This research addresses the packet delivery in vehicular communication systems using various reinforcement learning algorithms. The improvement process is executed using Q-learning, SARSA, DQN, and DDQN algorithms. The comparative analysis of algorithmic efficiency aims to identify the most effective reinforcement learning approach for enhancing communication reliability in vehicular networks, contributing to the advancement of intelligent transportation systems. The objectives of the research are to design and develop a simulation environment for VANETs and evaluate the performance of the various reinforcement learning models in improving network performance. The goals of the research are to demonstrate the effectiveness of reinforcement learning in improving network performance in VANETs, compare the performance of various reinforcement learning models in improving network performance, and provide a framework for developing more advanced algorithms for VANETs in the future. By designing a simulation environment tailored for VANETs, we evaluated different reinforcement learning models and found that they notably increase the packet delivery ratio (PDR), thus improving network performance and facilitating the successful delivery of a greater number of packets.
Reinventing Urban Mobility with Reinforcement Learning in Vehicular Ad-Hoc Networks
Lect. Notes in Networks, Syst.
International Conference on Computing and Machine Learning ; 2024 ; Rangpo, India March 29, 2024 - March 30, 2024
2025-01-24
13 pages
Article/Chapter (Book)
Electronic Resource
English
Vehicular Sensing for Improved Urban Mobility
GWLB - Gottfried Wilhelm Leibniz Bibliothek | 2024