Generative AI-Enabled Integrated Sensing and Communication

Integrated sensing and communication (ISAC) is an effective technique to enable both conventional communications and new capabilities by sharing spectrum resources and hardware platform, thus facilitating diverse vertical applications. Nevertheless, ISAC usually requires complex system design and signal processing to achieve the integration gain. Recently, generative artificial intelligence (GAI) has showcased its remarkable capability of digital content generation and data processing, which is expected to enable the ISAC from diverse perspectives. In this article, we first overview the ISAC and GAI, and discuss the motivation of introducing GAI to ISAC. Subsequently, we investigate both the direct and indirect applications of GAI for ISAC with the focus on the system modeling, data processing and decision making. Subsequently, we propose a GAI-enhanced deep reinforcement learning algorithm for the beamforming design in the double reconfigurable intelligent surface (RIS)-aided ISAC. The superiority of the proposed algorithm over benchmarks is verified via simulation. Finally, open research challenges are discussed.