ABSTRACT
The IEEE 802.11p stack of Vehicular Ad-hoc Networks (VANETs) is responsible for efficient channel access based on the network's throughput, latency, and fairness requirements. Due to collision, the IEEE 802.11p cannot guarantee efficient data transmission in a dense vehicular network. The Contention Window (CW) plays a significant role in the design of Media Access Control (MAC) protocol for vehicular communication, focusing on collision reduction. The proposed work dynamically adjusts the CW parameter to maximize the throughput of a vehicular network. To accomplish this, a Reinforcement Learning (RL) framework compensates for actions that result in high utility by using local channel observations to overcome the absence of system knowledge. The proposed model implements a learning-based IEEE 802.11p protocol for the MAC channel control approach. The actor-critic model effectively learns the VANET environment to provide the best reward. The simulation result shows the proposed learning-based CW mechanism significantly improves the throughput requirements of the traditional IEEE 802.11p standard.
