Internet of vehicles resource scheduling method, device and system, and storage medium

Abstract

The invention discloses an Internet of Vehicles resource scheduling method, device and system, and a storage medium, and the method comprises the steps: initializing the resource state information of each vehicle node according to the topological structure of each vehicle in the Internet of Vehicles; preheating the vehicle nodes through multi-round information exchange according to the resource state information of the vehicle nodes to obtain a spectrum resource allocation intermediate state under the current iteration; according to the vehicle node spectrum resource allocation intermediate state under the current iteration, the importance of each node in global resource allocation is evaluated, and a scheduling sequence of resource allocation is generated; and according to the scheduling sequence of resource allocation and the serial numbers of the nodes in the scheduling sequence, each vehicle node asynchronously executes spectrum resource allocation operation after receiving a completion signal of the node with the previous serial number, and outputs a resource scheduling result. By adopting the technical scheme of the invention, the interference management problem in resource allocation can be efficiently solved, and the performance of the whole system is optimized.

Description

Technical Field

[0001] This invention belongs to the field of vehicle network resource scheduling technology, and particularly relates to a vehicle network resource scheduling method, device, system, and storage medium. Background Technology

[0002] Vehicle-to-everything (V2X) technology, as a core component of future intelligent transportation systems, is widely used in fields such as autonomous driving, fleet management, and traffic monitoring. The multi-task communication resource allocation problem in V2X can be addressed by effectively scheduling resources to improve system performance. The resource scheduling problem is similar to the graph coloring problem, i.e., how to allocate limited spectrum resources among different vehicles to avoid interference and improve overall communication efficiency. Resource scheduling in V2X faces challenges in optimizing spectrum resource allocation and managing interference, especially when communication between vehicles and base stations is frequent and dynamically changing. Optimizing resource allocation, reducing conflicts, and improving communication quality is a crucial issue.

[0003] With the advancement of graph neural networks and reinforcement learning techniques, researchers have begun to model each vehicle node in the Internet of Vehicles (IoV) as an autonomous agent, achieving distributed resource allocation decisions through cooperation between agents. Reference 1, "Chao Zhu et al., “Multi-Task Communication Resource Allocation for MIMO-Based Vehicular Fog Computing,” IEEE Transactions on Vehicular Technology, vol. 73, no. 1, January 2024, proposes a multi-agent communication resource allocation framework that uses deep learning and optimization algorithms to allocate resources for tasks in the IoV environment. However, these methods suffer from the following problems in dynamic resource scheduling: On the one hand, traditional synchronization strategies are difficult to guarantee the stability and efficiency of resource scheduling in the complex communication environment and frequent signal interference of the Internet of Vehicles. On the other hand, although some methods use node importance ranking for resource scheduling, they lack clear modeling and measurement of vehicle "importance", which makes it impossible to accurately prioritize the scheduling of key vehicles, thus affecting the overall system performance. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method, device, system and storage medium for scheduling resources in the Internet of Vehicles (IoV) for optimizing the allocation of spectrum resources in a dynamic IoV environment; by introducing node importance assessment and asynchronous multi-agent scheduling mechanism, the interference management problem in resource allocation can be solved efficiently and the overall system performance can be optimized.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for scheduling vehicle-to-everything (V2X) resources includes: Step S1: Initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network; Step S2: Based on the resource status information of the vehicle nodes, perform preheating processing on the vehicle nodes through multiple rounds of information exchange to obtain the intermediate state of spectrum resource allocation under the current iteration; Step S3: Based on the intermediate state of vehicle node spectrum resource allocation in the current iteration, evaluate the importance of each node in the global resource allocation and generate a scheduling sequence for resource allocation; Step S4: According to the scheduling sequence of resource allocation, each vehicle node asynchronously performs spectrum resource allocation operation after receiving the completion signal from the node with the previous number, and outputs the resource scheduling result.

[0006] Preferably, in step S3, the importance evaluation score of each vehicle node in the global resource allocation process is calculated based on the overlap of resource probability distributions.

