Self-localization techniques for Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have a wide range of invaluable civil and military applications. Sensors localization is one of the fundamental steps in order to properly process the data for the target applications. The localization problem is particularly important in many applications, such as forest fire monitoring and wild life tracking. Traditional localization techniques have many limitations and are not adequate for some applications. For instance, global positioning system (GPS) receivers are limited to outdoor applications and are expensive if used on each sensor. Alternatively, other sensor localization techniques aim at finding the locations of all sensors in the network by measuring the pair-wise distance among sensors. In this paper, we consider the problem of cooperative localization in which some small number of sensors, called seeds act as reference nodes, with known location, and the rest, un-localized nodes, are cooperating to determine their own coordinates. Based on the local coordinates formed locally inside the WSNs, they can be tuned to global coordinates if at least two seeds in the WSNs have known coordinates or are equipped with GPS receivers. In this paper, the three main techniques used for localization in WSNs are analyzed in details. These techniques are based on the geometrical, statistical and optimization principles. In the geometrical principle, "triangulation" based on intersection of hyperbolas or circles are used to form triangles to locate the node. The statistical principle relies on an estimation model of range errors. The bound of the estimator covariance is a function of the number of nodes, sensor geometry, and other parameters that must be estimated. The optimization principle is based on one of two nonlinear optimization approaches. The first approach is based on adaptive sequence of Semi-Definite Programming (SDP). The second approach is based on gradient algorithm. This article introduces thorough analysis and discussions of the above techniques. Comparison between the different techniques, including cost, power consumption, and geometry, is presented. The deployment method used with each technique, number of nodes utilized, static/real time positioning and 2D vs. 3D localization complexity is also demonstrated.