Abstract
In the ultra-wide band (UWB)-based indoor wireless positioning scene, non-line-of-sight (NLOS) is a common phenomenon due to the existence of the obstacles and it may result in the significant positioning errors. So it is an important research issue to recognize the UWB NLOS signals from the line-of-sight (LOS) signals. As an emerging deep learning technology, convolutional neural network (CNN) has been applied successfully to handle this issue. However, the traditional CNN omits the influence of measurement noises and does not consider the mining of the signal’s temporal relationship. To solve this problem, a UWB NLOS recognition method, called wavelet Gramian CNN (WGCNN), is proposed by integrating the wavelet analysis and the Gramian angular field (GAF). Firstly, the discrete wavelet transform is used to extract the trend curve of the channel pulse response (CIR) signals by eliminating the measurement noises. Further, two GAFs, consisting of Gramian angular summation field (GASF) and Gramian angular difference field (GADF), are introduced to transform the one-dimensional CIR signal into the two-dimensional images, which describe the temporal correlation of the signals. Moreover, a GAF image re-organization strategy is designed to capture the key sub-images to eliminate the influence of invalid image parts. Finally, the re-organized sub-images are fed into the CNN model to build a LOS/NLOS classifier. The experiments are performed on a benchmark data set provided by the EU "Horizon 2020" programme, and the results shows that the proposed WGCNN method has higher recognition accuracy than the traditional CNN method.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.