwarding decision model and an inter-cluster forwarding decision model;

[0154] The steps for constructing the intra-cluster forwarding decision model include:

[0155] The distribution dispersion factor is calculated using the density of common neighbor nodes of cluster members and cluster head;

[0156] The coverage contribution value is calculated using the ratio of the cumulative cluster coverage area of ​​cluster members to that of the cluster head;

[0157] Calculate link stability metrics using historical transmission reliability data from cluster members;

[0158] The time slot synchronization degree is calculated using the absolute value of the delay difference between the time slot allocated to cluster members and the time slot transmitted by the cluster head;

[0159] Multi-dimensional weighted fusion of distribution dispersion factor, coverage contribution value, link stability index and time slot synchronization is performed to obtain the evaluation value of intra-cluster forwarding node, so as to complete the construction of intra-cluster forwarding decision model;

[0160] The steps for constructing the inter-cluster forwarding decision model include:

[0161] The position priority is calculated using the ratio of the distance from the candidate node to the destination node to the distance from the cluster head to the destination node;

[0162] Forwarding timeliness is calculated using the absolute value of the delay difference between the candidate node's allocated time slot and the cluster head's transmitted time slot;

[0163] Calculate link stability metrics using historical transmission reliability data from candidate nodes;

[0164] The location priority, forwarding timeliness, and link stability indicators are weighted and fused from multiple dimensions to obtain the inter-cluster forwarding node evaluation value, thereby completing the construction of the inter-cluster forwarding decision model.

[0165] In this embodiment, intra-cluster forwarding adopts proactive forwarding, with the cluster head selecting the optimal forwarding node by constructing a multi-dimensional optimization decision model. The intra-cluster forwarding decision model comprehensively considers key parameters such as the density of common neighbor nodes between cluster members and the cluster head, the cumulative cluster coverage ratio, link transmission reliability, and time slot resource allocation status. The specific calculation formula is as follows:

[0166]

[0167] (Coverage constraints for intra-cluster forwarding node selection);

[0168] In the formula N k N represents the number of members in cluster K; com (i,CH) represents cluster member CM i The number of nodes that share a common neighbor with the cluster head; S k S(i,CH) represents the area covered by cluster K; S(i,CH) represents the area covered by cluster member CM. i Cumulative intra-cluster communication area with cluster head; P suc For transmission reliability; N T T′ is the total number of time slots. slot (i) is a cluster member CM i Occupied time slot position; T CH The time slot occupied by the cluster head; μ1, μ2, μ3, and μ4 are weighting factors; This represents the dispersion factor, used to quantify the dispersion of location distribution (the larger the value, the more dispersed the distribution); P represents the coverage contribution value, used to measure the node's contribution to cluster range expansion; suc This represents a link stability metric, also known as historical transmission success rate, and is used to assess link reliability. This represents the time slot synchronization degree, used to evaluate the degree of time domain alignment.

[0169] With a large Re Intra (i) Cluster members are more likely to be selected as intra-cluster forwarding nodes, meaning that intra-cluster forwarding nodes tend to favor nodes with fewer common neighbors between cluster members and the cluster head (fewer common neighbors mean that the cluster member and the cluster head are more dispersed in the cluster); nodes with a larger cumulative intra-cluster communication coverage area between cluster members and the cluster head; cluster member nodes with higher transmission success rates; and cluster member nodes whose transmission time slots are closer to the cluster head's transmission time slot. Ultimately, the ratio of the cumulative cluster coverage area of ​​all forwarding nodes and the cluster head must be greater than or equal to 1. The goal is to achieve full coverage of cluster-wide communication by minimizing the number of forwarding nodes while ensuring transmission reliability and latency performance. Specifically, each cluster member node needs to dynamically maintain at least one reachable path to the cluster head. This path can adopt a single-hop direct path or a restricted multi-hop path (hop count ≤ 2) to effectively control end-to-end latency while ensuring network connectivity.

