Computation Efficient Task Offloading and Bandwidth Allocation in VEC Networks

With the development of intelligent connected vehicles, vehicular edge computing (VEC) systems can provide them with low latency and low energy consumption of task computation and data processing. In this paper, we propose a VEC-aided computing system by offloading tasks from vehicles to the edge computing node via the road side units (RSUs) to improve the computation efficiency. At first, we investigate the computation efficiency for task computation, regarded as the ratio of task bits to the cost in terms of the sum and latency and energy consumption. Then, we formulate a computation efficiency maximization problem to design the bandwidth allocation and task allocation. Due to the dynamically changing channels, available computational resource and bandwidth resource, the formulated optimization problem is non-convex. Therefore, we solve the joint bandwidth and task allocation problem by decoupling the original optimization problem into bandwidth allocation and task allocation sub-problems and solving them iteratively. At last, simulation results are provided to demonstrate that our proposed scheme performs better than the benchmark schemes.