DUAL Group

The Distributed compUting, optimizAtion, and Learning (DUAL) Group develops scalable, efficient, and reliable learning systems at the intersection of distributed computing, mathematical optimization, and artificial intelligence.

Our research links theory with deployable systems for real-world, resource-constrained environments. We work on federated and edge learning, distributed optimization, robust communication, and efficient large language model inference.

Research

We design algorithms and systems that improve the efficiency, robustness, fairness, and interpretability of modern AI systems. Current themes include large language models, wireless semantic communication, federated learning, and real-time collaborative inference.

Projects

Our projects span theoretical advances and practical systems, including WaSeCom for robust wireless semantic communication and GOODSPEED for efficient and fair LLM inference at the edge.

Team

DUAL brings together researchers, students, and collaborators across machine learning, distributed systems, optimization, and edge intelligence. We mentor emerging researchers while advancing practical and theoretical foundations for intelligent systems.

highlighted publication