[0170] Inter-cluster forwarding can be further divided into intra-cluster forwarding and inter-cluster forwarding outside the cluster. Intra-cluster forwarding relies on members within the cluster, while inter-cluster forwarding outside the cluster is accomplished by multi-hop forwarding by free nodes in a purely distributed structure. Specifically, multi-hop forwarding in a clustered structure is handled by... Figure 4As shown, the cluster is roughly divided into two cluster structures, cluster K1 and cluster K2. Multi-hop V2X messages are generated in cluster member CM1 of cluster K1, and forwarded to free nodes via the cluster head and forwarding node CM4 of cluster K1. Finally, they reach the cluster head of cluster K2 via forwarding node CM5. The selection of forwarding nodes within the cluster is jointly determined by the cluster head based on the distance information of cluster members, the radio resource slot location of cluster members, and the transmission reliability between cluster members. The specific calculation formula is as follows:

[0171]

[0172] In the formula, d(i,Des) represents the cluster member CM. i Distance to the target; d(CH,Des) represents the distance between the cluster head and the target; P suc For transmission reliability; N T T′ is the total number of time slots. slot (i) is a cluster member CM i Occupied time slot position; T CH μ1 represents the time slot position occupied by the cluster head; μ2, μ3 are weighting factors. This represents location priority and is used to quantify geographical advantages (the larger the value, the closer to the target); This represents forwarding timeliness and is used to assess the impact of time slot resources on forwarding latency.

[0173] With a large Re Inter (i) Cluster members will be preferentially selected as inter-cluster forwarding communication nodes within the cluster, i.e., cluster members that are closer to the target, have a time slot location closer to the cluster head, and have higher transmission stability.

[0174] In one alternative approach, the step of selecting an intra-cluster forwarding node or an inter-cluster forwarding node includes:

[0175] When selecting intra-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the intra-cluster forwarding decision model. Candidate nodes are selected as intra-cluster forwarding nodes in descending order of evaluation value. The sum of the cumulative cluster coverage area ratios of all selected intra-cluster forwarding nodes and the cluster head is calculated in real time, and the sum of the cumulative cluster coverage area ratios is always kept greater than or equal to 1.

[0176] When selecting inter-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the inter-cluster forwarding decision model, and the candidate node with the highest evaluation value is selected as the inter-cluster forwarding node.

[0177] When the delay difference between the allocation time slot of the selected candidate node and the transmission time slot of the cluster head is greater than the delay difference between the allocation time slot of the idle resource block and the transmission time slot of the cluster head, the dynamic resource reallocation process is triggered. The cluster head adds a forwarding node field to the message to be forwarded. The forwarding node field contains the selected cluster member ID. When the cluster member node resolves that the forwarding node field contains its own cluster member ID, it performs the message forwarding operation.

[0178] In one alternative approach, the dynamic resource reallocation step includes:

[0179] The signal strength weighted average of the idle resource blocks is calculated by distance weighted average, and it is determined whether it is lower than the preset average threshold.

[0180] If the value is below a preset average threshold, the original allocated time slot of the selected candidate node will be replaced with the allocated time slot of an idle resource block, and the cluster head will send a forwarding instruction containing the newly allocated time slot to the selected candidate node; otherwise, the original allocated time slot of the selected candidate node will be maintained.

[0181] In this embodiment, during the selection process of the intra-cluster forwarding node, a large Re Intra (i) Cluster members are more likely to be selected as intra-cluster forwarding nodes. Specifically, after calculating the evaluation value of each candidate node based on the intra-cluster forwarding decision model, they are sorted from high to low evaluation value. High-value nodes are selected sequentially, and the cumulative cluster coverage ratio is calculated in real time, always maintaining a cumulative cluster coverage ratio ≥ 1.

[0182] In the selection process of inter-cluster forwarding nodes, there is a large Re Intra (i) Cluster members will be preferentially selected as inter-cluster forwarding communication nodes within the cluster. Specifically, after calculating the evaluation value of each candidate node based on the inter-cluster forwarding decision model, the candidate node with the highest evaluation value will be selected as the inter-cluster forwarding node.

