Deep Reinforcement Learning Based Intelligent Approach to Channel Allocation for Effective Communications in Vehicle Platoon Systems

Vehicle platoons are the applications that provide greener, safer, faster, and economical transport as part of modern intelligent transportation systems. The advent of deep learning and reinforcement learning made it possible to address standard communication problems effectively using data-driven techniques. In this short research paper, we discuss the realm of vehicular platoon where deep and reinforcement learning can be applied and requires attention. We also propose an application of deep reinforcement learning to solve the resource allocation problem intelligently, addressing the specific needs of vehicle platoons. This application should pave the way for using deep reinforcement learning implements in vehicle platoon communications per se. We also discuss the scope for improvements required to well adopt deep reinforcement learning in this regime.

I. Introduction

In the last decade, on average, eighty million commercial and non-commercial vehicles were manufactured every year worldwide [1]. It indicates that a large number of vehicles have been registered for on-road operations every year. Besides, there is a significant increase in the migration of people toward urban areas. It brings about transportation-related problems such as traffic conditions, pollution, and safety for modern men in day-to-day life. Thus, it is time to update the current transportation systems accordingly.