Radio Resource Allocation in Vehicle-to-Everything with Different Optimization Objectives

As a core technology of intelligent transportation systems, vehicle-to-everything (V2X) enables a highly interconnected traffic environment through vehicle-to-vehicle, vehicle-to-pedestrian, vehicle-to-infrastructure, and vehicle-to-network communications. Radio resource allocation (RRA) is a key technology to guarantee the V2X communications, including the efficiency and reliability improvement, the latency reduction, etc. Currently, there is a lack of systematic reviews in terms of optimization objectives for RRA in V2X. This paper provides a comprehensive review of researches on single-objective and multi-objective optimization for RRA in V2X. Single-objective optimization mainly focuses on improving key performance indicators such as reliability, throughput, and latency, whereas multi-objective optimization aims to achieve the balance among the above-mentioned indicators. The applications and optimization effects of machine learning, deep learning and reinforcement learning methods are then summarized. Based on this analysis, this paper also explores future research directions. The systematic analysis, method classification, and practical guidance are offered for researchers, contributing to the future development of RRA in V2X networks.