Delay Optimization in Vehicular Edge Computing with Sensing Information Fusion and Heterogeneous Tasks

Integrated Sensing and Communication (ISAC) technology has attracted extensive attention and been widely studied in automatic drive. However, most of existing works related to automatic drive focused on the single-vehicle intelligence, i.e., each vehicle processes its radar sensing data independently. With the single-vehicle intelligence, the sensing results may not satisfy the accuracy requirements in application scenarios such as obstacle detection and congestion detection. In this work, we consider information fusion within a vehicular mobile edge computing system. Multiple vehicles transmit their unique sensing data to an edge server, which fuses and processes the data to enhance sensing accuracy. Then, we formulate a joint resource allocation and beamforming design problem to minimize the system delay with heterogeneous tasks while satisfying the sensing quality requirements. Due to the non-convexity of the problem, we decouple the problem into two sub-problems and propose an efficient method based on alternate optimization and successive convex approximation to approximate the optimal solution. Simulation results demonstrate that the proposed method improves the performance by 38.1 %, 43.8% and 65.7% in system delay compared with the existing algorithms, i.e., rate maximization, MEC computing and equal allocation.