FO-tensor ISAC

Integrated sensing and communication (ISAC), as a rapidly advancing technique, is proposed to address both the communication and sensing problem in a unified framework over the millimeter wave (mmWave) frequency bands. In this paper, a cost-saving bi-static architecture is designed for the mmWave ISAC system, and a corresponding effective estimation scheme is proposed utilizing the canonical polyadic (CP) decomposition, where a heterogeneous placement of stations is adopted, and both communications terminals (CTs) and passive targets (PTs) are jointly considered. Based on the proposed bi-static architecture, the estimation of cascaded and uplink channels are modeled as high-order tensor recovery problems, where the channels are expressed in full-order CP-form tensors, and channel parameters are obtained by estimating the factor matrices. Specifically, the proposed downlink estimation scheme consists of three phases, where a novel reconstruction scheme is provided to reduce the downlink pilot overhead. Additionally, the Cramer-Rao lower bound (CRLB) and the generic uniqueness condition of the channel parameters estimation are derived. Numerical results verify that both the proposed estimation and reconstruction schemes outperform the classic monostatic counterpart in terms of accuracy, and the reconstruction scheme approaches the estimation scheme with a reduction in downlink pilot overhead.