Range Free Localization in Wireless Sensor Networks for Homogeneous and Non-Homogeneous Environment

Abstract

Localization is one of the most important research subjects in terms of the wireless sensor networks (WSNs), because most of the information measured and distributed by the sensors are useful when sensors locations are well known. In this paper, a range free localization algorithm for sensor positioning is proposed. It is based on sensor distributed network model, coverage radius, hop count between each sensor and anchor, and improvement in the minimum mean square error, which computes coefficients for distance estimation between sensors and anchors. In the proposed method, the best coefficient for each hop count is computed with offline processing and Monte Carlo method. Then, these coefficients are stored in each sensor database and they are used for localization in the practical environment. Unlike some existent positioning methods, in this algorithm, it is not required to compute distance estimation coefficient by sensors. In the proposed method, it is assumed that all sensors use the omnidirectional antenna for their usual information transmission. High precision in geographical coordinate determining, less traffic load, and especially good performance in the homogeneous and non-homogeneous environment are the most important features of this algorithm. Simulation results show that the proposed algorithm has a good position determination and reduced traffic load for WSNs, as compared with some existent positioning schemes. Indeed, the proposed method improves localization precision and reduces traffic load simultaneously.