Research

Research Interest

Artificial Intelligence, Computational Game Theory, Mechanism Design, Machine Learning, Multi-Agent Reinforcement Learning

Overview of Research

My research integrates artificial intelligence, mechanism design, and communication networks to build trustworthy, adaptive, and efficient distributed systems. I have developed a coherent programme of work across three interconnected areas:

Trustworthy and Privacy-Preserving AI Systems

I develop intelligent systems that are robust to strategic behaviour and preserve privacy in large-scale, networked environments. My work combines differential privacy, mechanism design, and multi-agent learning to support secure and efficient coordination. Representative contributions include:

1. A network-aware differential privacy mechanism for crowdsourcing systems (Journal of Artificial Intelligence Research 2020 paper);

2. A budget-limited, category-aware incentive mechanism for information gathering (Artificial Intelligence 2021 paper)

3. A privacy-resilient peer-to-peer trading model over directed networks (IEEE Transactions on Smart Grid 2025 paper)

These methods provide foundational tools for secure data sharing, incentive alignment, and privacy protection in distributed AI systems, with applications in communication networks, edge intelligence, and federated infrastructures.

AI-driven Spatio-temporal Prediction and Control in Networked Systems

I design deep learning models to understand, predict, and control complex, nonlinear dynamics in distributed systems. Leveraging Transformer architectures, LSTM networks, and attention mechanisms, I address irregular and non-stationary behaviours in dynamic environments. Applications include:

1. Signal propagation forecasting in communication networks (Nature Electronics 2023, APL Photnoics 2023 (Editor's Pick))

2. Spatio-temporal traffic forecasting with external uncertainties (ECAI 2023)

3. Wind power prediction using Transformer-based models for irregular time series (working paper)

These methods are generalisable to network traffic prediction, adaptive scheduling, and AI-enhanced control across intelligent infrastructures. <\p>

Sustainable and Low-carbon Mobility Systems

I investigate how AI can be embedded into the design and operation of next-generation mobility systems and infrastructures. My research spans from low-latency physical-layer architectures to protocol-level coordination and resource provisioning. Key contributions include:

1. Energy management and charging scheduling (IEEE Internet of Things Journal 2025)

2. HD Map (IEEE Transaction on Mobile Computing)

3. Decarbonisation (working paper)

These works contribute to the development of AI-native mobility systems, enabling applications in edge intelligence, autonomous mobility, and sustainable network services