A wireless sensor network location algorithm based on insufficient fingerprint information

Abstract

To realize sensor network-based positioning, the use of conventional techniques based on fingerprints has considerable cost due to the need of establishing in an offline manner a fingerprint database. Besides, it is also time-consuming when searching the database for calculating the localization solution. To address these drawbacks, we first propose to utilize the inverse distance weighted (IDW) interpolation method to improve the spatial resolution of the fingerprint database. The genetic algorithm learning machine (GAELM) is introduced to speed up the database lookup while enhancing the positioning accuracy of the fingerprint-based localization. Experiments show that the proposed Extreme Learning Location algorithm based on Insufficient Fingerprint information (ELL-IF) offers improved positioning performance over the BP-neural network (BP-NN)-based and extreme learning machine (ELM)-based methods.