Wi-Fi Based Microwave GI Surveillance for Indoor Scenarios

Abstract

This thesis focuses on a novel design of Wi-Fi based microwave ghost imaging scheme for surveillance applications. The first research focus is demonstrating the feasibility of applying Wi-Fi signals into microwave ghost imaging. To meet the time-space independent requirements of microwave ghost imaging, the applied Wi-Fi signal source should be random and group-orthogonal. However, the preamble sequence of a Wi-Fi signal is full of repeating elements and the preamble sequences from different access points are the same. Based on a further analysis of Wi-Fi frame structures, I propose a detection time window to avoid using any preamble sequence for achieving a better performance of microwave ghost imaging. In addition, the performance of the Wi-Fi based microwave ghost imaging scheme for stationary objects has been numerically simulated. The second research focus is modifying the imaging reconstruction procedure in microwave ghost imaging scheme for surveillance in practical indoor scenarios where both stationary and moving targets appear. If the moving object appears, the reconstructed imaging will be blurred. To eliminate the blur caused by motion, the method of refocusing to compensate motion of the object has been proposed. The effectiveness of refocusing is validated in relative numerical simulations. Further, I investigate the proposed scheme in the scenario that contains both stationary and moving targets. The receiving signal should be processed to distinguish various reflecting signals from different targets before refocusing. Thus, the object reconstruction procedure of microwave GI is also modified to fit this scenario for surveillance purpose. Numerical simulation shows that the proposed modification can effectively isolate reflections from targets with different velocities and obtain focused images.