A Secure Framework for IoT Applications Using Blockchain and Artificial Intelligence

This research addresses key security and privacy challenges in IoT by proposing a multi-layered security framework. It includes formal verification of IoT protocols using the Scyther tool for secure communication, a hybrid intrusion detection system using LSTM and Deep Reinforcement Learning, and an ensemble model combining CNNs with a Quantum Neural Network to improve detection accuracy. Additionally, a blockchain-based system with Improved PCA and a GRU–Deep Belief Network classifier enhances accuracy and reduces false positives. Experiments demonstrate strong performance, with detection accuracy up to 97.26% and improved reliability through blockchain integration.