Blockchain-Enabled Ransomware Detection: A Hybrid Model Combining Behavioral Analysis and Machine Learning

The increasing prevalence of ransomware attacks in the present state of affairs posses a significant threat to security and privacy, integrating various sectors, necessitating robust measures for detection and prevention. This project addresses the challenge associated with ransomware attacks, so we propose a hybrid model for the detection along with a security framework using the Solana blockchain. This framework aims to enhance the security of the data and systems, embracing a decentralized structure.The proposed system employs behavior analysis, a feature extractor, and a hybrid machine learning model to analyze data and detect ransomware attacks. Using Program Derived Addresses (PDA), the signature generation algorithm in Solana is used to generate the signature for threat detection. Later, we develop a two-stage hybrid detection model. The evaluation metrics of the framework demonstrate the effectiveness and efficiency, showcasing the improved accuracy, F1 score, and other metrics of the hybrid model. The practical demonstration of the proposed system involves the detection of ransomware attacks with JSON reports.