Skip to Main Content
Sensor Network Indoor Localization Systems (SNILS) gain a significant attention these years, due to their ease of deployment and inexpensiveness. Ranging methods play basic role in the localization system, in which the RSSI (received signal strength indicator)-based ranging technique attracts the most attention. But the accuracy of the RSSI-based localization method remains a big challenge, because of the severe fading effects in the indoor environment. In this paper, a radio map method is proposed to improve the accuracy of the RSSI-based SNILS. This method contains two phases. The first phase is the radio map setting up phase. The radio maps are a set of probability density functions (pdf) indicating the radio fading pattern in the concerned environment, which are setup by a cooperative target. The second phase is the target tracking phase. Position probability matrixes (PPM) are used to indicate the positions of the targets, which are calculated by refering the stored radio maps according to the real-time RSSI values. To improve the localization accuracy, a radio map based Bayesian filter is proposed to iteratively calculate the PPM to speed up the convergence of the variance. Fully distributed algorithms of the localization method and the filter are designed and are implemented in the MICA2 system. The experimental accuracy is shown to be less than 1 meter with 80% probability, much better than current RSSI-based SNILS.