Positioning in wireless networks is today mainly used for yellow page services. Yet, its importance will grow when emergency call services become mandatory as well as with the advent of more advanced location-based services. It is also plausible that future resource management algorithms may rely on position estimation and prediction. The problem of NLOS is causing positioning to become unreliable. This paper presents positioning algorithms for cellular network-based vehicle tracking in severe non-line-of-sight (NLOS) propagation scenarios. The aim of the algorithms is to enhance positional accuracy of network-based positioning systems when the GPS receiver does not perform well due to the complex propagation environment. One-step position estimation method and another two-step method are developed. Constrained optimization is utilized to minimize the cost function which takes account of the NLOS error so that the NLOS effect is significantly reduced. In this we measure the vehicle velocity and heading direction by using an algorithm.