Wireless LAN based indoor positioning using radio-signal strength distribution modeling

Aiming to solve one of the serious problems of radio-signal strength (RSS) indoor positioning (namely, fluctuation of RSS values due to reflective waves), a novel indoor-positioning method based on RSS distribution modeling (called "RSS distribution modeling using the mirror-image method," RDMMI), was developed and evaluated. With RDMMI, a model of RSS distribution is created by using measured RSS and considering reflective waves. RDMMI achieves an average positioning error of 3.1 m, which exceeds that of conventional nearest-neighbor (NN) method by 0.6 m. It also accomplishes average positioning error of 3.7 m with a small set of randomly chosen RSS measurements. Furthermore, even in the case that radio source positions are unknown, RDMMI exceeds NN method.