[0183] If the time delay difference condition |T is satisfied slot (i)-T CH |>|T free (r)-T CH | and the signal strength of the free resource block Not exceeding the set threshold Then replace the original time slot with an available idle time slot T. free (r), otherwise maintain the original time slot.

[0184] Furthermore, the cluster head will comprehensively consider the idle resources in the cluster resource set, and if the transmission time slot T corresponding to the candidate forwarding node... slot (i) and the current cluster head time slot T CHThe delay between them is greater than the available time slot corresponding to a certain idle resource and the cluster head time slot T. CH The time delay difference between them satisfies the time delay difference condition |T slot (i)-T CH |>|T free (r)-T CH | and the candidate forwarding node is in resource block R m,n Signal strength weighted average Not exceeding the set threshold Right now This triggers resource reallocation, allowing the cluster head to allocate the originally planned T. slot (i) Replace with an available idle time slot T free (r).

[0185] If the final selected inter-cluster forwarding node is based on T free If (r) is selected, the cluster head will, while transmitting forwarding information to the cluster member nodes, also allocate the corresponding radio resource R. r,n This resource is allocated to the node for use in subsequent forwarding processes, thus realizing a dynamic reallocation mechanism for radio resources. This allows forwarding nodes to forward messages to the next node as quickly as possible, further reducing forwarding latency in inter-cluster communication. The cluster head selects inter-cluster forwarding nodes or adjusts idle resource allocation based on time slots, as illustrated below. Figure 5 As shown, the original time slot of cluster member CM4 has been adjusted by the cluster head and is now T free (r) is replaced.

[0186] After selecting a forwarding node, the cluster head adds a forwarding node ID field to the message. If a node parses the field and finds its own cluster member ID, it forwards the message; otherwise, it discards it. Furthermore, after selecting a forwarding node, if the destination node is outside the cluster head's communication range, the cluster head adds the selected cluster member ID to the message's forwarding node field. When that cluster member receives the message and determines it is a forwarding node through the forwarding node field, it will forward the message; other nodes will not forward it.

[0187] Figure 6 This diagram illustrates the structure of an embodiment of a vehicle cluster intra-layer resource scheduling and multi-hop forwarding cross-layer optimization system 200 provided by the present invention. Figure 6 As shown, the system 200 includes: a resource awareness module 210, a resource allocation module 220, and a multi-hop forwarding module 230;

[0188] The resource sensing module 210 is used to: periodically sense wireless channel resources through cluster member nodes, obtain sensing results containing resource occupancy information, and report the sensing results to the cluster head; wherein, in a network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members.

[0189] The resource allocation module 220 is used for: the cluster head to collaboratively evaluate the availability of resources within the cluster based on the perception results reported by all cluster members, through distance-weighted average calculation, to obtain the evaluation result, and to dynamically allocate transmission resources to cluster members according to the evaluation result;

[0190] The multi-hop forwarding module 230 is used to: combine the resource allocation status and the topological feature information of the cluster member nodes, and select intra-cluster forwarding nodes or inter-cluster forwarding nodes through a pre-constructed multi-dimensional decision model to achieve multi-hop message forwarding.

[0191] The technical solution in this embodiment significantly reduces the resource conflict rate of the wireless channel through intra-cluster collaborative resource awareness and dynamic distance weighted evaluation mechanism. By optimizing the selection of forwarding nodes by combining a multi-dimensional decision model with topological features, it simultaneously improves multi-hop transmission efficiency and communication reliability, thereby ensuring the low latency requirements of security services.

[0192] Based on the above scheme, the vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization system of the present invention can be further improved as follows.

[0193] In an alternative embodiment, the resource awareness module 210 is specifically used for:

[0194] Cluster member nodes perceive the occupancy status of resource blocks in the wireless channel and record resource occupancy information including the subframe number, subchannel number, and corresponding reference signal received power of the occupied resource block.

[0195] The recorded resource occupancy information is encapsulated into a sensing result, which includes: cluster ID, cluster member ID, resource-aware subframe, resource-aware subchannel, and reference signal received power.

