Collaborative V2X Communication And Computing Resource Allocation Technique Based On Deep Reinforcement Learning

In traditional cloud computing,data needs to be transmitted from distributed data sources to remote cloud data centers for high-performance computing.However,the transmission of tasks from data sources to cloud data centers typically incurs network transmission delays of hundreds of milliseconds or even longer,which cannot meet the low-latency requirements of V2 X systems.In contrast to cloud computing,mobile edge computing technology moves computing processing to the edge of the network,enabling collaborative communication and computing resources in V2 X systems.This approach significantly reduces response latency and data transmission overhead,and has gradually become the mainstream solution in V2 X systems.Therefore,this paper focuses on mobile edge computing technology and collaborative communication and computing resource coordination techniques for heterogeneous computing tasks and multi-communication modes.With the continuous development of intelligent connected vehicles,the types of invehicle applications have become increasingly diverse.However,most existing research on collaborative communication and computing resources is still based on single-task scenarios and cannot fully meet the requirements of current intelligent connected vehicles.Building on the existing research,this paper proposes a communication and computing resource coordination technique based on heterogeneous tasks,which can meet the requirements of latency and overhead when facing tasks with different latency requirements and different divisibility.This approach utilizes the communication and computing resources of roadside units(RSUs)and service vehicles to complete the computing tasks of the task vehicles within the latency requirements and with minimal overhead.Since this optimization problem is nonlinear,this paper adopts deep reinforcement learning to solve it.Finally,the performance of the proposed approach is verified through simulations.On the other hand,intelligent connected vehicles communicate with RSUs for perception fusion to improve the accuracy of functions such as advanced driver assistance systems.Existing solutions mostly consider only single communication modes,but V2 X communication modes are not limited to one,including DSRC and C-V2 X,which may result in resource waste and affect resource utilization.This paper studies communication and computing resource coordination schemes under scenarios where multiple communication modes are combined with perception capabilities,enabling rapid communication and computing resource coordination for vehicles with perception requirements.When vehicles require perception information with a certain accuracy,they generate computing demands and choose appropriate communication methods to transmit computing tasks through communication resources to service vehicles or roadside units,achieving the goal of collaborative communication and computing resource coordination.This approach comprehensively considers factors such as task offloading ratio,communication mode,accuracy constraints,and transmission power,and establishes a joint optimization problem for perception accuracy,latency constraints,and overhead.Finally,the effectiveness of the proposed approach is verified through simulation experiments.