Design and Optimization of Edge Computing for Data Fusion in V2I Cooperative Systems

Real-time data fusion combining with information from vehicles and roadside unit (RSU) is a promising solution to promote traffic safety and efficiency. In this paper, we design a multi-source data fusion scheme in edge computing-enabled vehicle-to-infrastructure (V2I) cooperative systems, where data fusion can be processed at RSU or vehicle. In order to balance the tradeoff between the vehicle speed and the fusion range, we define a new performance metric, namely fusion gain. We formulate the jointly data offloading decision, fusion range and computing resource allocation problem for maximizing the system fusion gain while minimizing local and edge computational resource consumption. We reformulate the stated problem and design a substitution-knapsack algorithm to reach a sub-optimal solution. Numerical results show that the proposed scheme has a significant performance gain and effectively promotes system utility in varying traffic environments.