基于BP神经网络的RSSI测距优化算法

姚军; 甄梓越; 马宇静

doi:10.12265/j.cjors.2021177

基于BP神经网络的RSSI测距优化算法

doi: 10.12265/j.cjors.2021177

西安科技大学通信与信息工程学院，西安 710054

基金项目: 教育部下一代互联网技术创新项目“IPv6在煤矿生产监测中的应用”（NGII20160301）

详细信息

作者简介:
姚军：（1974—），男，陕西人，西安科技大学副教授，兼任西安科技大学网络中心副主任

甄梓越：（1995—），男，陕西人，西安科技大学硕士研究生，主要研究领域为电子科学与技术

马宇静：（1997—），女，陕西人，西安科技大学硕士研究生，主要研究领域为信息与通信工程

通讯作者:
甄梓越 E-mail: 429531824@qq.com

中图分类号: TP391
计量
- 文章访问数: 157
- HTML全文浏览量: 42
- PDF下载量: 14
- 被引次数: 0
出版历程
- 收稿日期: 2021-06-29
- 录用日期: 2021-11-24
- 网络出版日期: 2021-11-24

RSSI ranging optimization algorithm based on BP neural network

College of Communication and Information Engineering, Xi’an University of Science and Technology, Xi’an 710054, China

摘要

摘要: 基于接收信号强度指示(received signal strength indication, RSSI)测距的研究应用领域很广泛，一直是物联网研究的热点. 为降低传统基于反向传播(back propagation，BP)神经网络的RSSI测距误差，文中提出一种基于K-means聚类算法对样本数据进行预处理的BP神经网络测距方法，来解决由于RSSI值衰减程度不同引起的不同距离区间RSSI值和真实距离之间映射关系不均匀的问题. 将K-means聚类算法应用在BP神经网络模型中，对样本数据进行距离区间划分，然后将已经分类好的数据分别输入BP神经网络建立网络模型并进行实验仿真. 结果显示，基于BP神经网络的RSSI测距算法的均方根误差为1.425 7 m，而经过K-means算法改进后的BP神经网络测距方法的均方根误差为1.288 7 m，降低了测距误差，并优化了目标RSSI与真实距离的映射关系.
- 路径损耗模型 /
- 接收信号强度指示(RSSI)测距 /
- K-means聚类算法 /
- 反向传播(BP)神经网络 /
- 测距误差 /
- 均方根误差
Abstract: The research of ranging based on received signal strength indication (RSSI) has been a hot spot in the Internet of Things due to its wide application. In order to reduce the traditional RSSI ranging error based on back propagation neural network, a back propagation neural network RSSI ranging method based on K-means clustering algorithm preprocessing sample data is proposed. The algorithm solves the problem of non-uniform mapping between RSSI value and real distance in different distance intervals due to different attenuation degree of RSSI value. The K-means clustering algorithm is applied to BP neural network model to divide the distance interval of the sample data, and then the classified data are input into BP neural network to build the network model and simulate the experiment. The results show that the root mean square error of RSSI ranging method based on BP neural network is 1.4257 m, while the root mean square (RMS) error of BP ranging method neural network improved by K-means algorithm is 1.2887 m. Compared with the existing ranging methods based on BP neural network, it reduces the ranging error and optimizes the mapping relationship between the target RSSI and the real distance.
- path loss model /
- RSSI ranging /
- K-means clustering algorithm /
- BP neural network /
- Ranging error /
- root mean square error

HTML全文

图 1 BP神经网络流程图

Fig. 1 Flowchart of BP neural network

下载: 全尺寸图片幻灯片

图 2 改进测距算法主要流程

Fig. 2 Process of the ranging optimization algorithm

下载: 全尺寸图片幻灯片

图 3 BP神经网络结构

Fig. 3 Network structure of BP neural network

下载: 全尺寸图片幻灯片

图 4 K-means分类前后数据分布

Fig. 4 Data distribution before and after K-means classification

下载: 全尺寸图片幻灯片

图 5 分类后BP神经网络训练效果

Fig. 5 Training effect of BP neural network after classification

下载: 全尺寸图片幻灯片

图 6 改进前后拟合效果对比

Fig. 6 Comparison of fitting effect before and after improvement

下载: 全尺寸图片幻灯片

表 1 补充后的训练数据

Tab. 1 Supplementary training data

RSSI/dBm	距离/m	RSSI/dBm	距离/m
−24	1.0	−41	11.5
−25	1.5	−41	12.0
−25	2.0	−42	12.5
−26	2.5	−43	13.0
−27	3.0	−43	13.5
−27	3.5	−43	14.0
−29	4.0	−43	14.5
−30	4.5	−43	15.0
−31	5.0	−44	15.5
−33	5.5	−44	16.0
−33	6.0	−45	16.5
−33	6.5	−46	17.0
−35	7.0	−48	17.5
−36	7.5	−48	18.0
−36	8.0	−49	18.5
−37	8.5	−50	19.0
−38	9.0	−50	19.5
−38	9.5	−51	20.0
−39	10.0	−53	20.5
−40	10.5	−53	21.0
−40	11.0

