An improved DV-Hop scheme based on similar path searching and Quasi-Newton iterative algorithm

The node localization process of the traditional DV-Hop (Distance Vector-Hop) scheme is sensitive to the heterogeneity of network topology, and the robustness of the scheme is poor while the network topology is diversiform due to random node deployment. Aiming at such a problem, improved strategies are provided. In the distance estimation step, a strategy of searching the most similar anchor-anchor path to anchor-node path is proposed; in the position calculation step, a modified Quasi-Newton iterative algorithm is added after the initial position of an unknown node is derived using Lateration algorithm. Simulation shows that compared with the original method and an existed representative improved scheme, our improvement can obtain higher distance estimation accuracy at first, and then, it can improve the robustness of the localization algorithm to range error. Thereby, the improvement can effectively improve the average localization accuracy of the unknown nodes in the network.