An Indoor Localization and Tracking System Using Successive Weighted RSS Projection

This letter proposes a novel successive weighted received signal strength (RSS) indoor localization and tracking system that projects previous time instance estimated mobile device (MD) position to provide projected RSS values. Such RSS projection increases the number of available RSS from Nm to Nm+ NAP, where NAP is the total number of access points and Nm is the number of RSS values measured by MD, ranging from 0 to NAP. Our proposed system thus resolves the issues associated with insufficient or no RSS values received by MD. Inertial navigation system (INS) is merged with RSS localization system to provide a weighted fusion of projected and measured RSS values. The weighting factors are derived based on the INS and RSS localization accuracy where the former is initially accurate but deteriorates with time and the latter is time-independent but environment-dependent. The proposed system was tested in indoor environments and outperformed other existing localization systems such as RSS and INS fusion using extended Kalman filter and non-line-of-sight (NLOS) selection scheme, especially in heavy multipath environment, by 42% and 75%, respectively.