MSc Thesis Defense "Transmission Power based Congestion Control using Q-Learning algorithm in Vehicular Ad Hoc Networks (VANET)" By Pooja Chandrasekharan

Thursday, September 7, 2023 - 13:00 to 15:00
The School of Computer Science is pleased to present…
MSc Thesis Defense Announcement
Transmission Power-based Congestion Control using Q-Learning Algorithm in Vehicular Ad Hoc Networks

MSc Thesis Defense by: Pooja Chandrasekharan
Date: Thursday, September 7, 2023
Time: 1:00 PM – 3:00 PM
Location: Essex Hall, Room #122

Vehicular ad hoc networks (VANETs), an emerging wireless technology used for vehicle-to-vehicle and vehicle-infrastructure
communication, are essential components to reduce road accidents and traffic congestion in Intelligent
Transportation Systems (ITS). It also provides additional services to vehicles and their users. However, vehicles must
balance awareness and congestion control in a dynamic environment to efficiently transmit basic safety messages (BSMs)
and event-driven warnings. The limited channel capacity makes the reliable delivery of BSMs a challenging problem for
VANETs. This paper aims to optimize the performance of VANETs by effectively managing channel load and reducing
congestion by maintaining the channel busy ratio (CBR) near the threshold value of 0.6. This is resolved using a
transmission power-based congestion control algorithm that employs a Markov decision process (MDP) and solves it using
a Q-Learning algorithm. The algorithm uses varying transmission power levels to lower the channel busy ratio while
maintaining high awareness for surrounding vehicles. According to simulation results for various traffic scenarios, the
suggested technique chooses a suitable transmission power depending on the present channel circumstances to achieve a
balance between awareness and bandwidth usage. The findings show that the proposed strategy reliably maintained the
channel load at or near the stipulated level without surpassing it for both low and high traffic densities.

Keywords: VANET, Congestion Control, Reinforcement Learning, Q-Learning

Thesis Committee:
Internal Reader: Dr. Shafaq Khan
External Reader: Dr. Animesh Sarker
Advisor: Dr. Arunita Jaekel
Chair: Dr. Christie Ezeife