On localization and tracking using received signal strength measurements /

Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Electrical and Electronics Engineering, Turkey

Approval Date: 2015




In this study, first, some received signal strength (RSS) based localization techniques, including maximum likelihood estimation (MLE), multidimensional scaling (MDS) and weighted least squares (WLS), are investigated and compared to each other via a simulation study within the perspective of a collaborative localization scenario. MLE using RSS measurement model, called RSS-MLE is known in the literature to be significantly biased. An important observation of this work is that the aforementioned bias can be clearly reduced in some collaborative localization scenarios when the non-connectivity information is incorporated into maximum likelihood (ML) cost function. We refer to the ML algorithm including the non-connectivity information as hybrid RSS-MLE (h-RSS-MLE). In order to support the reduced bias observation and determine the conditions in which h-RSS-MLE can mitigate the bias, we derive an analytical expression for the bias of the ML estimator based on a second order Taylor series expansion of MLE cost function by incorporating connectivity constraints into the problem. Since this analysis gives results which do not match the simulation results in a 2-D scenario, we also derive another expression based on a Taylor series expansion of the RSS measurements. The latter analysis is validated under some 2-D non-collaborative localization scenarios through a simulation study for MLE optimized by a grid-search. Finally, we make simulations as well as an experimental study to compare the localization algorithms with some conventional tracking methods including Kalman filters and a particle filter. It is observed in the experiments that the tracking methods can increase the accuracy about one meter compared to the localization algorithms for a non-collaborative case.