We consider a cell-free massive multiple-input multiple-output (CF-mMIMO) surveillance system, in which multiple multi-antenna monitoring nodes (MNs) are deployed in either observing or jamming mode to disrupt the communication between a multi-antenna untrusted pair. We propose a simple and effective channel state information (CSI) acquisition scheme at the MNs. Specifically, our approach leverages pilot signals in both the uplink and downlink phases of the untrusted link, coupled with minimum mean-squared error (MMSE) estimation. This enables the MNs to accurately estimate the effective channels to both the untrusted transmitter (UT) and untrusted receiver (UR), thereby yielding robust monitoring performance. We analyze the spectral efficiency (SE) performance of the untrusted links and of the monitoring system, taking into account the proposed CSI acquisition and successive MMSE cancellation schemes. The monitoring success probability (MSP) is then derived. Simulation results show that the CF-mMIMO surveillance system, relying on the proposed CSI acquisition scheme, can achieve monitoring performance close to that achieved by having perfect CSI knowledge of the untrusted link (theoretical upper bound), especially when the number of MNs is large.

翻译：本文研究一种无蜂窝大规模多输入多输出监控系统，其中多个多天线监控节点以观测或干扰模式部署，旨在破坏一对多天线非可信用户间的通信。我们提出了一种在监控节点处实现简单高效的信道状态信息获取方案。具体而言，该方法利用非可信链路上下行阶段的导频信号，结合最小均方误差估计，使监控节点能够准确估计到非可信发射机和非可信接收机的等效信道，从而获得鲁棒的监控性能。我们分析了非可信链路及监控系统的频谱效率性能，其中考虑了所提出的信道状态信息获取方案与连续最小均方误差消除方案。在此基础上推导了监控成功概率。仿真结果表明：基于所提信道状态信息获取方案的无蜂窝大规模MIMO监控系统能够达到接近已知非可信链路完美信道状态信息（理论上界）的监控性能，尤其在监控节点数量较大时表现显著。