Locating Mobile Stations with Non-line-of-sight Measurements

We consider the problem of locating mobile stations using range measurements from several base stations when the direct path from transmitter to receiver may be intermittently or, quite possibly, continuously blocked. This non-line-of-sight phenomenon is known to be a major source of error in position location because it systematically causes the mobile to appear farther away from the base station than it actually is. In this paper, we present a first order model for the non-line-of-sight error and show that, under certain conditions, it is possible to (1) detect the non-line-of-sight error using a hypothesis testing procedure and to (2) reconstruct the line-of-sight by exploiting a priori knowledge of the statistics of the standard measurement noise. Simulation examples are presented in order to demonstrate performance of this algorithm.