Cognitive Beamspace Algorithm for Integrated Sensing and Communications

This paper presents a novel cognitive beamspace algorithm for integrated sensing and communications (ISAC) systems, focusing on the optimization of spatial resources. The proposed method operates in the beamspace domain that enables an orthogonal design between the sensing and communication functions. Additionally, the approach leverages the principles of Thompson sampling, known for effectively balancing exploration and exploitation in uncertain environments. It enables the ISAC system to dynamically adjust radar target search strategies based on environmental feedback while maintaining acceptable communication rates with the user equipment. We demonstrate the algorithm's effectiveness over traditional methods through numerical simulations. These simulations reveal significant improvements in radar search functionality while meeting the communications quality of service constraints.