Improved probability matching model for mobile positioning based on GSM network

An improved probability positioning algorithm is proposed to enhance the accuracy of location estimation for outdoors under cellular network. The traditional probability algorithm models the received signal strength (RSS) by the standard Gaussian model from a base station. However, the propagation of the radio signal is based on a log-loss propagation model, which explains the relationship of the RSS and propagation distance. Hence, we proposed an improved asymmetric Gaussian model, according to the signal propagation model and the received signal strength. A genetic algorithm estimates the parameters of the proposed model. At the end of this paper, an experiment is conducted to estimate the locations of some given test samples and it is compared with some other approaches. The experiment result shows that our improved model could enhance the accuracy of location estimation.