Combined kernel distance and cross-correlation for object detection enhancement

Device-Free Localization (DFL) is an expanding field of research in the context of Wireless Sensor Networks (WSNs) and Internet-of-Things (IoT). Traditional positioning techniques concentrate on tracking localization tags carried by individuals. This may rise privacy concerns to individuals who are not willing to be identified. DFL schemes have advantages over traditional device-based schemes such as preserving an individual’s privacy and ability to localize individuals with no tracking tags. DFL systems that utilize Received-Signal-Strength-Indicator (RSSI) to acquire an individual’s location, operate on an individual’s effect on wireless signals. In this paper we implement RSSI-based DFL algorithm that combines Kernel Distance Radio Tomographic Imaging (KDRTI) with cross-correlation to improve KDRTI object detection metric. Experiments on static and moving individuals, have shown that cross-correlation has potential improvement for KDRTI method’s detection metric. Improvement is characterized by notable threshold created from correlation.