A probabilistic approach to determine mobile station location with application in cellular networks

Abstract

A probabilistic approach to the mobile station location problem is analyzed, and an algorithm that implements such an approach is presented. Information about the mobile station location is collected in the form of two-dimensional probability density functions provided from various sources. Combining the probability density functions into a joined probability density function is addressed. To provide computational efficiency, a method to limit the space of a possible location of the mobile station to a rectangular region of the minimal size is presented. Discretization of space is performed next, reducing the probability density functions to matrices of probabilities. The algorithm for combining the probability density functions is adjusted for an application with the matrices of probabilities. To reduce the computational burden further, probability density functions of the exclusion type, taking only two values, zero and nonzero, are introduced, providing savings in storage space and computational time. Information about the timing advance parameter value and the received signal level are interpreted by probability density functions of the exclusion type. Application of the algorithm is illustrated by two sets of measurements performed in an urban and a suburban region.