Simplex Optimization Localization Algorithm for Wireless Sensor Networks

Accurate, distributed localization algorithms are needed for large scale dense wireless sensor network applications. The Nelder-Mead Simplex Optimization Method (SOM) is applied to solve nondifferentiable problems and can often handle high discontinuity, particularly if it does not occur near the solution. This article proposes a distributed iterative multilateral algorithm. It formulates a function that presents the sum of range errors of an unknown node to its all neighboring anchors and applies the Nelder-Mead simplex optimization method to solve the minimum value of the function to calculate the unknown node´s location. The algorithm is implemented through event-driven scheme, applies timer trigger to make localization round after round so as to reduce error accumulations. Experiment results reveal that the average localization error that caused by applying SOM is far lower than that caused by applying maximum likelihood estimation.