How should decentralized users share a common channel?
We characterize stability, throughput, and delay in random-access networks—and design control mechanisms that approach their fundamental limits.
Professor Lin Dai · Department of Electrical Engineering · City University of Hong Kong
Theory for a connected world.
I study the mathematical foundations of communication and networking—developing models that reveal why complex systems behave as they do, and principles that remain useful as technologies change.
My recent work pursues two complementary directions: applying reinforcement learning to revisit fundamental access design in communication networks, and using the unified theory of random access I developed to explain and guide reinforcement learning-based access design.
Communication and networking theory · City University of Hong Kong
lindai@cityu.edu.hkAbout Lin →
Research philosophy
Questions before technologies.
Our goal is not simply to improve one protocol. We look for mathematical structures and general laws that explain entire families of communication systems.
When does sensing help—and when does it not?
We develop a coherent theory of random access that unifies Aloha and CSMA, revealing the value, cost, and fundamental limits of carrier sensing.
What makes information genuinely useful?
Beyond conventional data rates, we investigate timely information delivery under fading, delay constraints, and aging information.
How should large wireless systems be organized?
We study network decomposition, clustered cell-free systems, resource allocation, and infrastructure placement through analytical models.
Can learning reveal better access strategies?
We use reinforcement learning to revisit fundamental access problems, then connect learned strategies with analytical theory to explain and optimize their behavior.
Selected work
Four places to begin.
For prospective students who want to understand the intellectual arc of our work, these papers offer an accessible starting point.
A Unified Theory of Random Access
A unified analytical foundation for characterizing stability regions, determining operating regions, and guiding the optimal design of random-access networks.
Related publications
- Lin Dai, “A Theoretical Framework for Random Access: Stability Regions and Transmission Control,” IEEE/ACM ToN, 2022.
- Lin Dai, “A Theoretical Framework for Random Access: Effects of Carrier Sensing on Stability,” IEEE TC, 2025.
Learning to Access—Then Explaining Why
Reinforcement learning for distributed access design, interpreted and guided through analytical theory.
Related publications
- Huaqiang Zhang, Xinran Zhao, and Lin Dai, “Delay-Optimal Random Access: A Learning Framework,” IEEE TC, 2026.
- Nian Peng and Lin Dai, “Multi-Armed-Bandit-Based Framed Slotted Aloha for Throughput Optimization,” IEEE Communications Letters, 2024.
- Lin Dai, “Harnessing the Channel-Capture Phenomenon in Slotted Aloha for Achieving the Optimal Tradeoff between Throughput and Short-Term Fairness,” IEEE TC, 2025.
Delay-Constrained Maximum Information Output Rate
Rethinking reliable and timely communication when information loses value with age and channels vary over time.
Related publications
- Lin Dai, “Delay-Constrained Maximum Information Output Rate in Fading Channels,” IEEE TWC, 2024.
- Lin Dai, “Toward Reliable and Timely Communications: Maximizing the Information Output Rate with Aging Information,” WiOpt, 2025.
Network Decomposition for Large-Scale Wireless Networks
Dividing a large wireless network into parallel subnetworks while limiting inter-subnetwork interference and joint-processing complexity.
Related publications
- Lin Dai and Bo Bai, “Optimal Decomposition for Large-Scale Infrastructure-Based Wireless Networks,” IEEE TWC, 2017.
- Junyuan Wang, Lin Dai, Lu Yang, and Bo Bai, “Clustered Cell-Free Networking: A Graph Partitioning Approach,” IEEE TWC, 2023.
- Funing Xia, Junyuan Wang, and Lin Dai, “Optimizing Clustered Cell-Free Networking for Sum Ergodic Capacity Maximization with Joint Processing Constraint,” IEEE TWC, 2025.
Mentorship & outcomes
Training independent thinkers.
“A successful PhD is not a collection of results—it is the development of a research mind.”
Our students learn to identify fundamental questions, build tractable models, and communicate ideas with precision. Many alumni have continued their research careers in academia.
Selected PhD alumni
Xinghua SunAssociate Professor · Sun Yat-sen University
Yayu GaoAssociate Professor · Huazhong University of Science and Technology
Zhiyang LiuAssociate Professor · Nankai University
Junyuan WangAssociate Professor · Tongji University
Yitong LiAssociate Professor · Zhengzhou University
Wen ZhanAssociate Professor · Sun Yat-sen University
Yue ZhangAssociate Professor · Shantou University
Life in the group
Milestones and memories
Research is demanding, but it is also a shared journey—through graduations, discussions, friendships, and time together beyond campus.
Teaching
Foundations that make new ideas possible.
My courses connect mathematical principles with the design of real communication systems, from core undergraduate concepts to advanced wireless technologies.
Principles of Communications
Fundamental principles of analog and digital communications, including deterministic and random signal analysis, amplitude and frequency modulation, sampling and quantization, and digital modulation and demodulation.
Course information →Data Communications and Networking
Core concepts behind data networks, layered architectures, communication protocols, and network performance.
Course information →Wireless Communication Technologies
Fading, diversity, channel capacity, and centralized and distributed multiple access from a systems perspective.
Course information →Selected talks
Ideas in conversation.
Invited lectures, webinars, and research talks offer an accessible route into the questions, analytical frameworks, and design principles behind our work.
Intelligent Random Access for Next-Generation IoT Networks
How analytical theory and learning-based methods can work together to shape distributed access for future IoT networks.
A Unified Theory of Random Access
A common analytical foundation for understanding and optimizing random-access schemes across performance metrics.
Random Access for Machine-to-Machine Communications: Challenges and Prospects
Challenges, analytical foundations, and prospects for massive distributed access · IEEE Distinguished IoT Webinar.
Prospective students
Do fundamental questions keep you curious?
I welcome students who want to understand not only how a system works, but why—and what its limits must be.
Introduce yourselfYou may thrive here if you…
01Enjoy probability, optimization, information theory, stochastic processes, and reinforcement learning.
02Like turning complicated systems into clean mathematical questions.
03Value depth, rigor, persistence, and clear thinking over fashionable labels.
04Want to grow into an independent researcher who can identify and define new problems.
When you write
Please include your CV, transcripts, research interests, and a few sentences about one theoretical problem or paper that genuinely interested you—and why.