Device-Free Passive Localization

This report describes a Device-Free Passive Localization System (DfP). The system provides a software solution over nominal WiFi equipment to detect the presence and track intruders in an area of interest. The system is based on measuring the changes in the received RSSI at a fixed receiver form a fixed transmitter. Using RSSI values for multiple transmitters and monitoring points, we show a system is able to detect intruders with a high probability and very low false positive rate. Moreover, the system is able to track the intruder with an accuracy of a few feet. 1 Background In the recent years, 802.11[1] wireless LANs (WLAN) installations have become increasing common, providing communication capabilities in the office, at home, and public places. In an office environment WLANs usually employ infrastructure mode in which a number of Access Points (APs) are carefully positioned to provide access from any location in the office. While the wireless communication is the primary reason for deploying WLANs, the deployed infrastructure can also be used for other purposes. For example, the RF signals have been used for determining the location of a receiver [1-7, 9]. In one such technology developed at the University of Maryland, Horus, the position of a wireless card can be tracked using the received signal strength in a pure software solution, to an accuracy of a few feet. Based on the user location, many context-aware applications can be implemented in a WLAN environment such as location-sensitive information retrieval and direction finding inside a building. All extensions on the use of WLANs require an active participation of a device with a NIC (Network Interface Card) in it. In contrast, a DfP system operates without requiring any active participation of the person being monitored or tracked. The DfP technology is based on both the knowledge that RF signals are affected by the presence of people in the environment and preliminary measurements indicating the RF signal changes are significant. The extent of the impact depends on the location of the person relative to the NIC. The DfP system infers the presence or movement of people from changes in received signal strength. Using the signal strength: 1. By monitoring the RSSI at one or more locations in an area/building in which WiFi is deployed, we can reliably detect the presence of people. 2. As we can detect the presence of people, we can quantify such detection. For example, we can determine the number of people and their location. We need to determine the accuracies that can be achieved in this process. The DfP system primarily uses the standard access point and wireless card components of any Wi-Fi installation. 2 System Architecture The basic architecture of a DfP system is shown in Figure 1. It consists of the Access Points (APs) and Monitoring Points (MPs) along with a DfP server. The APs of any WLAN deployment double as DfP APs. MPs monitor the RSSI of AP beacons and report thee values to the DfP Server. As we expect the location of MPs to be fixed, any stationary desktop computers which are normally used by the users can be used as MPs. The DfP server is a PC which performs computations on the RSSI streams and initiates actions as necessary. The DfP System is envisioned to support three modes of operation: Monitoring Mode In this mode, DfP System expects no movement of people anywhere in its protected area and raises alarms on detecting any movements. This mode may be appropriate for providing night time protection. In this mode MPs recognize the change of RSSI values and inform the DfP Server accordingly. The NS then takes the appropriate security steps reflecting the specific configurations.