Sensing-Assisted Channel Estimation for OFDM ISAC Systems

The advent of integrated sensing and communication (ISAC) has brought new perspectives to the age-old topic of channel estimation. Leveraging sensing information as prior knowledge holds the potential to assist channel estimation, thereby enhancing communication performance. However, how to leverage the sensing information for channel estimation is still an open problem. We demonstrate that even small sensing errors can lead to significant channel estimation inaccuracies in wideband systems for existing channel estimation methods that directly utilize sensing information. To this end, a sensing-assisted channel estimation scheme is developed for the bistatic orthogonal-frequency-division-multiplexing (OFDM) system in this paper. An architecture for sensing-assisted channel estimation is first proposed. Next, a sensing-assisted linear minimum mean square error (LMMSE) estimation algorithm is developed, which is capable of effectively integrating inaccurate sensing information. The proposed algorithm incorporates delay and Doppler tolerance factors in the construction of channel correlation matrices for LMMSE, so as to combat the impact of errors in delay and Doppler estimation. Simulation results show that the proposed algorithm can significantly outperform the existing methods.