MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641113117 A) filed by Dr. M. G. R. Educational & Research Institute on September 21, 2026, for A System And A Method Of Edge-Ai Intrusion Detection For Resource-Constrained Iot Devices.
Inventors include A. Srikrishnan; K. Sudhaman; N. Lakshminarayanan; B. Ramesh; R. G. Balaji; and K. Manivannan.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: Disclosed is a system (100) for edge-AI based intrusion detection and autonomous mitigation in a resource-constrained Internet of Things device. The system (100) comprises a resource-constrained microcontroller platform (102), a data acquisition module (104), a feature extraction module (106), an edge-AI inference engine (108), a decision-making module (110), a response and mitigation module (112), an optional cloud reporting interface (114), and a secure update and configuration module (116). Operational and communication data is continuously monitored and converted into a compact feature vector using sliding-window statistical and behavioural analysis. The edge-AI inference engine (108) executes a TinyML-optimized machine-learning model locally using quantization, pruning, or memory-efficient inference techniques. Based on the inference result, the decision-making module (110) classifies device behaviour and the response and mitigation module (112) autonomously initiates corrective security actions. Raw monitored data remains local, thereby enabling low-latency, privacy-preserving, bandwidth-efficient, and cloud-independent intrusion detection and response for constrained IoT environments. Fig. 1 is the reference figure.
Disclaimer: Curated by HT Syndication.
