UWB NLOS/LOS Classification Using Deep Learning Method

Ultra-Wide-Band (UWB) was recognized as its great potential in constructing accurate indoor position system (IPS). However, indoor environments were full of complex objects, the signals might be reflected by the obstacles. Compared with the Line-Of-Sight (LOS) signal, the signal transmitting path delay contained in None-Line-Of-Sight (NLOS) signal would induce positive distance errors and position errors. Before employing ranging information from the channels to calculate the position, LOS/NLOS classification or identification was necessary for selecting the “clean” channels. In conventional method, features extracted from the UWB channel impulse response (CIR) or some other signal properties were employed as the input vector of the machine learning methods, e.g. Support Vector Machine (SVM), Multi-layer Perception (MLP). Deep learning methods represented by Convolutional neural network (CNN) and Long Short-Term Memory (LSTM) had performed superior performance in dealing with time series data classification. In this pap er, deep learning method CNN-LSTM was employed in the UWB NLOS/LOS signal classification. UWB CIR data was directly input to the CNN-LSTM. CNN was employed for exploring and extracting the features automatically, and then, the CNN outputs were fed into the LSTM for classification. Open source datasets collected from seven different sites were employed in the experiments. Classification accuracy of CNN-LSTM with different settings was compared for analyzing the performance. The results showed that CNN-LSTM obtained stat e-of-art classification performance.