Indoor positioning using multipath components in wireless networks.

Abstract

In wireless propagation, the transmitted signals are scattered, reflected and also diffracted by objects in the environment. In urban canyons or indoors, the signal that reaches the receiving antenna consists of multiple replicas of the transmitted signal that are called multipath components. The localization accuracy might be drastically reduced due to the distorted received signal thus localization algorithms need to resolve the impact of multipath components on the received signal in order to obtain accurate results when estimating the position of a user. In this project report, rather than mitigating the multipath components we exploit them in position estimation and this is called multipath assisted positioning (MAP). The algorithm to do this is called Channel-SLAM and the basic idea of this algorithm is to interpret the multipath components as signals that are originating from the so called virtual transmitters. There is inherent synchronization of these virtual transmitters to the physical transmitter and they are also static in their positions. Also in this report, we show that localization is possible even when only one physical transmitter is capable of sending signals to the receiver. The minimum signals for localization are three so the multipath components can make up this number since they also contain that can be useful for localization. To use the information contained in the multipath components, Channel-SLAM estimates the position of the virtual transmitters without the need for any prior information such as a room layout. Instead of mapping the physical environment, Channel-SLAM maps the virtual transmitter positions and interprets them as landmarks. In order to ease the complexity in computation, a particle filter known as the Rao-Blackwellization Particle Filter is applied to estimate the position of each virtual transmitter using a separated particle filter. In order to analyze the performance and location accuracy of Channel-SLAM, we applied it on a two dimensional environment having a user moving in a random track. We quantified the performance basing on the Root Mean Square Error and other parameters like signal to noise ratio.