Localization Projects at the iMPACT Lab
Our research includes several projects targeting the difficult problem of wireless sensor localization.
A Location-Based Access Control Architecture for Wireless Home Networks
Location-based access control maintains user privacy, as a person's location, not her identity, is used to provide access. It simplifies the management of authentication information by relying on a limited set of authorized locations, rather than thousands of authorized IDs. It also links physical access control to resource control. Authentication is already required for physical access. Location-based authentication removes this duplication.
Our project's goals are to develop an authentication mechanism providing location-based access to computing resources. We focus on determining location in a typical residential or office environment.
Challenges for Received Signal Strength (RSS) Based Approaches
Indoor radio signals suffer from multipath propagation, shadowing and scattering effects. Small changes in position, antenna orientation, and pedestrian traffic cause an extreme local variation in signal strength. The radio environment is highly location-dependent and dynamic over time. Solutions using scene analysis by constructing a localized radio model have high infrastructure and deployment costs.
Location-Based Access Control using Proximity Localization
Our use of proximity detection to perform localization is motivated by the inherent limitations of a system based on radio signal strength measurement, as revealed through experimentation. Radio proximity detection is based on the receiver's ability to successfully demodulate and decode a radio signal. Extremely low-power transmission, used in close range proximity detection, limits the effects of scattering, shadowing, and multipath propagation, has low infrastructure and deployment costs, and does not depend on modeling the radio environment. We found that proximity detection has several advantages over signal-strength based localization, including a higher degree of reliability mere connectivity being easier to gauge than variable signal strength and range adjustability, from a few meters to as little as twenty centimeters.
The architecture consists of wireless base stations, one positioned at each resource, and a badge transmitting low-power beacon packets. Briefly, when a base station receives beacons, it requests the current authentication information for that badge from an authentication server. Next, the base station conducts a challenge-response with the badge at low power, verifying the badge’s response using the information received from the authentication server.
Meanwhile, the other base stations engage in low power jamming on the same frequency, preventing eavesdropping of the challenge-response by creating a radio-hostile environment and helping to guarantee true proximity of the badge. Distributed listening stations monitor the same frequency, ensuring that no distant high-power transmitter is trying to spoof the system by pretending to be in close proximity. If a listening station can hear the challenge-response, location spoofing is indeed occurring, and the listener sends an authentication fail message to the authentication server. Otherwise, the base station sends an authentication success message to the authentication server, which contacts the resource directly out-of-band, using a wired network or a different frequency, instructing it to allow access.
Funded by: CEINT - Consortium for Embedded and Internetworking Technologies
Principle Investigators: Dr. Sandeep Gupta, Dr. Partha Dasgupta, Associate Professors, Arizona State University
Graduate Students: Valliappan Annamalai, John Quintero