Soft Channel Estimation and Localization for Millimeter Wave Systems With Multiple Receivers

In millimeter wave (mmWave) communications, user position information can enable various position-based communication services, such as resource allocation, beam tracking and alignment, interference control, and synchronization. Classical localization methods focus on hard localization information, but soft localization provides the confidence levels in position estimates and thus enables the information to be efficiently fused with different measurements and application layers to realize integrated communication and localization. In this study, we propose a soft channel estimation and localization algorithm for an mmWave systems with multiple base stations. We present the Newtonized variational inference spectral estimation algorithm to extract soft information of position-related channel parameters. The soft localization algorithm estimates user position by using soft channel parameters from the expectation propagation simultaneous localization and mapping algorithm framework. The proposed algorithms realize channel estimation and localization in the communication process and refine the channel estimation through the localization information. Numerical results show that the proposed algorithms approach the Cramér-Rao lower bound for channel estimation and localization, and are thus verified to be effective.