This research theme focuses on developing trustworthy, explainable, robust and privacy-preserving AI systems through federated learning, split learning, explainable AI (XAI), privacy-enhancing technologies and blockchain-based trust mechanisms. Our research investigates secure collaborative learning, protection against data and model poisoning, privacy leakage prevention and transparent AI decision-making. We also explore differential privacy, homomorphic encryption, secure multiparty computation and robust aggregation mechanisms to protect sensitive information and distributed AI models.
A major research direction involves integrating blockchain and distributed ledger technologies (DLT) with collaborative AI to enable decentralised trust management, secure data sharing, verifiable model updates, data provenance and accountable AI operations. The objective is to establish secure, transparent, privacy-preserving and trustworthy intelligent systems for next-generation networks and distributed digital ecosystems.
Key research topics
- Privacy-preserving federated learning and split learning.
- Secure and privacy-preserving collaborative intelligence for 5G/6G networks.
- Differential privacy, homomorphic encryption and secure multiparty computation.
- Explainable AI (XAI) for cybersecurity and intelligent network management.
- Adversarial robustness against data poisoning, model poisoning and inference attacks.
- Secure aggregation and trustworthy distributed AI model training.
- Blockchain-enabled federated learning and decentralised trust management.
- Blockchain-based identity management, authentication and secure data sharing.
- Data provenance, model integrity and verifiable AI operations.
- Reputation-based trust evaluation and smart-contract-enabled accountability.
- Privacy-by-design and trustworthy AI-native 6G architectures.
- Non-fungible tokens (NFTs).