Authentication of Smartphone User Using RSSI Geolocation
Abstract
This thesis attempts to authenticate a smartphone user by pattern of life based on a smartphone user s geolocation throughout the course of a day. Current smartphone technology uses the global positioning system (GPS) as the primary source for geolocation because of its accuracy. However, services such as Google Location Service and Skyhook use Receive Signal Strength Indicator (RSSI)-based geolocation in GPS-degraded environments, such as inside a building. By using a smartphone s Wi-Fi application programming interface, a smartphone would detect all wireless access points Wi-Fi signals and associated signal strength over a discrete time interval. A hidden Markov model is used to model various smartphone users and used as an authentication method. The resulting f-score from the experiments ranged between 0.76 and 0.80, which is well above the 0.20 baseline. It is feasible to use RSSI-based geolocation as an element in combination with other methods to continuously authenticate a smartphone user. For an acceptable authentication method, the evaluation criteria must be as close to 1.0 as possible. Future research could combine authentication from RSSI-based geolocation with gait and keystroke analysis to improve results by leveraging other sensors on a smartphone.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2014
- Accession Number
- ADA608080
Entities
People
- Vincent K. Nguyen
Organizations
- Naval Postgraduate School