[0007] The present invention also provides a vehicle network resource scheduling device, comprising: The first processing module is used to initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network. The second processing module is used to preheat the vehicle nodes by exchanging information in multiple rounds based on the resource status information of the vehicle nodes, so as to obtain the intermediate state of spectrum resource allocation under the current iteration. The third processing module is used to evaluate the importance of each node in the global resource allocation based on the intermediate state of the vehicle node spectrum resource allocation in the current iteration, and generate a scheduling sequence for resource allocation. The fourth processing module is used to perform spectrum resource allocation operations asynchronously by each vehicle node after receiving the completion signal from the preceding node in its scheduling sequence, and output the resource scheduling results, according to the scheduling sequence of resource allocation.

[0008] As a preferred option, the third processing module calculates the importance evaluation score of each vehicle node in the global resource allocation process based on the overlap of resource probability distributions.

[0009] The present invention also provides a vehicle network resource scheduling system, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a vehicle network resource scheduling method when executed by the processor.

[0010] The present invention also provides a storage medium storing a computer program, which executes a vehicle network resource scheduling method when running.

[0011] This invention designs a node importance metric based on the overlap of resource probability distributions to dynamically identify critical vehicles that have a significant impact on resource scheduling. It prioritizes high-importance vehicle nodes through an asynchronous scheduling mechanism and optimizes resource allocation strategies by combining local resource distribution and neighboring vehicle information.

[0012] Compared with traditional methods, the present invention has the following advantages: 1) This invention proposes a clear and calculable node importance index. Based on the local conflict intensity of the probability distribution overlap of vehicle resources, it can effectively identify vehicle nodes that have a key impact in the allocation of vehicle network spectrum resources, thereby improving the rationality of resource allocation order and the pertinence of decision-making.

[0013] 2) This invention adopts an asynchronous multi-agent scheduling mechanism. Each node (vehicle) agent in the Internet of Vehicles makes resource scheduling decisions independently based on local perception and policy model under asynchronous conditions, avoiding the computational bottleneck problem that may occur in traditional synchronous execution and improving the scheduling flexibility and adaptability of the system.

[0014] 3) The resource scheduling process of this invention introduces a local spectrum conflict probability estimation mechanism. When a vehicle node finds that its effective spectrum resource set is empty, it can dynamically expand the spectrum resource options, thereby effectively controlling the conflict propagation and solution space constraints in the resource allocation process and improving the model's adaptability to different graph structures.

[0015] 4) The intelligent agent design and scheduling mechanism adopted in this invention has good modularity, supports flexible replacement and integration of algorithm components, and has the theoretical basis and engineering feasibility to be further extended to other graph theory decision problems (such as graph segmentation, covering, labeling, etc.), and can further optimize other resource scheduling problems in the Internet of Vehicles. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 This is a flowchart of the vehicle network resource scheduling method according to an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1: like Figure 1 As shown, this embodiment of the invention provides a vehicle network resource scheduling method, including: Step S1: Initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network; Step S2: Based on the resource status information of the vehicle nodes, the vehicle nodes are preheated through multiple rounds of information exchange to fully integrate the information between the nodes, thereby forming the intermediate state of spectrum resource allocation under the current iteration, that is, the resource allocation probability distribution of each node, which provides a basis for subsequent asynchronous optimization allocation.

[0021] Step S3: Based on the intermediate state of vehicle node spectrum resource allocation formed during the preheating stage, i.e. the resource allocation probability distribution of each node, assess the importance of each node in the global resource allocation, and generate the scheduling sequence required for subsequent optimization allocation accordingly.

[0022] Step S4: According to the scheduling sequence, after each vehicle node receives the completion signal from the preceding node in the scheduling sequence, it triggers an asynchronous resource selection operation and broadcasts the completion signal to all nodes after completion, so that subsequent nodes can determine whether it can be executed, and finally outputs the scheduling result.

[0023] As one embodiment of the invention, step S2 includes: Step S21: Each vehicle node encapsulates its initial resource status information into a message containing a vehicle identifier, a spectrum resource allocation probability vector, and a list of neighboring nodes, and broadcasts it to its neighboring vehicle nodes, and receives information from neighboring vehicle nodes, thereby realizing the first information aggregation among neighboring vehicle nodes. Step S22: Each vehicle node updates its own spectrum resource probability distribution based on the aggregated information of its neighboring vehicle nodes, so that it can more accurately reflect the resource usage and interference relationship between nodes in its local network environment. Step S23: Repeat steps S21-S22 until the preset number of iterations (e.g., 3 times) is reached to form an intermediate state of spectrum resource allocation for vehicle nodes, i.e., the resource allocation probability distribution of each node, which is used for subsequent asynchronous resource optimization allocation.

