Localization for mobile sensor network based on unscented Kalman filter with clipping and uncorrelated conversion

• DOI: 10.1177/1550147717741104
• Publication Date: 2017-11-30T13:04:02Z
• Authors: Dan Jia; Weihua Li; Peng Wang
• Keywords: Uncorrelated; Clipping (morphology)

This article studies localization for mobile sensor network with incomplete range measurements, that is, the ranges between some pairs of sensors cannot be measured. Different from existing works localize one by one, problem in this is solved whole network. According to whether can measured or not, all are grouped construct basic units. For units, a constrained nonlinear model first established formulate their relative motion, where motion states chosen as and cosine values angles ranges. Then, based on model, unscented Kalman filter adopted provide state estimation. In filter, clipping technique introduced handle constraints, uncorrelated conversion make full use measurements. Hence, estimation accuracy improved. Finally, distributed multidimensional scaling-map method used using estimated ranges, algorithm presented. The effectiveness advantages proposed demonstrated through several simulation examples.