Asrld: Adaptive Secure Reinforcement Learning-Based Route Discovery

Asrld: Adaptive Secure Reinforcement Learning-Based Route Discovery

Authors- B. Swetha, U. Madhu Sandhiya, U. Nowmitha Priya, Dr. J. Viji Gripsy

Abstract-Efficient route discovery in mobile ad hoc networks (MANETs) and vehicular ad hoc networks (VANETs) is a critical issue due to considerations such as augmented latency, energy consumption, and control message overhead. Traditional routing systems, though optimal in certain environments, tend to perform suboptimally in resource-constrained and dynamic environments. The work introduces the Adaptive Secure Reinforcement Learning-Based Route finding (ASRLD) model, combining adaptive threshold-based broadcasting, reinforcement learning, and blockchain security mechanisms to improve route finding efficiency. The ASRLD model reduces unnecessary route request (RREQ) messages, optimizes path choice, and improves overall routing effectiveness. Key components include an adaptive broadcasting mechanism that adjusts rebroadcasting probability based on network conditions, a Q-learning agent that self-detects best paths, an energy-aware routing scheme that considers residual energy levels, and blockchain-based security features for safe route establishment. Simulation results using NS-3 show significant improvement over traditional AODV and DSR protocols in terms of reduced routing overhead, enhanced packet delivery ratio, lower end-to-end latency, and optimized energy consumption. The proposed model delivers a robust, energy-aware, and secure methodology for route detection under challenging network scenarios.

DOI: /10.61463/ijset.vol.12.issue6.955