基于卷积神经网络的OFDM-UWB信道环境识别

doi:10.19678/j.issn.1000-3428.0058181

摘要/Abstract

摘要： 超宽带（Ultra-wideband, UWB）无线通信技术如今被广泛应用在室内定位领域中，识别出多径信道中的非视距（non line of sight，NLOS）信道，有助于去除影响信号的非理想因素，提升定位精度。本文基于 OFDM 方案的 UWB 超宽带系统，提出一种利用卷积神经网络对信道估计出的信道冲激响应时频图像进行训练，从而识别出信道环境的方法，将信道识别问题转化为图像识别问题，同时对时频处理参数对识别结果的影响进行分析研究。该方法的识别率随着通信系统比特信噪比（EbN0）的增加而提升，在 EbN0 增加至 20dB 时稳定在 90%。同时使用传统的信道识别方法与本文研究方法进行对比，本研究方法具有 10%的性能提升。

Abstract: In the field of indoor positioning, Ultra-Wideband (UWB) wireless communication technology is widely used. Identifying non-line of sight (NLOS) channel in multi-path channel is helpful to remove non-ideal factors and improve positioning accuracy. This paper proposes a LOS/NLOS channel identification method by using a convolutional neural network (CNN) to identifying the impulse response figures of the channels based on OFDM-UWB system, transform the channel identification problem into an image recognition problem. The influence of time and frequency processing parameters on the identification results is analyzed. The recognition rate of this method increases with the increase of the bit signal-to-noise ratio (EbN0), and it is stable at 90% when EbN0 increases to 20dB. At the same time, the traditional channel recognition method is compared with the research method in this paper, and this research method has a 10% performance improvement.

王斐 , 徐湛 , 职如昕. 基于卷积神经网络的OFDM-UWB信道环境识别[J]. 计算机工程, doi: 10.19678/j.issn.1000-3428.0058181.

参考文献

[1] Maria-Gabriella, Di Benedetto, Guerino Giancola. Understanding ultra wide band radio fundamentals[M]. (葛利嘉，朱林，袁晓芳，等译). Beijing: Publishing House of Electronics Industry, 2005. 1~2.
[2] Kim D H , Kwon G R , Pyun J Y , et al. NLOS identification in UWB channel for indoor positioning[C]// 2018 15th IEEE Annual Consumer Communications & Networking Conference (CCNC). IEEE, 2018.
[3] MENG W, HUANG B, GaodFrag G. A novel approach to NLOS identification in sensor localization. In: Proc. 26th Chinese control and decision conf., Changsha, China, Jul. 2014. p. 876–80.
[4] YAN L, LU Y, ZHANG Y. An improved NLOS identification and mitigation approach for target checking for wireless sensor networks, IEEE Access, 5 (2017), pp. 2798-2807
[5] WU C, YANG Z, ZHOU Z, et al. PhaseU: Real-time LOS identification with WiFi[C]. Computer Communications. IEEE, 2015
[6] ABBASIA, LIU H. Improved line-of-sight/non-line-of-sight classification methods for pulsed ultrawideband localisation. Iet Communications 8(5), 680-688(2014).
[7] ALMAZROUEI, E, SINDI, N.A., Al-Araji, S.R., et al. Measurement and analysis of NLOS identification metrics for WLAN systems. In: IEEE International Symposium on Personal (2015).
[8] ZHANG Hao, LIANG Xiaolin, LV Tingting, et al. A novel Non-Line-of-Sight identification algorithm based on skewness[J]. Telecommunication engineering, 015, 000(005):484-490. (in Chinese)张浩, 梁晓林, 吕婷婷,等. 一种新颖的基于偏度的非视距区分算法[J]. 电讯技术, 2015, 000(005):484-490.
[9] TABAA M, DIOU C, AROUSSI M E, et al. LOS and NLOS identification based on UWB stable distribution[C]. International Conference on Microelectronics. IEEE, 2014.
[10] AFIF J A, SEONG L K, Krishnan S, A SVM approach to UWB-IR based positioning under NLOS conditions[C]. IEEE International Conference on Communication Systems. IEEE, 2011.
[11] CHITAMBIRA B, ARMOUR S, Wales S, et al. NLOS identification and mitigation for geolocation using least-Squares support vector machines[C]. IEEE Wireless Communications & Networking Conference. IEEE, 2017.
[12] LEE K, OH J, YOU K. Closed-Form solution of TDOA-based geolocation and tracking: a recursive weighted least square approach[J]. wireless personal communications, 2016, 94(4):1-14.
[13] CHU H, WU C D. A non-parametric propagation condition identification method and non-line of sight mitigation algorithm for wireless sensor network[J]. Open Electrical & Electronic Engineering Journal, 2016, 10(1):80-87.
[14] YANG X, ZHAO F, CHEN T. NLOS identification for UWB localization based on import vector machine[J]. AEU - International Journal of Electronics and Communications, 2018: S1434841117322379.
[15] TABAA M, SAADAOUI S, CHEHAITLY M, et al. NLOS identification for UWB body communications[J]. International Journal of Computer Applications, 2015, 124(6):12-17.
[16] MIAOZ M, ZHAO L W, YUAN W W, et al. Application of one-Class classification in NLOS identification of UWB positioning[C]. 2016 International Conference on Information System and Artificial Intelligence (ISAI). IEEE, 2016.
[17] TANG S , PAN Z , ZHOU X . Low-rank and sparse based deep-fusion convolutional neural network for crowd counting[J]. Mathematical Problems in Engineering, 2017, 2017(pt.9):5046727.1-5046727.11.
[18]YU Yuhai, LIN Hongfei, MENG Jiana, et al. Assembling deep neural networks for medical compound figure detection[J]. Information (Switzerland), 2017, 8(2):48.
[19] LU Jiwen, Abdulnabi, et al. Multi-task CNN model for attribute prediction[J]. IEEE Transactions on Multimedia, 2015.
[20] KECELI A S. Viewpoint projection based deep feature learning for single and dyadic action recognition[J]. Expert Systems with Application, 2018, 104(aug.):235-243.[21] WANG Fei, XU Zhan, ZHI Ruxin, et al. LOS/NLOS channel identification technology based on CNN[C]. NAFOSTED Conference on Information and Computer Science (NICS), 2019, Thailand.

选择文件类型/文献管理软件名称

选择包含的内容