RSS-Based Self-Localization Framework for Future Wireless Networks

The localization is an important asset in all existing and emerging wireless networking solutions since it can extend the radio environmental awareness and assist in providing better network operation. The received signal strength (RSS) based non-Bayesian transmitter localization is especially interesting due to the inherent presence of the RSS observations in all commercial radio devices and the instantaneous estimations without the need for extensive training/learning phases. The existing RSS based localization solutions neglect the problems arising from the inherent sensing network topology uncertainty. The sensors, usually assumed to have a-priori known positions, are often a subject of a previous estimation which propagates errors in the transmitter localization procedure, and, hence, results in significant transmitter localization performance degradation. This paper presents a recently developed generic RSS based joint transmitter/sensors localization framework, founded on the assumption of uncertain topology information. The derived joint maximum likelihood (JML) algorithm simultaneously estimates the transmitter and uncertain sensor positions providing twofold gains: improving the transmitter localization and reducing the network topology uncertainty. The paper broadly evaluates the JML algorithm, emphasizing the substantial localization gains originating from the joint transmitter/sensor position estimation. The results prove up to 85 % sensor position uncertainty reduction with the general system model with multiple transmitters locations and multiple previous estimations of the sensors positions. The paper also derives the theoretical lower bounds of the joint estimation framework, and proves the convergence of the JML algorithm. The presented joint estimation framework is applicable to a variety of wireless networking applications. It can provide self-awareness in future wireless networks and cope with the environment and topology dynamism in wireless ad-hoc networks.