[0196] Among the above-mentioned optional methods, the structured reporting of perception results is achieved by standardizing and encapsulating resource occupancy information, providing a unified data foundation for cluster head collaborative evaluation and significantly improving the efficiency and accuracy of resource status perception.

[0197] In an alternative embodiment, the module further includes a message processing module, which is used to:

[0198] The cluster member nodes receive messages from the wireless channel, extract the cluster ID field from the message, and compare it with the local stored cluster ID.

[0199] When the cluster ID in the message does not match the cluster ID to which it belongs, the resource-aware subframe, resource-aware subchannel, and reference signal received power occupied by the message are recorded and updated in the locally stored resource occupancy information.

[0200] When the cluster ID in the message matches the cluster ID to which it belongs and the message contains a resource allocation field, the subframe and subchannel information allocated to itself is extracted based on the cluster member ID in the message and updated in the local transmission resource record.

[0201] Among the above optional methods, messages are processed by comparing and splitting based on cluster IDs, and the local resource occupancy information database and transmission resource records are dynamically updated to achieve real-time synchronization of resource occupancy status and allocation results, avoid resource conflicts and ensure the continuity of data transmission.

[0202] In an alternative embodiment, the resource allocation module 220 is specifically used for:

[0203] For the target cluster member whose resources are currently to be allocated, the cluster head assigns dynamic weights to other cluster members based on the Euclidean distance between other cluster members and the target cluster member;

[0204] By assigning dynamic weights, the reference signal received power of other cluster members on each resource block is weighted and averaged to generate the resource block interference assessment matrix corresponding to the target cluster member.

[0205] Resource blocks with evaluation values ​​lower than a preset evaluation value threshold in the interference evaluation matrix are identified as available transmission resources for the target cluster member, and the resource block with the lowest evaluation value is selected and assigned to the target cluster member.

[0206] Among the above-mentioned optional methods, Euclidean distance is used to dynamically allocate weights, accurately quantify the interference impact of neighboring nodes, generate an interference evaluation matrix specific to the target member, optimize the selection of low-interference resources, and reduce the intra-cluster communication conflict rate.

[0207] In an alternative embodiment, the resource adjustment module is further included, the resource adjustment module being used for:

[0208] The cluster member nodes add a resource adjustment request field to the perception results to declare the required number of sub-channels.

[0209] By responding to the resource adjustment request by the cluster head, the distance-weighted average calculation is re-executed to generate an updated interference assessment matrix for the target cluster members, and transmission resources are reallocated to the cluster members based on the updated interference assessment matrix.

[0210] Among the above optional methods, cluster members can actively declare their resource requirements, triggering the cluster head to recalculate the interference assessment matrix in real time and reallocate resources, thereby improving resource utilization and business adaptability and meeting the dynamic needs of rapid vehicle movement scenarios.

[0211] In an alternative embodiment, the path maintenance module is further included, the path maintenance module being used for:

[0212] At least one reachable path to the cluster head is dynamically maintained through cluster member nodes, and the maximum number of hops of the reachable path is ensured not to exceed a preset hop count threshold.

[0213] The connectivity of reachable paths is verified by periodically performing path probing by cluster member nodes.

[0214] Among the above-mentioned optional methods, the number of hops on reachable paths is forcibly limited, and path connectivity is ensured through periodic probing. This maintains network topology stability while controlling end-to-end latency, thus meeting the hard requirements for low latency in security services.

[0215] In one alternative approach, the multi-dimensional decision model includes: an intra-cluster forwarding decision model and an inter-cluster forwarding decision model; it also includes: a model building module, which is used for:

[0216] The distribution dispersion factor is calculated using the density of common neighbor nodes of cluster members and cluster head;

[0217] The coverage contribution value is calculated using the ratio of the cumulative cluster coverage area of ​​cluster members to that of the cluster head;

[0218] Calculate link stability metrics using historical transmission reliability data from cluster members;

[0219] The time slot synchronization degree is calculated using the absolute value of the delay difference between the time slot allocated to cluster members and the time slot transmitted by the cluster head;

