Ambient backscatter devices (tags and readers) use existing radio frequency (RF) signals to transmit data. Most prior works consider single-antenna tags, but this paper investigates the case of multiple-antenna tags, which are capable of simultaneous energy harvesting and data transmission. However, the multi-antenna channel between the tag and the reader, and the unpredictable nature of RF signals due to uncontrollable RF sources (e.g., location and transmit power), make signal detection highly challenging. Thus, the detection process becomes a hypothesis testing problem with unknown parameters. Consequently, we design a blind detector based on the generalized likelihood ratio test (GLRT) without using channel state information (CSI), signal power and noise variance. The decision threshold and detection probability of it are also analyzed in detail. Furthermore, to maximize its detection performance, we develop the optimal backscatter antenna selection scheme. Interestingly, we show that the detector performs best when only two backscatter antennas are selected. Finally, extensive simulation results validate the analysis and illustrate the effectiveness of the proposed detector.

中文翻译：

---

多天线标签环境后向散射通信的信号检测和最佳天线选择

环境反向散射设备（标签和阅读器）使用现有的射频 (RF) 信号来传输数据。大多数先前的工作都考虑单天线标签，但本文研究了多天线标签的情况，它能够同时进行能量收集和数据传输。然而，标签和阅读器之间的多天线通道，以及由于不可控制的 RF 源（例如，位置和发射功率）导致的 RF 信号的不可预测性，使得信号检测极具挑战性。因此，检测过程变成了一个未知参数的假设检验问题。因此，我们在不使用信道状态信息 (CSI)、信号功率和噪声方差的情况下，基于广义似然比测试 (GLRT) 设计了一个盲检测器。对其决策阈值和检测概率也进行了详细分析。此外，为了最大限度地提高其检测性能，我们开发了最佳反向散射天线选择方案。有趣的是，我们表明当只选择两个反向散射天线时，检测器的性能最好。最后，大量的仿真结果验证了分析并说明了所提出的检测器的有效性。