IRS‑assisted wireless to enable dynamic coexistence of communication and sensing systems with GNN‑based channel understanding, physics‑regulated DRL resource allocation, and fairness across co‑located networks.
Next‑generation networks face spectrum congestion, path loss at mmWave/THz, and increasing coexistence with radar sensing. This project leverages Intelligent Reflecting Surfaces (IRS) to guide multipath, improve link reliability, and reduce interference while enabling equitable spectrum sharing. We pair graph neural networks (GNNs) for channel estimation and IRS configuration with physics‑regulated deep reinforcement learning (DRL) for resource allocation under strict QoS and fairness.
Currently recruiting undergraduate researchers