Deep Learning-Driven SIoT Security: Detecting Anomalies and Community Attacks with SPY IoT Devices

The rapid growth of Social Internet of Things (SIoT) networks, new security vulnerabilities have evolved, including irregularity detection and community-based hacks. Given the large degree of decentralization in SIoT systems, traditional security methods are frequently ineffective. The study employs a deep learning (DL)-based security solution that detects anomalies and community attacks in real time using SPY IoT sensors. The proposed approach employs AutoEncoders(AEs) for SIoT behaviour, Convolutional Neural networks(CNN)+Long Short Term Memory(LSTM) for spatial and temporal features and Reinforcement Learning(RL) for adaptive Learning. The novel approach described in this study improves the detection of various attack situations, hence increasing the overall security and resilience of SIoT networks. The findings of the proposed study outperforms with an accuracy of 97.2 compared to other state o art methods. This study is related to building and developing intelligent and self-adjusting safety procedures for the future Internet of Things.