A novel hybrid fuzzy logic and federated learning framework for enhancing cybersecurity and fraud detection in IoT-enabled metaverse transactions

Increasing integration of the Internet of Things (IoT) with virtual environments like the Metaverse has opened up new avenues in the applicability of technologies but faces severe challenges to security and fraud detection. Most of the existing frameworks are incapable of efficiently managing trust and detecting fraudulent activities in a decentralized, resource-constrained environment. In this article, a novel framework of cybersecurity is proposed that integrates hybrid fuzzy logic-based Trust Management with a decentralized model of Federated Learning. The proposed approach assesses and manages at runtime to maintain the degree of trust using fuzzy logic in dynamic conditions of the Metaverse. The optimized federated learning model for IoT devices implements lightweight algorithms with hierarchical aggregation that reduce computational and communication overhead to enhance fraud detection capabilities. The performance evaluation is conducted on different attack scenarios like $O{n}_{\text{off}}$, Whitewashing, DDOS, and Bad Mouthing attacks. It is observed that the proposed approach performs better in comparison with existing approaches by achieving a 0.93 trust score value in low-network scenarios. It reduces computational energy consumption by 25%, thus proving the effectiveness and strength of the framework in fraud detection within IoT-enabled Metaverse environments.