Joint Offloading and Resource Allocation for Scalable Vehicular Edge Computing

Edge computing assisted autonomous driving technology has become a promising method to satisfy exacting computation requirements of achieving high or even full automation. However, the computation and spectrum resources of multi-access vehicular edge computing (VEC) system are also limited, which may not guarantee the best experience for all users, so we make a tradeoff between resource consumption and user experience. First, according to the characteristics of task scalability in driving assistance applications, we model the problem as maximizing system utility under the deadline constraint and the total resource constraint. Then, we formulate a collaborative computation offloading and resource allocation optimization scheme (JORA). Since the problem is NP-Hard, the JORA scheme eventually solves the problem by the mutual iteration of the two sub-algorithms, which includes offloading strategy and resource allocation. Simulation results prove that the proposed algorithm can effectively improve the system utility.

I. Introduction

Nowadays, autonomous driving has attracted widespread attention in academia and industry, as a desirable technology in connected vehicles[1]. Intelligent vehicles can use the information obtained by various sensors to activate a variety of applications to assist drivers and passengers, such as augmented reality and speech recognition [2]. At present, the computation capacity of the on-board processor is often difficult to meet the requirements of the compute-intensive applications with rigorous delay constraints.