Design and Implementation of a Consensus-Driven Malware Analysis System using Deep Learning and Blockchain

This research presents a decentralized malware analysis framework that integrates deep learning techniques with blockchain-based consensus mechanisms. Traditional centralized malware detection systems often suffer from single points of failure and lack transparency in decision-making. The proposed system performs static feature extraction from Android applications and trains multiple deep learning models including MLP, DNN, CNN, and LSTM for malware classification. To enhance reliability and prevent manipulation, the system uses a consensus-driven approach where multiple detection nodes analyze applications independently. A Practical Byzantine Fault Tolerance (PBFT) consensus mechanism is used to aggregate the predictions of distributed nodes. Blockchain technology ensures transparency, immutability, and secure storage of malware analysis results. Experimental evaluation demonstrates that the proposed system improves detection reliability and robustness in distributed environments. This work highlights the potential of combining deep learning with blockchain technology to build a secure and trustworthy decentralized malware detection system.