Efficient Repeated Service Caching Scheme in Ultra-dense Vehicular Edge Computing Networks

With the exponential growth of communication and computation requirements for both the vehicles and passengers (e.g., self-driving, online game) in urban traffic system, the ultra-dense vehicular edge computing network (Ud-VECN) is consid-ered as a promising technology. Through employing computation and caching resources at the edge of wireless access networks (e.g., RSUs), Ud-VECN is able to provide the high quality of services (QoS) for numerous vehicle users. In this paper, we explore an Ud-VECN architecture, the overlapped coverage areas of adjacent RSUs are considered. To balance the computation and storage resources in the edge nodes, we design a repeated service caching framework to satisfy the task completed delay constraints of multiple vehicle users. To minimize the system energy consumption, we formulate a mixed-integer programming problem that jointly considers the service caching and task offloading decisions. Then, an optimal solution method that leveraging the genetic algorithm (GA) and a distance-based algorithm is proposed. Simulation results show the outstanding performance of proposed algorithm.