Two Tier Hybrid Routing Schemes for Vehicular Ad Hoc Networks Based on Reinforcement Learning

This research addresses the issue of routing messages from vehicles to specific geographical locations in VANETs. To overcome this challenge we propose a hybrid system consisting of traffic-aware located RSUs and a reinforcement learning routing strategy. Our proposed method includes two main tiers. We first divided the geographical area into equal-sized grids; then using the traffic flow patterns of the area, the protocol will suggest the optimal locations for RSUs with the objective of minimizing the number of RSUs in the system to cut down the cost while improving the delivery ratio at the lowest possible delay.