An Efficient Low-Complexity Algorithm for MI-Based MIMO ISAC Beamforming Design

The paper investigates MIMO integrated sensing and communication (ISAC) systems with extended radar clutter interference. Different from previous work considering a point target scatterer, we focus on a more general scenario with an extended target scatterer. We aim to maximize the sensing mutual information (MI) while satisfying constraints on communication quality of service (QoS) and base station (BS) transmit power. The existing algorithm exhibits high computational complexity in the general scenario proposed in this paper. Hence, we propose an efficient low-complexity algorithm that employs a parallel optimization strategy and closed-form solutions to significantly reduce the complexity of each iteration. Moreover, we derive a lower bound on sensing MI, which is a function of transmit beampattern, and provide a sufficient condition for its tightness. Numerical results demonstrate the low complexity and good performance of our proposed algorithm compared to the existing benchmarks and validate the relationship between the lower bound and sensing MI.