Cramer-Rao Bound Analysis of Localization Using Signal Strength Difference as Location Fingerprint

• Published in: 2010 Proceedings IEEE INFOCOM pp. 1 - 9
• Main Authors: Hossain, A.K.M. Mahtab, Wee-Seng Soh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-03-2010
• Subjects: Algorithm design and analysis; Bluetooth; Communications Society; Computer errors; Fingerprint recognition; Phase estimation; Signal analysis; Signal design; System analysis and design; Training data
• ISBN: 1424458366; 9781424458363
• ISSN: 0743-166X; 2641-9874
• DOI: 10.1109/INFCOM.2010.5462020

In this paper, we analyze the Cramer-Rao Lower Bound (CRLB) of localization using Signal Strength Difference (SSD) as location fingerprint. This analysis has a dual purpose. Firstly, the properties of the bound on localization error may help to design efficient localization algorithm. For example, utilizing one of the properties, we propose a way to define weights for a weighted K-Nearest Neighbor (K-NN) scheme which is shown to perform better than the K-NN algorithm. Secondly, it provides suggestions for a positioning system design by revealing error trends associated with the system deployment. In both cases, detailed analysis as well as experimental results are presented in order to support our claims.