NYCU-ISAC
Prompted by 3GPP TR 22.837, which highlights 32 key applications—many in smart cities and intelligent transportation—this project addresses the growing need for large-scale Integrated Sensing and Communication (ISAC) networks. While ISAC has seen significant research over the past five years, most efforts focus on isolated technologies, with little attention to scalable, perceptive wireless systems.
Funded by NSTC, Taiwan, this project was officially launched in Aug. 2025 to bridge that gap, aiming to develop AI-driven ISAC networks that unify sensing and communication for real-world deployment in smart urban and transportation environments.
News and Activities
We published the paper entitled "Deep unfolding learning-based beamforming design for multi-user MIMO-OFDM integrated sensing and communication systems" at IEEE GLOBECOM 2025.
This work targets the runtime bottleneck of optimization-based beamforming for multi-user MIMO-OFDM ISAC. Building on a WMMSE-SCA formulation, we unfold the iterative procedure into a lightweight neural architecture (UBeD) and train it with an unsupervised loss directly tied to the ISAC objective. Training tricks such as RD-bin weighting and dynamic tradeoff control further stabilize convergence. Simulation show that UBeD achieves superior sensing-communication tradeoffs while slashing per-design runtime from hours to seconds and maintaining scalability as system dimensions grow.
Ming-Chun delivered a tutorial talk entitled “MIMO-OFDM ISAC Systems: Fundamentals, Recent Advances, and Opportunities” at the 25th Asia-Pacific Network Operations and Management Symposium.
Li-Chun, Ming-Chun, and Jerry visited the Department of ECE, Texas A&M University
Li-Chun, Ming-Chun, and Jerry visited Network Sensing Lab @UTS
Li-Chun, Ming-Chun, and Jerry visited UNSW
We published the paper entitled "On the scaling law tradeoff of integrated sensing and communication networks" at IEEE ISIT 2025.
In this work, we fully characterize the optimal scaling-law tradeoff between communication throughput and sensing range of large ad hoc ISAC wireless networks. Our results show that by slightly reducing throughput, we can significantly extend the sensing range. Specifically, the improvement in sensing range is proportional to the degree of throughput reduction, scaled by the relative difference in how signals weaken over distance in communication versus sensing.
The NYCU ISAC team was officially formed, comprising five professors at NYCU and three international collaborators: Jinhong Yuan (UNSW, Australia), Vahid Jamali (TU Darmstadt, Germany), and Geoffrey Li (Imperial College London, UK).