Research

Our society is at an inflection point of technological transformation, driven by the next generation of wireless chip- scale systems. Advances in radio-frequency (RF), millimeter-wave (mmWave), and sub-terahertz (THz) integrated circuits and systems have already enabled crucial applications, including 5G communication, autonomous driving radar, gesture recognition, interplanetary internet etc. Today’s RF/mmWave systems are extremely complex; however, design productivity has not kept pace with system complexity, and the full potential of these systems remains unrealized due to three major challenges within the ecosystem: 1) In terms of design methodologies, there exists no generalized approach to efficiently synthesis of such complex high-frequency circuits. The use of templated, narrow design spaces limits performance potential, while the time-intensive, human-guided design process prolongs semiconductor development cycles significantly. 2) In terms of system architectures, the current approach to multi-antenna wireless interfaces have remained largely unchanged for two decades, relying on frequency-static front-ends with narrow bandwidth and application-specific designs, limiting versatility and universality. 3) In terms of circuits and device, mainstream digital-compatible, low-cost silicon- based transistors are increasingly unable to meet the heterogeneous demands of high-frequency performance including power, efficiency, bandwidth, noise etc.

Addressing the multi-tiered challenges of future intelligent wireless chip-scale systems—from bit to THz, and from communication to sensing—requires a transformative co-design approach, which integrates algorithms, architectures, and circuits to achieve better performance and design productivity across diverse applications (ie. integrated communication, sensing and power beaming). The approach is expected to leverage AI to discover new design spaces and accelerate design cycles; to develop universal system architectures adaptable to diverse spectrum and applications; and to enhance circuits and devices through beyond-silicon materials to achieve unprecedented performance. Therefore, my research vision is founded on these three pillars: 1) Smarter wireless chip design beyond human intuition through AI/ML-driven methodologies; 2) Greater wireless system universality through novel phased array architectures; 3) Superior circuits and electromagnetics performance through advanced materials and co-design.