Neuro-Heal: An AI-Powered Autonomous Self-Healing Framework for Resilient IoT Networks
Year: 2025
Type: article
Abstract: This research introduces Neuro-Heal, an AI-powered autonomous self-healing framework designed to enhance the resilience of Internet of Things (IoT) networks. The increasing scale and heterogeneity of IoT deployments—spanning smart cities, industrial automation, and other connected environments—make them prone to node failures, connectivity disruptions, and performance degradation. Traditional fault management approaches, which rely heavily on manual intervention, often result in increased downtime and reduced service reliability. Neuro-Heal leverages predictive analytics and deep learning models to detect potential network faults proactively, classify failure types in real time, and automatically reconfigure network routes to restore connectivity with minimal delay. Reinforcement learning is employed to continuously refine recovery strategies based on prior fault-handling experiences, improving healing efficiency over time. Simulation results using NS-3 and OMNeT++ demonstrate that Neuro-Heal outperforms traditional fault management methods, achieving up to 35% faster recovery, reducing packet loss by 28%, and significantly enhancing overall network availability. This framework supports the development of self-sufficient, resilient IoT infrastructures capable of maintaining uninterrupted service in dynamic and failure-prone environments.
Author: Guugilam, Sai Subash,
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7111328512
FWCI:
Cites:
Cited by:
Related to: