Intelligent Resource Allocation Scheme Using Reinforcement Learning for Efficient Data Transmission in VANET.

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Vehicular ad hoc networks (VANETs) utilize Wireless Access in Vehicular Environments (WAVE) standards for vehicle communication.
• The current IEEE 802.11p WAVE standard uses fixed CCH interval (CCHI) and SCH interval (SCHI) durations, limiting adaptability to traffic loads.
• Fixed intervals lead to network performance degradation and packet collisions due to channel congestion.

Purpose of the Study:

• To propose an adaptive resource allocation technique for efficient data transmission in VANETs.
• To dynamically adjust SCHI and CCHI to enhance network performance under varying traffic conditions.
• To reduce data collisions and optimize channel access using intelligent algorithms.

Main Methods:

• Developed an adaptive resource allocation technique that dynamically adjusts SCHI and CCHI.
• Implemented a reinforcement learning (RL) based intelligent channel access algorithm.
• Simulated the proposed scheme to evaluate its performance against existing methods.

Main Results:

• The proposed adaptive technique effectively manages channel resources.
• Reinforcement learning optimizes backoff distribution, reducing data collisions.
• Simulations show significant improvements in network throughput and transmission delays.

Conclusions:

• Adaptive adjustment of CCHI and SCHI is crucial for efficient VANET operation.
• Reinforcement learning provides an intelligent approach to channel access in VANETs.
• The proposed scheme offers a promising solution for high-throughput, low-delay VANET communication.