Abstract

Blockchain-Based Secure Data Sharing in IoT with Machine Learning-Based Hybrid Anomaly Detection is an advanced hybrid framework that works with IoT security, blockchain trust, and ML intelligence combined. Internet of Things (IoT) devices are growing rapidly across smart homes, healthcare, smart cities, and industrial systems, but with the rapid growth here comes serious security issues, like unauthorized data access, data tampering, fake sensor data injection, and distributed attacks on IoT systems. A big issue is that data is stored in centralized servers, which is like a single point of failure, so the blockchain technique is used to solve this problem. It provides a decentralized ledger and tamper-proof data storage, ensuring secure data sharing in a trustless environment. But blockchain itself is not a complete solution because malicious or abnormal data can also be stored in the blockchain. Major approaches of this system are enhanced security, data integrity, and trustworthy data sharing where only clean and verified data is shared. The main goal is to build a lightweight, scalable and secure framework that will work in IoT environments for real-time anomaly detection while enabling secure data sharing. The proposed model addresses limitations like high computational cost, false positives, and lack of real-time adaptability by building a lightweight ML model with edge computing integration and efficient blockchain consensus mechanisms.