[0024] As one embodiment of the invention, step S3 includes: Step S31: Obtain the current spectrum resource allocation status of all vehicle nodes in the vehicle network; Step S32: Based on the current spectrum resource allocation status, calculate the importance evaluation score of each vehicle node in the global resource allocation process by the overlap of resource probability distribution between the vehicle node and its neighboring nodes; Step S33: Sort all vehicle nodes without allocated spectrum resources in descending order according to their importance score to generate a scheduling sequence for resource allocation.

[0025] Furthermore, in step S32, the importance score of the vehicle node is represented by a calculation method based on the overlap of resource probability distributions, which is used to measure the potential impact of the node's resource allocation decision on the performance of vehicle network resource scheduling. The calculation method is as follows: ; in, This represents the importance score of the i-th vehicle node in the global resource allocation process; This represents the probability that node i selects the k-th spectrum resource; This represents the probability that node i's neighbor node j selects the k-th spectrum resource; This represents the set of neighboring nodes that have a direct communication connection with node i. This indicates the total amount of spectrum resources currently available for allocation.

[0026] As one embodiment of the present invention, step S4 specifically includes: Step S41: After the first vehicle node in the scheduling sequence completes the aforementioned stable process of resource allocation probability distribution, it can automatically perform spectrum resource allocation operation without waiting for the completion signal, and broadcast the completion signal and its own number to the entire network after completion to start the entire scheduling process. Step S42: For subsequent vehicle nodes in the scheduling sequence, the resource selection process can be triggered only after receiving the completion signal from the previous numbered node in the scheduling sequence. Based on its spectrum resource distribution probability and the actual resource occupancy of neighboring nodes, the optimal spectrum resource is selected for allocation decision. Specifically, each vehicle node first selects the spectrum resource with the lowest conflict probability from its own available spectrum resource set for occupancy decision based on the pre-determined spectrum resource probability distribution (the stable probability distribution state obtained from step S21) and the currently determined resource occupancy information of neighboring nodes (i.e. the spectrum resources actually allocated by neighboring nodes).

[0027] The spectrum resource allocation decision specifically includes: (1) The probability distribution of vehicle nodes based on the currently available spectrum resources In addition to the actual spectrum resources occupied by neighboring nodes, determine its own effective spectrum resource set. That is, the set of spectrum resources that do not conflict with the resource occupation of neighboring nodes.

[0028] (2) Vehicle nodes from the set of effective spectrum resources Select the spectrum resource with the highest occupancy probability. Specifically expressed as .

[0029] (3) If the current vehicle node has an effective spectrum resource set If the spectrum is empty (meaning all resources are occupied or conflicting with each other), the node dynamically adds new spectrum resources through the resource expansion mechanism and immediately occupies them to avoid resource conflicts.

[0030] (4) After the vehicle node completes the spectrum resource allocation decision, it immediately broadcasts the occupancy status to the neighboring nodes and updates its status for other nodes to refer to.

[0031] Step S43: After completing resource selection, the current node broadcasts a completion signal and its own number to all vehicle nodes so that the next node can determine whether it can be executed, thereby ensuring that the entire scheduling sequence proceeds in an asynchronous environment. Step S44: If the current vehicle node finds that its effective spectrum resource set is empty (i.e., there is no available spectrum resource), then it dynamically introduces additional spectrum resource options to expand the available spectrum set and immediately occupies the one with the lowest conflict probability to avoid resource allocation failure.

[0032] In summary, the embodiments of the present invention have the following characteristics: 1. Dynamic scheduling optimization decision-making. By dynamically adjusting the spectrum resource scheduling order according to the importance of nodes, priority is given to vehicle nodes that have a significant impact on the efficiency of spectrum resource allocation in the vehicle network. This can quickly solve resource scheduling bottlenecks and avoid the inefficiencies and high computational costs caused by random or fixed scheduling orders in traditional methods.

[0033] 2. Improved computing resource utilization. This invention can prioritize critical nodes and update resource allocation decisions in real time, thereby reducing backtracking calculations and resource waste caused by spectrum conflicts; especially in large-scale vehicle network resource scheduling tasks, it effectively reduces computing resource consumption, significantly improves computing resource utilization, and ensures the efficient operation of the vehicle network system.

