Date of Award


Publication Type

Doctoral Thesis

Degree Name



Electrical and Computer Engineering


Fuzzy logic, indoor localization, IR-UWB, Ranging error


Kemal Tepe




The cutting-edge technology to support high ranging accuracy within the indoor environment is Impulse Radio Ultra Wide Band (IR-UWB) standard. Besides accuracy, IR-UWB’s low-complex architecture and low power consumption align well with mobile devices. A prime challenge in indoor IR-UWB based localization is to achieve a position accuracy under non-line-of-sight (NLOS) and multipath propagation (MPP) conditions. Another challenge is to achieve acceptable accuracy in the conditions mentioned above without any significant increase in latency and computational burden. This dissertation proposes a solution for addressing the accuracy and reliability problem of indoor localization system satisfying acceptable delay or computational complexity overhead. The proposed methodology is based on rules for identification of line-of-sight (LOS) and NLOS and the range error bias estimation and correction due to NLOS and MPP conditions. The proposed methodology provides accuracy for two major application domains, namely, wireless sensor networks (WSNs) and indoor tracking and navigation (ITN). This dissertation offers two different solutions for the localization problem. The first solution is a rules-based classification of LOS / NLOS and geometric-based range correction for WSN. In the first solution, the Boolean logic based classification is designed for identification of LOS/NLOS. The logic is based on channel impulse response (CIR) parameters. The second solution is based on fuzzy logic. The fuzzy based solution is appealing well for the stringent precision requirements in ITN. In this solution, the parametric Boolean logic from the first solution is converted and expanded into rules. These rules are implemented into a fuzzy logic based mechanism for designing a fuzzy inference system. The system estimates the ranging errors and correcting unmitigated ranges. The expanded rules and designed methodology are based on theoretical analysis and empirical observations of the parameters. The rules accommodate the parameters uncertainties for estimating the ranging error through the relationship between the input parameters uncertainties and ranging error using fuzzy inference mechanism. The proposed solutions are evaluated using real-world measurements in different indoor environments. The performance of the proposed solutions is also evaluated in terms of true classification rate, residual ranging errors’ cumulative distributions and probability density distributions, as well as outage probabilities. Evaluation results show that the true classification rate is more than 95%. Moreover, using the proposed fuzzy logic based solution, the residual errors convergence of 90% is attained for error threshold of 10 cm, and the reliability of the localization system is also more than 90% for error threshold of 15 cm.