Blockchain-Driven Cybersecurity Framework for Smart Homes: Integrating IoT and Machine Learning for Secure Automation

Abstract

With the increasing use of Internet Things (IoT) devices in smart homes has resulted in issues like data security, privacy, and how we manage everything. The Centralized architecture which we have been using can fail easily and bring down the whole system, in this research we are using a new approach where the smart home devices are handled using blockchain. This approach makes use of the decentralized approach and data replication of the blockchains so that the system is not controlled by one central server instead data is stored in distributed ledger so it would not be changed by the smart home data. With the use of IoT sensors and machine learning models (like Support Vector Regression, Linear Regression, Random Forest, and ARIMA), appliance monitoring and predictions are more correct, reaching an accuracy of 95.77%. The system uses tough security with cryptographic hashing and stores data across a decentralized network using IPFS. The system implementation costs and benefits analysis in Pakistani Rupees (PKR) confirms that this idea is worth it financially, paying itself off in 25 months. The plan comes with an easy-to-use dashboard for homeowners to watch and control things as they happen. Tests confirm that the system does a good job at boosting security, saving energy, and giving homeowners useful information.