|
In this paper, we investigate cell-free integrated sensing and communication (ISAC) systems and propose a sensing-assisted beam tracking scheme based on multiple access point (AP) cooperation. Leveraging the Bayesian filtering frame work, the proposed scheme integrates prior state transitions with posterior filtering to achieve optimal tracking. Since the echoes received at each AP may contain contributions from multiple user terminals (UTs), we apply point-wise division and the two dimensional Fourier transform to map these echoes into the delay-Doppler domain, thereby enabling distinguishing signals from different UTs.
To further reduce computational complexity, we discard low-power components and focus on dominant parts by exploiting the sparsity of UT signals in the delay-Doppler domain. Moreover, the nonlinearity of state transition and observation models renders the direct computation of prior and posterior probability density functions intractable. To address this challenge, we employ particle filtering to provide a Monte Carlo approximation of the Bayesian filtering. Simulation results demonstrate the effectiveness of the proposed scheme.
Full Article: IEEE Transactions on Vehicular Technology, Early Access
| 
