An energy-efficient V2X Resource Allocation for User Privacy Protection: A Learning-Based Approach

Abstract

In Intelligent Transportation Systems(ITS), vehicles are mainly considered to travel in vehicle groups on highways, and in cities, C-V2X is more prone to data eavesdropping when communi-cating at vehicle convergence sections such as intersections, and with limited spectrum resources, communication quality needs to be guaranteed and enhanced. It is a great challenge to improve the spectrum efficiency (SE) and energy efficiency (EE) of the V2X network while satisfying the C-V2X confidentiality rate. To solve this problem, this paper proposes a deep reinforcement learning based SE and EE enhancement algorithm. It establishes an objective optimization function that considers both SE and EE, and uses the secrecy rate of C-V2X as the key constraint of this function. The optimization problem is transformed into a spec-trum and transmission power selection problem for V2V and V2I links using the Deep-Q-Network ( DQN ). The simulation results show that the overall efficiency and V2V link secrecy rate of the proposed algorithm is significantly higher than that of the ran-dom algorithm when the number of vehicles is between 20 and 40, with an average secrecy rate increase of 82.86%.