Node Localization Algorithm Based on Matrix Pencil for Wireless Sensor Network

Wireless sensor network (WSN) is a new way to collect and process information. The node localization is the basis for most WSN applications. As matrix pencil (MP) algorithm is a high-resolution method of parameter estimation, in this paper, we put forward a novel range-based node localization method for WSN, which combines MP algorithm, multilateral localization with Taylor algorithm or Chan algorithm. MP algorithm is used to estimate the time of arrival (TOA) of the signal from an unknown node to an anchor node. Compared with traditional localization methods, the proposed method has some key advantages, such as higher positioning accuracy, smaller computational amount, suppression over non-Gaussian noise, energy saving, faster executing, etc. In the simulation, we also study the impact of signal to noise ratio (SNR) and the number of nodes on positioning accuracy.