Location determination is an essential component in large sensor networks. In those types of networks, sensors are data centric, and hence, the location of each sensor is critical. Furthermore, the need for a cost-effective means for sensor localization eliminates a variety of options including GPS. In this paper, we use particle filters for estimating the location of sensors assuming a noisy Rayleigh fading channel. Simulation results show that the algorithm converges to approximately two-meter error range.