A Hybrid Deep IoT Network-Driven Anomaly Detection using Multi-Scale Deep Representation Learning

Due to the exponential increase in IoT device production, the IoT (Internet of Things) business has experienced rapid expansion on the market, which gives attackers a larger attack surface from which to launch potentially more devastating assaults. There has been a rise in cyber-attacks. When intruders perform cyber-attacks utilizing unique and inventive ways, many of these attacks have effectively fulfilled the maliciousintentions. Conventional machine learning approaches seem ineffective in the context of unanticipated network technology and various penetration strategies. The introduction of new vulnerabilities is a result of cyber-physical applications leveraging Internet of Things (IoT) devices. Because of the cross-domain, cross-layer, and multidisciplinary nature of the emerging security and dependability concerns, a comprehensive solution is required.