Performance Evaluation of Object Detection Utilizing Channel State Information in Wireless LAN Systems with Distributed Antennas

In this paper, we investigate a machine learning based object detection scheme utilizing channel state information (CSI) in wireless local area network (WLAN) systems with multiple distributed antennas, and evaluate the impact of antenna placement on achieved detection probability, where CSI frames captured from nearby wireless devices are used for machine learning based object detection. To improve the objective detection accuracy, we also investigate a method to distribute transmit and receive antennas in a target area for acquiring sufficient amount of CSI, where support vector machine (SVM) is used as a supervised machine learning algorithm. For performance evaluation, we used compressed CSI specified in IEEE 802.11ac standard and conduct a ray-tracing simulation in an indoor propagation environment for analyzing data and detecting a target object. Simulation results show that object detection probability can be improved by properly distributing antenna elements in a target area.