Title:MovISAC: Coherent Imaging of Moving Targets with Distributed Asynchronous ISAC Devices

Abstract:Distributed integrated sensing and communication (ISAC) devices can overcome the traditional resolution limitations imposed by the signal bandwidth, cooperating to produce high-resolution images of the environment. However, existing phase-coherent imaging approaches are not suited to imaging multiple moving targets, since the Doppler effect causes a phase rotation that degrades the image focus and biases the targets' locations. In this paper, we propose MovISAC, the first coherent imaging method for moving targets using distributed asynchronous ISAC devices. Our approach obtains a set of high-resolution images of the environment, in which each image represents only targets moving with a selected velocity. To achieve this, MovISAC performs over-the-air (OTA) synchronization to compensate for timing, frequency, and phase offsets among distributed ISAC devices. Then, to solve the prohibitive complexity of an exhaustive search over the targets' velocities, we introduce a new association algorithm to match the Doppler shifts observed by each ISAC device pair to the corresponding spatial peaks obtained by standard imaging methods. This gives MovISAC the unique capability of pre-compensating the Doppler shift for each target before forming the images, thus recasting it from an undesired impairment to image formation to an additional means for resolving the targets. We perform extensive numerical simulations showing that our approach achieves extremely high resolution, superior robustness to clutter thanks to the additional target separation in the Doppler domain, and obtains cm- and cm/s-level target localization and velocity estimation errors. MovISAC significantly outperforms existing imaging methods that do not compensate for the Doppler shift, demonstrating up to 18 times lower localization error and enabling velocity estimation.