Cooperative blind localization of primary user in a cognitive radio environment

Blind localization has emerged as an important topic addressing the challenges faced by any cognitive radio (CR) network. The estimation of the primary user (PU) position in a CR network is made difficult because generally cooperation between the PU and the secondary user (SU) does not exist. This presents a challenge since PU parameters such as transmit power, path loss exponent, noise variance and the location of the PU are not known. Hence, the need for a robust, accurate and range-free localization technique that does not require any knowledge of such parameters. In this paper, we investigate the localization performance when combining the database technique and the centroid localization techniques. We adopt a two-stage localization strategy. In the first stage, a mean-square error (MSE) method is used to find a close match between the actual received power and the database power and subsequently the PU estimates. Centroid localization (CL) and weighted centroid localization (WCL) techniques are then adopted in the second-stage as fusion strategies to give an estimate of the PU. We also analyse the root-mean squared error (RMSE) of the CL and WCL techniques. We further compare our results with the direct centroid localization techniques without involving the database.