Dr. AVLN Sujith Kumar, Devender Avula 02 07 This review explores the evolution of IoT Intrusion Detection Systems IDS transitioning from traditional anomalybased models to advanced intelligent detection frameworks It synthesizes findings from fifteen peerreviewed studies focusing on graphbased neural networks distributed learning architectures and resourceefficient optimization techniques The review highlights innovations like analogical training generative model creation ensembledistillation methods and protocolspecific detection strategies It evaluates datasets including CoAPDoS and other benchmarks to assess detection accuracy The study concludes that modern IDS technologies are becoming protocoladaptable and more intellectually robust aiming to balance computational performance detection precision and effective protection across diverse and resourceconstrained IoT network environments Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026