This paper is motivated by our interest in a localization technique for a mobile sensor network, deployed in a harsh environment and a set of interesting/surprising results obtained from simulations of two state-of-the-art localization techniques for mobile sensor networks, namely MCL and MSL. Range-based localization methods require an estimate of the distance or angle between two nodes to localize and may operate in both absolute and relative coordinate systems.

Methods requiring specialized hardware include precise measurement of acoustic phase difference, optical sensors/reflectors. We have proposed FUZLOC, a fuzzy logic-based localization method suitable for wireless sensor nodes that are mobile in noisy, harsh environments. The constituent systems use fuzzy multilateration and a grid predictor to compute the location of a node as an area. The multilateration technique presented in the previous section assumes the presence of a sufficient number of anchors, typically three or more., in mobile sensor networks with low anchor densities, it might frequently be the case that a node does not have enough anchors for multilateration. To address this problem, we extend our fuzzy logic-based localization framework to predict an area, e.g., a cell in a grid, where the node might be. The idea is inspired from cellular systems. virtualize the anchors, so that a node is within a set of Virtual Anchors at any point in time. A Virtual Anchor is a fictitious anchor which is assumed to be located at a known, fixed location in the field of deployment, the distance to which can be found in an approximate way from the node.