Hybrid Attribute-Based Access Control Framework for Intelligent Computing in Consumer IIoT

The protection of consumer and industrial IoT (CIoT/IIoT) systems against modern cyber threats including AI-based attacks needs suitable access control systems. Current access control models fail to provide flexible operation and strong security because they do not support real-time policy changes and multiple data sources. The research establishes HY-ABAC as a hybrid Attribute-Based Access Control system which unites blockchain technology with machine learning algorithms (policy gradient and random forest) to enhance security and operational speed and enable flexible policy administration for IoT systems. The system depends on a random forest classifier for making exact decisions and policy gradient for creating and updating policies in real-time. The system uses a local blockchain to track access transactions through decentralized and transparent and immutable methods which establish trust and defend against unauthorized changes. The proposed model supports the complete integration of sensing technology with communication systems and computing power and control mechanisms for secure intelligent system operation in consumer IoT. Using the ToN-IoT dataset, the model achieves 98% accuracy and processes 25% more transactions per block than Fabric-IoT and 11% more than ABAC-HLFBC. It further reduces policy configuration and verification time compared with Fabric-IoT, ABAC-IoT, ABAC-HLFBC, and DACSB.