Indoor localization of wireless emitter using direct position determination and particle swarm optimization

Many methods are introduced to accomplish determining the position of emitters with respect to knownposition receivers in indoor localizations. Among them, the direct position determination (DPD) approach advocates using the received signals by all the base stations together in order to estimate the locations in a single step. However, DPD is not very accurate due to the use of a gridding area, the effect of noise, and the multipath phenomenon. In order to improve the DPD performance, we derive an analytic model based on weighted least square estimation that uses simultaneously the effect of delay, Doppler, attenuation, and angle of reception of the signals. In addition, a new approach to define a cost function based on the analytic model is proposed that is optimized by particle swarm optimization (PSO). A combination of the improved DPD and the proposed PSO-based technique is also used to decrease the computation volume and increase the resolution. Finally, the accuracy of the proposed algorithms is investigated by Monte Carlo computer simulation in a wireless local area network. Numerical results show that the localization by PSO, the improved DPD, and previous DPD are more accurate in that order.