All ETDs from UAB

Advisory Committee Chair

Thomas C Jannett

Document Type


Date of Award


Degree Name by School

Doctor of Philosophy (PhD) School of Engineering


This dissertation focuses on energy-based target localization in wireless sensor networks (WSNs). A WSN consists of a fusion center and a large number of sensors. The fusion center localizes a target using a maximum likelihood estimation (MLE) method based on decisions received from sensors whose positions are known. However, several factors may degrade localization performance, including imperfect communication channels, sensor faults, sensor position uncertainty, and sensor sleep. Localization performance may be improved if these factors are accounted for in the MLE framework. In this dissertation, models were developed that describe how localization performance is affected by imperfect communication channels, multi-hop communication channels, sensor position uncertainty, sensor faults, and sensor sleep. These models were incorporated into the MLE framework either simultaneously or separately. The Cramer-Rao lower bound on the estimation error (CRLB) was also derived. Simulation results showed that the new MLE methods gave better localization performance in terms of root mean square (RMS) estimation errors than current methods which ignore these factors that degrade localization performance. Moreover, the RMS estimation errors were close to the CRLB. Based on the sensor sleep model, an energy saving strategy was developed in which the sensor sleep probability is used to adjust energy consumption of the WSN. However, although the sensor sleep strategy saves energy, sleeping sensors cannot send decisions to the fusion center and missing decisions can degrade localization performance. Therefore, the system design must consider tradeoffs between energy consumption and localization performance. Models of energy consumption and localization performance were developed for one-dimensional and two-dimensional sensor arrays. These models were used in a multi-objective optimization method that balances energy consumption and localization performance. The Pareto-fronts generated by the multi-objective optimization method allow the system designer to make tradeoffs between energy consumption and localization performance.

Included in

Engineering Commons