Received signal strength localization with an unknown path loss exponent

• Published in: 2011 24th Canadian Conference on Electrical and Computer Engineering(CCECE) pp. 000456 - 000459
• Main Authors: Chan, Y. T., Lee, B. H., Inkol, R., Chan, F.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.05.2011
• Subjects: Cramer-Rao bounds; Fading; Geometry; Localization; Maximum likelihood estimation; received signal strength; Receivers; Simulation; Wireless sensor networks
• ISBN: 9781424497881; 1424497884
• ISSN: 0840-7789
• DOI: 10.1109/CCECE.2011.6030492

Received Signal Strength (RSS) measurements obtained at locations in the vicinity of an emitter can be used to estimate the emitter location given a suitable path loss model. For commonly used propagation models, the RSS has a dependence on the emitter to sensor separation, d, of the form P ∝ d -α , where the path loss exponent, α, is typically between 2 and 4 depending on the terrain. This uncertainty is a problem since a solution for the emitter location obtained using a suboptimal choice of α will suffer from degraded accuracy. To resolve this issue, a near maximum likelihood (ML) estimator for both α and emitter position (x,y) has been developed. By expressing (x,y) in terms of α and substituting the results into the ML-function, the 3-D minimization problem is simplified to 1-D. A simple search in α then gives the (α,x,y) estimates. Simulation experiments corroborate the proposed approach and demonstrate estimation accuracy close to the Cramer-Rao Lower Bound.