Performance of WLAN-based Object Detection with Distributed Antenna and Spatially Concatenated CSI

Wireless local area network (WLAN) based object detection techniques have attracted much attention. However, the detection performance of WLAN-based object detection is highly dependent on the antenna topology used. In this paper, we experimentally evaluate performance of antenna topology on the achieved object detection performance using channel state information (CSI) in WLAN systems, where spatially concatenated CSI is utilized as effective feature information for machine learning (ML)-based object detection. Based on the experiment results, we discuss how much object detection performance improvement is expected when the proper antenna topology is used at the access point side. We also experimentally clarify that the detection accuracy is improved when concatenated CSI is used as feature information for ML-based object detection for various antenna placement scenarios.