下载: 导出CSV

参考文献(15)

[1]	邓昀, 朱彦, 杨逸夫, 等. 基于BP神经网络的RFID室内定位算法研究[J]. 小型微型计算机系统,2019,40(8):1707-1712. DOI: 10.3969/j.issn.1000-1220.2019.08.024 DENG Y, ZHU Y, YANG Y F, et al. Research on RFID indoor location algorithm based on BP neural network[J]. Journal of Chinese computer systems,2019,40(8):1707-1712. (in Chinese) DOI: 10.3969/j.issn.1000-1220.2019.08.024
[2]	WANG J, WANG Y, GUAN X. An indoor localization system based on backscatter RFID tag[C]// IEEE Wireless Communications and Networking Conference. IEEE, 2016: 1-6.
[3]	宋宁佳, 崔英花. 基于GA-GRNN的RFID室内定位算法[J]. 计算机工程,2019,45(11):298-302+308. SONG N J, CUI Y H. RFID indoor positioning algorithm based on GA-GRNN[J]. Computer engineering,2019,45(11):298-302+308. (in Chinese)
[4]	邹东尧, 陈鹏伟, 刘宽. 一种改进的RSSI测距定位算法[J]. 电讯技术,2019,59(10):1191-1196. DOI: 10.3969/j.issn.1001-893x.2019.10.014 ZOU D Y, CHEN P W, LIU K. An improved RSSI ranging algorithm[J]. Telecommunication engineering,2019,59(10):1191-1196. (in Chinese) DOI: 10.3969/j.issn.1001-893x.2019.10.014
[5]	谭志, 张卉. 无线传感器网络RSSI 定位算法的研究与改进[J]. 北京邮电大学学报,2013,36(3):88-91+107. TAN Z, ZHANG H. Research and improvement of RSSI localization algorithm in wireless sensor networks[J]. Journal of Beijing University of Posts and Telecommunications,2013,36(3):88-91+107. (in Chinese)
[6]	石琴琴, 徐强, 张建平, 等. 基于RSSI测距的室内目标定位方法[J]. 电子测量与仪器学报,2018,32(2):86-91. SHI Q Q, XU Q, ZHANG J P, et al. Indoor targeting method based on RSSI ranging[J]. Journal of electronic measurement and instrument,2018,32(2):86-91. (in Chinese)
[7]	SASIWAT Y, JINDAPETCH N, BURANAPANICHKIT D, et al. An experimental study of human movement effects on RSSI levels in an indoor wireless network[C]//The 12th Biomedical Engineering International Conference (BMEiCON). IEEE, 2019: 1-5.
[8]	IVANIĆ M, MEZEI I. Distance estimation based on rssi improvements of orientation aware nodes[C]// Zooming Innovation in Consumer Technologies Conference (ZINC). IEEE, 2018: 140-143.
[9]	SHI Y, LONG Y, LU F, et al. Indoor RSSI trilateral algorithm considering piecewise and space-scene[C]// IEEE International Conference on Smart Cloud (SmartCloud), 2017: 278-282.
[10]	费扬, 杜庆治. 基于BP神经网络模型的RSSI测距方法研究[J]. 电波科学学报,2018,33(2):195-201. FEI Y, DU Q Z. RSSI ranging method based on BP neural network model[J]. Chinese journal of radio science,2018,33(2):195-201. (in Chinese)
[11]	LI H, HUANG Z, SUN H, et al. A dynamic adaptive indoor ranging model based on RSSI[C]//The 39th Chinese Control Conference (CCC). IEEE, 2020: 2850-2855.
[12]	余振宝, 卢小平, 刘英, 等. 一种改进BP神经网络的接收信号强度测距算法[J]. 测绘科学, 2020, 45(11): 48-52+67. YU Z B, LU X P, LIU Y, et al. A ranging algorithm of received signal strength based on improved BP neural network [J]. Science of surveying and mapping, 2020, 45 (11): 48-52 + 67. (in Chinese)
[13]	TANG J, HAN J. An improved received signal strength indicator positioning algorithm based on weighted centroid and adaptive threshold selection[J]. Alexandria engineering journal,2021,60(4):3915-3920. DOI: 10.1016/j.aej.2021.02.031
[14]	WANG C, WU F, SHI Z, et al. Indoor positioning technique by combining RFID and particle swarm optimization-based back propagation neural network[J]. Optik,2016,127(17):6839-684. DOI: 10.1016/j.ijleo.2016.04.123
[15]	盖晴晴, 韩玉兵, 南华, 等. 基于深度卷积神经网络的极化雷达目标识别[J]. 电波科学学报,2018,33(5):575-582. GAI Q Q, HAN Y B, NAN H, et al. Polarimetric radar target recognition based on depth convolution neural network[J]. Chinese journal of radio science,2018,33(5):575-582. (in Chinese)