Robust Techniques for Accurate Indoor Localization in Hazardous Environments

The challenging conditions prevalent in indoor environments have rendered many traditional positioning methods inept to yield satisfactory results. Our work tackles the challenging problem of accurate indoor positioning in hazardous multipath environments through three versatile super resolution techniques: time domain Multiple Signal Classification (TD-MUSIC), frequency domain MUSIC (FD-MUSIC) algorithms, and frequency domain Eigen value (FD-EV) method. The advantage of using these super resolution techniques is twofold. First for Line-of-Sight (LoS) conditions this provides the most accurate means of determining the time delay estimate from transmitter to receiver for any wireless sensor network. The high noise immunity and resolvability of these methods makes them ideal for cost-effective wireless sensor networks operating in indoor channels. Second for non-LoS conditions the resultant pseudo-spectrum generated by these methods provides the means to construct the ideal location based fingerprint. We provide an in depth analysis of limitation as well as advantages inherent in all of these methods through a detailed behavioral analysis under constrained environments. Hence, the bandwidth versatility, higher resolution capability and higher noise immunity of the TD-MUSIC algorithm and the FD-EV method’s ability to resurface submerged signal peaks when the signal subspace dimensions are underestimated are all presented in detail.