Mobile user location in dense urban environment using unified statistical model

A statistical approach for mobile subscriber (MS) location in urban environment is derived in this work. Availability of novel 4G mobile broadband technologies motivated a renewed interest in applications which require MS positioning. Implementation of these applications in urban environment depends on the ability to locate the MS in NLOS propagation conditions. A recently proposed statistical model of the propagation conditions in urban built-up environment, is adopted in this work. The statistical model is parameterized by the MS location, and is interpreted as a likelihood function. The proposed method does not involve any significant data collection during the training process, it requires only a single base station (BS), does not require identification or mitigation of the NLOS conditions, and is computationally efficient. Source localization performance of the proposed method was evaluated using a measured data, and acceptable localization accuracy was achieved.