Integrated Communication and Localization for Next-Generation Ultra-Massive MIMO System

The evolution towards ultra-massive multiple-input-multiple-output (UM-MIMO) system holds immense promise to accommodate a more ambitious vision of next-generation communications. Nevertheless, the emerging UM-MIMO not only implies a sharp increase in the number of antennas, but also necessitates a paradigm shift in signal processing methods since its Fresnel region expands rapidly. In this paper, we exploit its near-field nature to enable an integrated communication and localization (ICL) scheme. Specifically, we first propose an ICL signal frame structure including uplink training and data transmission stage. In the uplink training stage, user terminals (UTs) transmit reference signals for integrated channel estimation and localization, wherein the near-field projection matrix is leveraged to achieve sparse channel estimation and sense the scatterers' angular and range parameters simultaneously. Then, treat the sensed scatterers as virtual anchors, the location of UT is derived in line with the ray-tracing propagation model and a weighted least-squares strategy. Exploiting the measurable time difference of arrival of multipath components of the broadband system, we further propose to refine the localization results in an iterative manner. While in the downlink/uplink data transmission stage, the acquired channel state information and location information can contribute to more efficient payload data transmission. Simulation experiments demonstrate the effectiveness of the proposed ICL scheme, which opens up a new avenue for target localization in sharp contrast to the traditional localization methods.