Oscillator Drift Compensation by Line-of-Sight Tracking for Distributed Multisensor ISAC

Abstract: We observed synchronization mismatches in the form of non-smooth phase progressions and drifts within mobile multisensor channel sounding measurements. However, performing Doppler estimation in a distributed multisensor integrated sensing and communications (ISAC) system requires coherence among the nodes, which implies a continuously differentiable phase progression of the received signals. To correct the sounding data in post-processing, we extend traditional geometry-based drift compensation algorithms by utilizing Kalman filtering for line-of-sight (LoS) tracking, which improves the robustness of the LoS estimate in multipath scenarios. This approach smooths the phase progression and enables the correction of time-varying drifts while preserving relative sensor motion. Furthermore, we propose using the relative residual power after high-resolution parameter estimation (HRPE) as a metric for ground-truth-independent comparison of post-processing synchronization methods for recorded channel sounding data. Results show that the proposed approach outperforms traditional LoS estimation heuristics, reducing the relative residual power by more than 5 dB and the delay-Doppler estimate root mean square errors (RMSEs) by approximately 60 %.