MP-based Algorithm in Node Localization of Wireless Sensor Network

Abstract

W ireless sensor network (WSN) has attracted more and more research interest b ecause of its various applications in distributed monitoring, communication, and computing . For most applications, developing a distributed practical positioning algorithm is probably the most important and challenging task. As a high-resolution method of parameter estimation , m atrix pencil (MP) algorithm can be used in node localization. W e put forward a novel range-based node localization method for WSN i n th is paper . The new scheme employ s MP algorithm for multi-path time-delay estimation , multilateral localization for absolute positioning and Taylor algorithm or Chan algorithm to solve non-linear equations . T he proposed method has some key superiority over traditional loca lization methods, such as higher positioning accuracy, smaller computational amount, suppression over non-Gaussian noise, energy saving, faster executing, etc. Simulations show that the accuracy of MP algorithm for time of arrival (TOA) estimation is low (ns-level) and non-Gaussian noise can be well inhibited. T he impact of some factors such as signal to noise ratio (SNR) and the number of nodes are also studied in this paper.