Autonomic Routing Performance Using AI with Trust and Fuzzy Logic

Mobile Ad Hoc Networks (MANETs) suffer from dynamic topology, resource limitations, and vulnerability to routing attacks such as blackhole and wormhole. Conventional protocols like AODV, DSR, and DSDV lack adaptability and struggle to maintain performance in such environments. This research proposes an autonomic AI-driven routing model that integrates Q-learning, trust evaluation, and fuzzy logic to enhance performance, reliability, and security. The proposed AI-AODV dynamically learns optimal routes, filters malicious nodes through trust scoring, and handles uncertain conditions using fuzzy inference. Simulation results show significant improvements in packet delivery ratio, delay, throughput, convergence time, and attack detection accuracy compared to traditional protocols. These findings demonstrate that AI-based autonomic routing is a promising direction for next-generation secure MANETs. “The proposed model was validated through simulation-based experimental analysis, not real classroom or real-world deployment. ” Unlike existing approaches, this work integrates reinforcement learning, trust evaluation, and fuzzy inference into a unified autonomic routing framework.”