A Decentralized Architecture Integrating Blockchain and Artificial Intelligence to Ensure Security, Privacy, and Trust in the Internet of Things
A Decentralized Architecture Integrating Blockchain and Artificial Intelligence to Ensure Security, Privacy, and Trust in the Internet of Things
Rezvan Mahmoudie1
1) Department of Computer Engineering, Asadabad Branch, Islamic Azad University, Asadabad, Iran
IOI Code
:
AAHD-ABKAGH
Publication :
National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
The simultaneous advancement of the Internet of Things (IoT), blockchain, and artificial intelligence (AI) has ushered in a new era of intelligent decentralized systems. However, existing centralized architectures, due to their reliance on cloud servers, face three fundamental crises: cybersecurity threats, privacy violations, and the lack of inherent trust among heterogeneous nodes. This paper proposes a fully innovative three-layer architecture called DAIoT, which for the first time integrates trust-based federated learning, self-evolving smart contracts, and an adaptive consensus system within a coherent framework. The key innovation of this research lies in the design of a hybrid consensus mechanism that dynamically switches between Proof of Authority (PoA) and Practical Byzantine Fault Tolerance (PBFT) algorithms based on the network s risk level, thereby improving energy efficiency by up to 87% and reducing end-to-end latency to below 800 milliseconds. Furthermore, the decentralized intelligence layer, leveraging reinforced federated learning, not only fully preserves the privacy of raw data but also ensures proactive network security through real-time detection of behavioral anomalies in nodes. The results of extensive simulations conducted in the NS-3 environment and Hyperledger Fabric on a smart city scenario with 500 IoT nodes demonstrate that the proposed architecture, compared to the best existing approaches, increases security by 42%, privacy by 38%, and trust levels by 51%, while reducing computational overhead by up to 65%. This architecture can be implemented as a reference model for critical IoT systems (healthcare, smart cities, critical infrastructures) and paves a novel path toward realizing secure, transparent, and decentralized next-generation Internet of Things.
Keywords : decentralized architecture adaptive blockchain reinforced federated learning hybrid consensus secure iot privacy oriented trustworthiness daiot