Communication is more than technology; it is a foundational pillar of human civilization. For decades, its evolution, from theory to the pervasive reality of 5G and WiFi, has been a triumph of science and engineering, a testament to the pioneers who built the invisible infrastructure of our modern world.
That very success now invites a profound question: Has the era of fundamental innovation in communication passed? Some observe a mature field and suggest the core work is complete.
We see a different reality. The discipline is not ending; it is entering its most critical phase. Like physics or mathematics, communication is a bedrock science. Its prior revolutions constructed the connected world we rely upon; its next evolution will enable and define the future itself. To view this field as "solved" is to mistake a foundation for a finished building. The foundation must now be strengthened, expanded, and re-imagined to support the structures of tomorrow: the networks of distributed intelligence, the Internet of Agents, and paradigms of interaction we have yet to conceive.
Our Research Agenda
The Science of Perfection ('X' for Communication)
Current networks are a monumental achievement, but they are not static endpoints. We treat standards like 5G and WiFi as living, adaptive platforms. Our work here focuses on their continuous transformation through strategic integration. We ask: How can the latest breakthroughs in artificial intelligence, advanced computing, and novel materials be fused into the communication fabric to make it radically more efficient, intelligent, and resilient? This is "X for Communication": the deliberate application of adjacent revolutions to perfect our most critical global infrastructure.
The Architecture of the Future (Communication as an Enabler)
Beyond evolving existing systems, we are designing the fundamental communication primitives for tomorrow's breakthroughs. The next frontiers in science and technology, from distributed AI and swarm robotics to brain-computer interfaces and planet-scale sensing, are constrained not by ideas, but by their need for new forms of dialogue. We engineer the protocols, algorithms, and system designs that allow distributed intelligences to collaborate, that turn fragmented data into coherent insight, and that make emerging technologies truly scalable and synergistic. Here, we are not merely improving communication; we are building the connective tissue of future progress.
Research Topics
Topic 1 (Source): DeepJSCC and DL-based compression
We leverage DL to design Joint Source-Channel Coding schemes. This allows communication systems to be optimized directly for source-related metrics, moving beyond traditional bit-error-rate objectives and enabling goal-oriented semantic communication.
Topic 2 (Baseband): Wireless Signal Processing
We focus on optimal transceiver design for the physical layer by employing advanced model-based signal processing techniques. Our work addresses the core challenge of reliable communication by tailoring solutions to the specific characteristics of the wireless channel.
Topic 3 (Transmission): Flexible Antenna Systems and Electromagnetic Co-Optimization
We pursue holistic communication system optimization by jointly designing across domains: from electromagnetic theory and antenna design to wave propagation and baseband processing. This integrated approach unlocks performance gains beyond isolated component design.
Topic 4 (Feedback): Rich-ARQ
We innovate in feedback mechanisms by embedding rich, beyond-1-bit information into acknowledgment packets. A key focus is utilizing deep learning to design efficient feedback channel codes, significantly improving system adaptation and reliability.
Topic 5 (MAC and NET): Efficient Access and Networking
We tackle critical challenges at the Medium Access Control and Network layers. Our research aims to solve fundamental pain points related to resource allocation, access coordination, and scalable networking in modern wireless systems.
Topic 6 (Communication as an Enabler): Multi-Agent Systems
Our research explores communication in two key agent-related areas: 1) Designing communication protocols to resolve cognitive misalignment among multiple agents; 2) Designing intelligent agents capable of orchestrating IoT systems and executing complex task planning.
Topic 7 (Communication as an Enabler): Federated Learning via AirComp
We address the communication bottleneck in Federated Learning by leveraging the Over-the-Air Computation framework. This technique allows efficient model aggregation directly in the wireless domain, dramatically improving the scalability of distributed learning over networks.