A Multi-Objective Optimization Approach for Roadside Unit Deployment Strategy in IoV

Roadside unit (RSU) deployment plays a crucial role in enhancing the quality of service (QoS) in Vehicular Ad-hoc Networks (VANET). To determine the optimal quantities and locations of deployed RSUs for maximum effectiveness, previous research categorized this as a single-objective optimization problem, considering only the performance indicators for vehicle-to-infrastructure (V2I) communication and solving it using heuristic algorithms or learning-based algorithms. Nonetheless, in the event of information interaction among RSUs as a supplementary component to V2I communication, the process of data transmission and dissemination will achieve a heightened level of comprehensiveness and efficiency within the network. In this paper, we construct a multi-objective roadside unit deployment model (MORD) that takes both V2I communication and data interaction between RSUs into account. The optimization problem of MORD is solved by a multi-objective evolutionary algorithm. We perform a series of simulation evaluations in the ideal grid scenario and the real road network scenario. The results prove that compared to other schemes, the proposed algorithm can better balance the relationship between various objectives in MORD. Moreover, the algorithm can provide decision makers with non-dominated solution sets to choose an appropriate deployment scheme that is more suitable for the actual conditions.