Bluetooth Fingerprint based Indoor Localization using Bi-LSTM

Bluetooth fingerprint has been widely applied in the filed of indoor localization due to its advantages of low power consumption and easy deployment. Currently, most Bluetooth fingerprint based localization methods construct the location estimators by use of the machine learning techniques, such as k-nearest neighbor (KNN), support vector machine (SVM) and random forest (RF), which hardly make full use of massive localization data. Aiming at this problem, this paper introduces in the idea of deep learning and proposes a Bluetooth fingerprint based localization algorithm using bidirectional long short-term memory (Bi-LSTM) network. In the proposed algorithm, the Bi-LSTM network is used instead of machine learning techniques to fully learn from the localization data and improve the localization accuracy. Experimental results show that, compared with the traditional localization methods using machine learning, the proposed algorithm achieves less root mean square error and is superior in localization accuracy.