[0220] Multi-dimensional weighted fusion of distribution dispersion factor, coverage contribution value, link stability index and time slot synchronization is performed to obtain the evaluation value of intra-cluster forwarding node, so as to complete the construction of intra-cluster forwarding decision model;

[0221] The position priority is calculated using the ratio of the distance from the candidate node to the destination node to the distance from the cluster head to the destination node;

[0222] Forwarding timeliness is calculated using the absolute value of the delay difference between the candidate node's allocated time slot and the cluster head's transmitted time slot;

[0223] Calculate link stability metrics using historical transmission reliability data from candidate nodes;

[0224] The location priority, forwarding timeliness, and link stability indicators are weighted and fused from multiple dimensions to obtain the inter-cluster forwarding node evaluation value, thereby completing the construction of the inter-cluster forwarding decision model.

[0225] Among the above-mentioned optional methods, by integrating four-dimensional parameters such as topology dispersion, coverage contribution, link stability, and time slot synchronization, it is possible to construct a quantitative evaluation value for forwarding nodes, realize global optimization of forwarding node selection, and balance coverage completeness and resource efficiency.

[0226] In one alternative embodiment, the multi-hop forwarding module 230 is specifically used for:

[0227] When selecting intra-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the intra-cluster forwarding decision model. Candidate nodes are selected as intra-cluster forwarding nodes in descending order of evaluation value. The sum of the cumulative cluster coverage area ratios of all selected intra-cluster forwarding nodes and the cluster head is calculated in real time, and the sum of the cumulative cluster coverage area ratios is always kept greater than or equal to 1.

[0228] When selecting inter-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the inter-cluster forwarding decision model, and the candidate node with the highest evaluation value is selected as the inter-cluster forwarding node.

[0229] When the delay difference between the allocation time slot of the selected candidate node and the transmission time slot of the cluster head is greater than the delay difference between the allocation time slot of the idle resource block and the transmission time slot of the cluster head, the dynamic resource reallocation process is triggered. The cluster head adds a forwarding node field to the message to be forwarded. The forwarding node field contains the selected cluster member ID. When the cluster member node resolves that the forwarding node field contains its own cluster member ID, it performs the message forwarding operation.

[0230] Among the above-mentioned optional methods, intra-cluster forwarding is sorted by evaluation value until the cumulative coverage ratio is greater than or equal to 1, which can achieve full coverage of the cluster range with the minimum number of nodes; inter-cluster forwarding directly selects the node with the highest evaluation value, and combines the forwarding field parsing mechanism to achieve precise directional control, simultaneously improving transmission efficiency and reducing channel conflicts.

[0231] In an alternative embodiment, the multi-hop forwarding module 230 is further specifically used for:

[0232] The signal strength weighted average of the idle resource blocks is calculated by distance weighted average, and it is determined whether it is lower than the preset average threshold.

[0233] If the value is below a preset average threshold, the original allocated time slot of the selected candidate node will be replaced with the allocated time slot of an idle resource block, and the cluster head will send a forwarding instruction containing the newly allocated time slot to the selected candidate node; otherwise, the original allocated time slot of the selected candidate node will be maintained.

[0234] Among the above-mentioned optional methods, the candidate node time slot resources are dynamically replaced based on the comparison of time delay difference and the determination of signal strength threshold, which shortens the forwarding waiting time and significantly optimizes the end-to-end latency performance of inter-cluster services.

[0235] The parameters and steps for implementing the corresponding functions of each module in the vehicle cluster resource scheduling and multi-hop forwarding cross-layer optimization system 200 of this embodiment can be referred to the parameters and steps in the embodiment of the vehicle cluster resource scheduling and multi-hop forwarding cross-layer optimization method above, and will not be repeated here.

