Earticle

Localization is one of the most important research topics in the wireless sensor network applications. To improve the indoor localization accuracy, the centroid localization algorithm based on Mamdani fuzzy system has been adopted to attain the weight between sensor node and anchor node. This paper proposes a novel optimized input membership function by bat algorithm in fuzzy inference system using the data of received signal strength in real indoor condition. The author has realized the algorithm on Zigbee platform and the experimental comparison on other different centroid localization algorithms indicates that Mamdani fuzzy inference adopting the membership function optimized by bat algorithm renders smaller mean localization errors.

As one of the most important applications of wireless sensor network, positioning technology has become a extremely hot research filed. Taking into account of the position of beacon nodess, they make an effect on the positioning accuracy of unknown nodes. Therefore, in this paper, it proposes a new weighted centroid localization algorithm based on the traditional algorithm. For the selection of weight, the distance between beacon nodes and unknown nodes, and the slide length of the triangle are used to formed as the weighted factor. Experiment Simulation results show that this algorithm increased localisition accuracy than that of traditional algorithm.