Date of Award
Doctor of Philosophy (PhD)
Rastko R. Selmic
We present new results in detection and localization of a hidden emitter using Wireless Sensor Networks (WSNs). While this type of problem has been explored extensively, this research combines multiple dynamics, such as mobility and adaptation, to extend the value beyond the typical search mission. Using Micro-Aerial Vehicles (MAVs), in conjunction with multiple static wireless sensor nodes, we create a hybrid sensor network capable of detection and localization of a hidden electro-magnetic (EM) emitter.
In order to localize the emitter, a Received Signal Strength Indicator (RSSI) is used as an approximation of distance from the transmitter to the revolving receivers. As a result, an algorithm for on-line estimation of the Path Loss Exponent (PLE) is used in modeling the distance based on Received Signal Strength (RSS) measurements. Based on surrounding sensors' RSS values, the emitter position estimation is then calculated using Least-Square Estimation (LSE) method.
RSS-based localization techniques can be inaccurate due to the noisy and uncertain nature of RSS in different mediums and environments. In order to improve localization accuracy, a technique called Position-Adaptive Direction Finding (PADF) is developed in which a team of MAVs coordinate their sensing missions, adapt their position in real-time, and localize the unknown emitter. We enhance the adaptation segment of the PADF by providing an algorithmic framework for MAVs to reposition themselves, thus avoiding obstructions or locations that may distort the propagation of the emitter and reduce the accuracy of the receivers' combined emitter location estimation. Given the cross-PLEs between the static and mobile nodes, we propose a cost function for MAVs' position adjustments that is based on the combination of such cross-PLEs and RSSIs. The mobile node adjusts current position by minimizing a quadratic cost function such that the PLE of surrounding receivers is decreased, while increasing RSSI from the mobile node to the target, thereby, reducing the inconsistency of the environment created by echo and multi-path disturbances. In the process, the mobile node moves towards a more uniform measuring environment that ultimately increases localization accuracy.
This Dissertation presents our recent results on a novel, multi-platform, RF emitter localization PADF technique. The position-adaptive approach shows potential for an accurate emitter localization in challenging, embedded, multi-path environments such as urban environments. We also present a recent development of a three-state machine-based MAV cooperative control algorithm that is used in search for a hidden emitter.
We provide simulation and experimental results that illustrate proposed methods. We describe the testbed and laboratory development for experimentation and discuss obtained results. Future work is proposed that includes complex cepstrum between mobile nodes and the hidden emitter as a metric for MAV control. Complex cepstrum is correlated with a received echo in EM signals and reducing the cepstrum is expected to improve measurements and estimation accuracy by moving the MAV sensor nodes towards low echo space.
Gates, Miguel D., "" (2013). Dissertation. 291.