[0034] 3. Resolving conflict issues in complex vehicle-to-everything (V2X) architectures. This invention effectively addresses spectrum resource conflicts between vehicle nodes in V2X by introducing a dynamic spectrum expansion mechanism and a conflict-sensitive reward function, thereby reducing spectrum conflicts during resource scheduling.

[0035] 4. This invention employs a decentralized scheduling strategy, enabling each node to independently make resource scheduling decisions based on its local information and importance score. This avoids the computational bottlenecks and synchronization problems caused by a central scheduler in traditional methods. Through an asynchronous multi-agent scheduling mechanism, each vehicle node makes resource allocation decisions based on its local perception, improving scheduling flexibility and adaptability.

[0036] 5. Easy to integrate and deploy. This invention is highly flexible and can quickly integrate with existing vehicle networking architectures. Through a simplified deployment process, it can be efficiently executed in various vehicle networking application scenarios and seamlessly interface with traditional communication platforms, while providing a convenient deployment path for practical applications.

[0037] Example 2: This invention also provides a vehicle network resource scheduling device, comprising: The first processing module is used to initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network. The second processing module is used to preheat the vehicle nodes by exchanging information in multiple rounds based on the resource status information of the vehicle nodes, so as to obtain the intermediate state of spectrum resource allocation under the current iteration. The third processing module is used to evaluate the importance of each node in the global resource allocation based on the intermediate state of the vehicle node spectrum resource allocation in the current iteration, and generate a scheduling sequence for resource allocation. The fourth processing module is used to perform spectrum resource allocation operations asynchronously by each vehicle node after receiving the completion signal from the previous numbered node in the scheduling sequence, and output the resource scheduling results, according to the scheduling sequence of resource allocation.

[0038] As one embodiment of the present invention, the third processing module calculates the importance evaluation score of each vehicle node in the global resource allocation process based on the overlap of resource probability distribution.

[0039] Example 3: This invention also provides a vehicle network resource scheduling system, including: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a vehicle network resource scheduling method when run by the processor.

[0040] Example 4: This invention also provides a storage medium storing a computer program, which executes a vehicle network resource scheduling method during runtime.

[0041] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for scheduling vehicle network resources, characterized in that, include: Step S1: Initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network; Step S2: Based on the resource status information of the vehicle nodes, perform preheating processing on the vehicle nodes through multiple rounds of information exchange to obtain the intermediate state of spectrum resource allocation under the current iteration; Step S3: Based on the intermediate state of vehicle node spectrum resource allocation in the current iteration, evaluate the importance of each node in the global resource allocation and generate a scheduling sequence for resource allocation; Step S4: According to the scheduling sequence of resource allocation, each vehicle node asynchronously performs spectrum resource allocation operation after receiving the completion signal from the node with the previous number, and outputs the resource scheduling result.

2. The vehicle network resource scheduling method as described in claim 1, characterized in that, In step S3, the importance evaluation score of each vehicle node in the global resource allocation process is calculated based on the overlap of resource probability distribution.

3. A vehicle-to-everything (V2X) resource scheduling device, characterized in that, include: The first processing module is used to initialize the resource status information of each vehicle node according to the topology of each vehicle in the vehicle network. The second processing module is used to preheat the vehicle nodes by exchanging information in multiple rounds based on the resource status information of the vehicle nodes, so as to obtain the intermediate state of spectrum resource allocation under the current iteration. The third processing module is used to evaluate the importance of each node in the global resource allocation based on the intermediate state of the vehicle node spectrum resource allocation in the current iteration, and generate a scheduling sequence for resource allocation. The fourth processing module is used to perform spectrum resource allocation operations asynchronously by each vehicle node after receiving the completion signal from the preceding node in its scheduling sequence, and output the resource scheduling results, according to the scheduling sequence of resource allocation.

4. The vehicle network resource scheduling device as described in claim 3, characterized in that, The third processing module calculates the importance evaluation score of each vehicle node in the global resource allocation process based on the overlap of resource probability distributions.

5. A vehicle-to-everything (V2X) resource scheduling system, characterized in that, include: A memory and a processor, wherein the memory stores a computer program executed by the processor, the computer program executing the vehicle network resource scheduling method as described in any one of claims 1-2 when run by the processor.

6. A storage medium, characterized in that, The storage medium stores a computer program, which executes the vehicle network resource scheduling method as described in any one of claims 1-2 when it runs.

Images