28 October 2024 in IEEE Transactions on Green Communications and Networking
A near-field based on a semi-passive intelligent reflecting surface (IRS)-aided integrated sensing and communication (ISAC) framework is proposed. Our objective is to identify the optimal beam direction from the base station to both the communicating user and the sensing target through downlink beam scanning signals. Especially, in the communication module, the beam scanning process in the downlink by IRS’s passive reflecting elements (REs) searches for the optimal reflected beam and sends the signal from the base station to the communicating user. Meanwhile, in the sensing module, the IRS can estimate the location of nearby targets by analyzing the echo signals received by the active sensing elements (SEs) on the IRS. To achieve the above objectives, in this paper, we design the near-field codebook by considering the near-field cascaded array steering vectors of the IRS and match the optimal codewords of the IRS through an exhaustive training process, so as to efficiently complete the near-field beam training under the matched near-field channel model. We have done a performance analysis of this ISAC system for the Cramer-Rao bound (CRB), and CRB provides performance constraints for near-field target sensing. Specifically, the CRB curve shows a monotonic decrease with increasing beam scan duration, while the achievable rate initially increases and then decreases. Therefore, we need to set the size of the model codebook space reasonably to balance the perception and communication performance. Simulation results show that the perceptual performance of our proposed scheme in the ISAC system is close to the derived CRB.
26 June 2025 in IEEE Transactions on Vehicular Technology
In movable antennas (MAs) systems, implementing channel estimation remains a significant challenge due to pilot overhead and sparsity issues. To address these issues, we study a 5-D tensor-based Bayesian factorization for angles of departure (AOD) and angles of arrival (AOA) estimation in wireless communications with MAs in this paper. The MAs are utilized to reconstruct the channel response at different antenna positions, and we propose a algorithm that utilizes constructed received signals to reconstruct a fifth-order complex tensor model. To reduce computational complexity for angle estimation, the array structure properties are used to transform the tensor to a real-valued tensor model. Finally, we derive the Bayesian fifth-order tensor factorization from the constructed real-valued model and extract angular information using the approximated factor matrices obtained through the decomposition. We also derive the Cramér-Rao Bound (CRB) for angle estimation, where the angle parameters to be estimated rely on a finite set of channel measurements collected at random locations by the MA at both the transmitter and receiver. Our proposed method offers higher estimation accuracy compared to other angle estimation methods and is also effective when the number of transmission paths is unknown. The simulation results confirm the stability and effectiveness of the proposed angle estimation method based on Bayesian inference.