Deep Q-Learning on Internet of Things System for Trust Management in Multi-Agent Environments for Smart City

Abstract

Smart Cities are vital to improving urban efficiency and citizen quality of life due to the fast rise of the Internet of Things (IoT) and its integration into varied applications. Smart Cities are dynamic and complicated, making trust management in multi-agent systems difficult. Trust helps IoT devices and agents in smart ecosystems connect and cooperate. This study suggests using Deep Q-Learning and Bidirectional Long Short-Term Memory (Bi-LSTM) to manage trust in multi-agent Smart City settings. Deep Q-Learning and Bi-LSTM represent long-term relationships and temporal dynamics in the IoT network, enabling intelligent trust-related judgments. The architecture supports real-time trust assessment, decision-making, and response to smart city changes. The suggested solution improves dependability, security, and trustworthiness in the IoT system's networked agents. A complete collection of studies utilizing real-world IoT data from a simulated Smart City evaluates the system's performance. The Deep Q-Learning and Bi-LSTM technique surpasses existing trust management approaches in dynamic, complicated multi-agent environments. The system's capacity to adapt to changing situations and improve decision-making make IoT device interactions more dependable and trustworthy, helping Smart Cities expand sustainably and efficiently.