Revenue and Energy Efficiency-Driven Delay-Constrained Computing Task Offloading and Resource Allocation in a Vehicular Edge Computing Network: A Deep Reinforcement Learning Approach

For in-vehicle application, task type and vehicle state information, i.e., speed, bear a significant impact on the delay requirement. However, joint of speed constraint has not been studied, this lack study may cause mismatch between requirement allocated computation wireless resources. In article, we propose speed-aware offloading resource allocation strategy to decrease vehicle’s energy cost for executing tasks increase revenue processing within constraint. First, establish model. Then, delay, cost, execution in vehicular edge computing (VEC) server, local terminal, terminals other vehicles are calculated. Based from execution, utility function is acquired. Next, formulate optimization maximize level subject constraints resources, To obtain near-optimal solution formulated problem, based multiagent deep deterministic policy gradient (JORA-MADDPG) algorithm proposed vehicles. Simulation results show that our can achieve superior performance completion vehicles’ revenue.