Joint Traffic Signal and Connected Vehicle Control in IoV via Deep Reinforcement Learning

In this paper, we propose to exploit the interconnection in the Internet of Vehicles (IoV) to realize efficient traffic network control, which is indispensable in building intelligent transportation systems (ITS). In addition to control the traffic signals as in conventional traffic network control schemes, we propose to control the detouring behavior of the connected vehicles as well, with an objective to further enhance the traffic efficiency. Specifically, we formulate the joint traffic signal and connected vehicle control problem as a reinforcement learning (RL) problem, the action and state spaces of which are specifically designed to take into account the connected vehicles. To characterize the detouring behavior of the connected vehicles while keeping the decision process simple, we introduce a new concept termed as detouring ratio, which is defined as the fraction of connected vehicles that detour. Moreover, we also design an effective rewarding mechanism that takes into account the impact of the detouring on the network traffic efficiency. By utilizing tools from deep RL, we put forward an efficient algorithm to jointly control the traffic signals and the connected vehicles. Numerical results demonstrate the validity of our proposed models and show that the proposed joint control algorithm can significantly enhance the network traffic efficiency in terms of the travel time and the waiting time.