[0236] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for resource scheduling and multi-hop forwarding across layers within a vehicle cluster, characterized in that, include: The cluster member nodes periodically sense wireless channel resources, obtain sensing results containing resource occupancy information, and report the sensing results to the cluster head; in the network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members. Based on the perception results reported by all cluster members, the cluster head collaboratively assesses the availability of resources within the cluster through distance-weighted average calculation, obtains the assessment results, and dynamically allocates transmission resources to cluster members according to the assessment results; The cluster head combines the resource allocation status and the topological characteristics of the cluster member nodes, and selects intra-cluster forwarding nodes or inter-cluster forwarding nodes through a pre-built multi-dimensional decision model to achieve multi-hop message forwarding. The step of the cluster head collaboratively assessing the availability of resources within the cluster based on the perception results reported by all cluster members and through distance-weighted averaging includes: For the target cluster member whose resources are currently to be allocated, the cluster head assigns dynamic weights to other cluster members based on the Euclidean distance between other cluster members and the target cluster member; By assigning dynamic weights, the reference signal received power of other cluster members on each resource block is weighted and averaged to generate the resource block interference assessment matrix corresponding to the target cluster member. Resource blocks with evaluation values ​​lower than a preset evaluation value threshold in the interference evaluation matrix are identified as available transmission resources for the target cluster member, and the resource block with the lowest evaluation value is selected and assigned to the target cluster member.

2. The vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization method according to claim 1, characterized in that, The step of obtaining a sensing result containing resource occupancy information by periodically sensing wireless channel resources through cluster member nodes includes: Cluster member nodes perceive the occupancy status of resource blocks in the wireless channel and record resource occupancy information including the subframe number, subchannel number, and corresponding reference signal received power of the occupied resource block. The recorded resource occupancy information is encapsulated into a sensing result, which includes: cluster ID, cluster member ID, resource-aware subframe, resource-aware subchannel, and reference signal received power.

3. The vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization method according to claim 2, characterized in that, It also includes message processing steps: The cluster member nodes receive messages from the wireless channel, extract the cluster ID field from the message, and compare it with the local stored cluster ID. When the cluster ID in the message does not match the cluster ID to which it belongs, the resource-aware subframe, resource-aware subchannel, and reference signal received power occupied by the message are recorded and updated in the locally stored resource occupancy information. When the cluster ID in the message matches the cluster ID to which it belongs and the message contains a resource allocation field, the subframe and subchannel information allocated to itself is extracted based on the cluster member ID in the message and updated in the local transmission resource record.

4. The method for intra-cluster resource scheduling and multi-hop forwarding cross-layer optimization according to claim 1, characterized in that, It also includes resource adjustment steps: The cluster member nodes add a resource adjustment request field to the perception results to declare the required number of sub-channels. By responding to the resource adjustment request by the cluster head, the distance-weighted average calculation is re-executed to generate an updated interference assessment matrix for the target cluster members, and transmission resources are reallocated to the cluster members based on the updated interference assessment matrix.

5. The method for intra-cluster resource scheduling and multi-hop forwarding cross-layer optimization according to claim 1, characterized in that, It also includes path maintenance steps: At least one reachable path to the cluster head is dynamically maintained through cluster member nodes, and the maximum number of hops of the reachable path is ensured not to exceed a preset hop count threshold. The connectivity of reachable paths is verified by periodically performing path probing by cluster member nodes.

6. The method for intra-cluster resource scheduling and multi-hop forwarding cross-layer optimization according to claim 1, characterized in that, The multi-dimensional decision-making model includes: intra-cluster forwarding decision-making model and inter-cluster forwarding decision-making model; The steps for constructing the intra-cluster forwarding decision model include: The distribution dispersion factor is calculated using the density of common neighbor nodes of cluster members and cluster head; The coverage contribution value is calculated using the ratio of the cumulative cluster coverage area of ​​cluster members to that of the cluster head; Calculate link stability metrics using historical transmission reliability data from cluster members; The time slot synchronization degree is calculated using the absolute value of the delay difference between the time slot allocated to cluster members and the time slot transmitted by the cluster head; Multi-dimensional weighted fusion of distribution dispersion factor, coverage contribution value, link stability index and time slot synchronization is performed to obtain the evaluation value of intra-cluster forwarding node, so as to complete the construction of intra-cluster forwarding decision model; The steps for constructing the inter-cluster forwarding decision model include: The position priority is calculated using the ratio of the distance from the candidate node to the destination node to the distance from the cluster head to the destination node; Forwarding timeliness is calculated using the absolute value of the delay difference between the candidate node's allocated time slot and the cluster head's transmitted time slot; Calculate link stability metrics using historical transmission reliability data from candidate nodes; The location priority, forwarding timeliness, and link stability indicators are weighted and fused from multiple dimensions to obtain the inter-cluster forwarding node evaluation value, thereby completing the construction of the inter-cluster forwarding decision model.

