A high accuracy localization algorithm with DV-Hop and fruit fly optimization in anisotropic wireless networks
Wenxian Jia, Qi Guohong, Menghan Liu, Jie Zhou
Abstract
Wireless sensor networks (WSNs) technology has been applied in many fields. At the same time, the realization of these applications needs to be combined with the location information of nodes. Therefore, the accuracy and real-time performance of node location information have a great impact on the entire WSNs system. Compared with the idealized isotropic network, the anisotropic network (AN) is more suitable for practical applications. But the topological structure of AN has irregular and non-uniform characteristics, and the Distance Vector-Hop (DV-Hop) algorithm determines the distance by the number of hops. If an obstacle is encountered, the number of hops generated by the detour cannot indicate its true distance. Thus, the positioning accuracy of DV-Hop drops sharply in AN and cannot meet the positioning requirements. Therefore, we propose a new localization algorithm, the DV-Hop with Adaptive Step Variation Chaotic Fruit Fly Optimization Algorithm (CAFOA-DV-Hop). This algorithm has a greatly low localization error compared with two algorithm, DECHDV-Hop and DV-Hop on several topologies. In addition, a novel adaptive search step size strategy raised by us has strong global search ability and fully strengthens the balance of global and local optimization ability. When the global optimal solution is no longer updated, we also design a new chaotic strategy to jump out of the local optimal solution, which makes the method converge extremely fast. We evaluate the performance of our scheme using extensive simulations with localization accuracy under the influence of C-, S-, H-, O- and W-shaped topologies. The simulation results show that our proposed CAFOA-DV-Hop achieves higher localization accuracy than DECHDV-Hop and DV-Hop in different topology models, whether it is the change of the communication radius or the change of the anchor node ratio. Besides, the convergence speed of our proposed algorithm is also faster than the other two compared algorithms.