Distributed Recognition of Reference Nodes for Wireless Sensor Network Localization

Abstract

All known localization techniques for wireless sensor and ad-hoc networks require certain set of reference nodes being used for position estimation. The anchor-free techniques in contrast to anchor-based do not require reference nodes called anchors to be placed in the network area before localization operation itself, but they can establish own reference coordinate system to be used for the relative position estimation. We observed that contemporary anchor-free localization algorithms achieve a low localization error, but dissipate significant energy reserves during the recognition of reference nodes used for the position estimation. Therefore, we have proposed the optimized anchor-free localization algorithm referred to as BRL (Boundary Recognition aided Localization), which achieves a low localization error and mainly reduces the communication cost of the reference nodes recognition phase. The proposed BRL algorithm was investigated throughout the extensive simulations on the database of networks with the different number of nodes and densities and was compared in terms of communication cost and localization error with the known related algorithms such as AFL and CRP. Through the extensive simulations we have observed network conditions where novel BRL algorithm excels in comparison with the state of art.

Keywords

• Recognition
• reference nodes
• wireless sensor networks
• localization
• communication cost
• simulation