7. The vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization method according to claim 6, characterized in that, The step of selecting an intra-cluster forwarding node or an inter-cluster forwarding node includes: When selecting intra-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the intra-cluster forwarding decision model. Candidate nodes are selected as intra-cluster forwarding nodes in descending order of evaluation value. The sum of the cumulative cluster coverage area ratios of all selected intra-cluster forwarding nodes and the cluster head is calculated in real time, and the sum of the cumulative cluster coverage area ratios is always kept greater than or equal to 1. When selecting inter-cluster forwarding nodes, the evaluation value of each candidate node is calculated based on the inter-cluster forwarding decision model, and the candidate node with the highest evaluation value is selected as the inter-cluster forwarding node. When the delay difference between the allocation time slot of the selected candidate node and the transmission time slot of the cluster head is greater than the delay difference between the allocation time slot of the idle resource block and the transmission time slot of the cluster head, the dynamic resource reallocation process is triggered. The cluster head adds a forwarding node field to the message to be forwarded. The forwarding node field contains the selected cluster member ID. When the cluster member node resolves that the forwarding node field contains its own cluster member ID, it performs the message forwarding operation.

8. The method for intra-cluster resource scheduling and multi-hop forwarding cross-layer optimization according to claim 7, characterized in that, The dynamic resource reallocation steps include: The signal strength weighted average of the idle resource blocks is calculated by distance weighted average, and it is determined whether it is lower than the preset average threshold. If the value is below a preset average threshold, the original allocated time slot of the selected candidate node will be replaced with the allocated time slot of an idle resource block, and the cluster head will send a forwarding instruction containing the newly allocated time slot to the selected candidate node; otherwise, the original allocated time slot of the selected candidate node will be maintained.

9. A vehicle cluster intra-resource scheduling and multi-hop forwarding cross-layer optimization system, characterized in that, include: The resource awareness module is used to periodically sense wireless channel resources through cluster member nodes, obtain the sensing results containing resource occupancy information, and report the sensing results to the cluster head. In the network architecture composed of multiple vehicle clusters, each cluster containing vehicle nodes that meet preset similarity conditions consists of a cluster head and multiple cluster members. The resource allocation module is used by the cluster head to collaboratively evaluate the availability of resources within the cluster based on the perception results reported by all cluster members, through distance-weighted average calculation, to obtain the evaluation results, and to dynamically allocate transmission resources to cluster members according to the evaluation results. The step of the cluster head collaboratively assessing the availability of resources within the cluster based on the perception results reported by all cluster members and through distance-weighted averaging includes: For the target cluster member whose resources are currently to be allocated, the cluster head assigns dynamic weights to other cluster members based on the Euclidean distance between other cluster members and the target cluster member; By assigning dynamic weights, the reference signal received power of other cluster members on each resource block is weighted and averaged to generate the resource block interference assessment matrix corresponding to the target cluster member. Resource blocks with evaluation values ​​lower than a preset evaluation value threshold in the interference evaluation matrix are identified as available transmission resources for the target cluster members, and the resource block with the lowest evaluation value is selected from them and assigned to the target cluster members. The multi-hop forwarding module is used by the cluster head to combine the resource allocation status and the topological characteristics of the cluster member nodes, and select intra-cluster forwarding nodes or inter-cluster forwarding nodes through a pre-built multi-dimensional decision model to achieve multi-hop message forwarding.

Images