基于环境信息的无线信道指纹研究
发布时间：2018-07-15 17:46
【摘要】：无线信道与周围的环境密切相关,不同环境下的无线信道具有一些差异化的特征。分析提取这些差异化的特征并将其应用,是当前的一个研究热点。类比人类指纹,本文将上述无线信道的差异化特征称为无线信道"指纹",在分析无线信道传播特性的基础上,通过对不同场景下的实测数据进行多维度的分析和处理,提取出相应场景的信道冲激响应作为特征研究无线信道指纹,表征不同无线传播环境下信道指纹的差异性,建立无线信道"指纹"特征模型,并展开相应研究分析。具体工作如下:1、针对根据测试数据求解信道冲激响应问题,基于无线传播环境中信道稀疏的重要特征,提出了稀疏正则最小二乘模型,在利用接收信号求解信道系数时兼顾求解精准度和信道稀疏特点,并给出了其基于二阶锥规划的求解方法。仿真实验证明,该方法准确重构了原始信号,同时在一定程度上减少了信道造成的失真影响。2、在对无线信道"指纹"特征模型的分析中,提出了用于表述不同场景传输特性差异的"指纹"特征模型,该模型能刻画不同场景中主要信道数目以及信道系数幅度的变化规律;提出了基于"指纹"特征模型的的场景识别分类器,该模型通过已知场景下的"指纹"特征训练BP神经网络建立映射关系,用以判别不同场景间的差异性进而用于场景的分类;提出了基于"指纹"特征邻段聚类的连续路段场景聚类划分模型,通过"指纹"特征模型实现连续区域路段下复杂场景的划分,从而建立连续路段指纹库并可为精准定位服务;提出了基于"指纹"特征的精准定位模型,将定位的过程被简化为将一个未知位置的"指纹"特征与指纹库中的信息进行比对匹配的过程,该模型的定位精度较高且可控。并对提出的模型进行仿真分析。
[Abstract]:The wireless channel is closely related to the surrounding environment. It is a research hotspot to analyze and extract the characteristics of these differences and apply them. Analogous to human fingerprint, this paper refers to the difference characteristic of wireless channel as "fingerprint" of wireless channel. On the basis of analyzing the propagation characteristics of wireless channel, we analyze and process the measured data in different scenarios in many dimensions. The channel impulse response of the corresponding scene is extracted as the feature to study the fingerprint of the wireless channel, the differences of the fingerprint in different wireless propagation environment are represented, the characteristic model of the fingerprint of the wireless channel is established, and the corresponding research and analysis are carried out. The specific work is as follows: 1. In order to solve the impulse response problem based on the test data, a sparse regular least square model is proposed based on the important characteristics of channel sparsity in wireless propagation environment. When the received signal is used to solve the channel coefficient, the accuracy of the solution and the channel sparsity are taken into account, and the method based on the second-order cone programming is given. The simulation results show that the method can reconstruct the original signal accurately and reduce the distortion effect of the channel to a certain extent. In the analysis of the "fingerprint" characteristic model of wireless channel, the simulation results show that the proposed method can effectively reconstruct the original signal and reduce the distortion caused by the channel to a certain extent. A fingerprint feature model is proposed to describe the difference of transmission characteristics between different scenes. The model can describe the variation of the number of the main channels and the amplitude of the channel coefficients in different scenarios. A scene recognition classifier based on "fingerprint" feature model is proposed. BP neural network is trained by "fingerprint" feature of known scene to establish mapping relationship, which can be used to distinguish the difference between different scenes and then be used for scene classification. The scene clustering model of continuous road sections based on the clustering of "fingerprint" feature adjacent segment is proposed. The "fingerprint" feature model is used to realize the classification of complex scenes under the continuous section of road, so that the fingerprint database of continuous section can be established and can be used for accurate location. An accurate location model based on "fingerprint" features is proposed. The process of location is simplified to the process of matching the "fingerprint" feature of an unknown position with the information in the fingerprint database. The location accuracy of the model is high and controllable. The proposed model is simulated and analyzed.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92
本文编号：2124872
[Abstract]:The wireless channel is closely related to the surrounding environment. It is a research hotspot to analyze and extract the characteristics of these differences and apply them. Analogous to human fingerprint, this paper refers to the difference characteristic of wireless channel as "fingerprint" of wireless channel. On the basis of analyzing the propagation characteristics of wireless channel, we analyze and process the measured data in different scenarios in many dimensions. The channel impulse response of the corresponding scene is extracted as the feature to study the fingerprint of the wireless channel, the differences of the fingerprint in different wireless propagation environment are represented, the characteristic model of the fingerprint of the wireless channel is established, and the corresponding research and analysis are carried out. The specific work is as follows: 1. In order to solve the impulse response problem based on the test data, a sparse regular least square model is proposed based on the important characteristics of channel sparsity in wireless propagation environment. When the received signal is used to solve the channel coefficient, the accuracy of the solution and the channel sparsity are taken into account, and the method based on the second-order cone programming is given. The simulation results show that the method can reconstruct the original signal accurately and reduce the distortion effect of the channel to a certain extent. In the analysis of the "fingerprint" characteristic model of wireless channel, the simulation results show that the proposed method can effectively reconstruct the original signal and reduce the distortion caused by the channel to a certain extent. A fingerprint feature model is proposed to describe the difference of transmission characteristics between different scenes. The model can describe the variation of the number of the main channels and the amplitude of the channel coefficients in different scenarios. A scene recognition classifier based on "fingerprint" feature model is proposed. BP neural network is trained by "fingerprint" feature of known scene to establish mapping relationship, which can be used to distinguish the difference between different scenes and then be used for scene classification. The scene clustering model of continuous road sections based on the clustering of "fingerprint" feature adjacent segment is proposed. The "fingerprint" feature model is used to realize the classification of complex scenes under the continuous section of road, so that the fingerprint database of continuous section can be established and can be used for accurate location. An accurate location model based on "fingerprint" features is proposed. The process of location is simplified to the process of matching the "fingerprint" feature of an unknown position with the information in the fingerprint database. The location accuracy of the model is high and controllable. The proposed model is simulated and analyzed.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92
【参考文献】
中国期刊全文数据库 前7条
1 张剑云;夏杰;周青松;毛云祥;;基于基追踪算法的宽带LFM信号信道化侦察[J];数据采集与处理;2016年04期
2 李琰;;基于移动通信中的无线信道“指纹”特征模型研究[J];信息通信;2015年11期
3 刘留;陶成;陈后金;周涛;孙溶辰;邱佳慧;;高速铁路无线传播信道测量与建模综述[J];通信学报;2014年01期
4 孟小峰;慈祥;;大数据管理:概念、技术与挑战[J];计算机研究与发展;2013年01期
5 刘留;陶成;邱佳慧;陈后金;;高速铁路平原场景无线信道小尺度衰落特征的研究[J];铁道学报;2012年05期
6 刘留;陶成;邱佳慧;余立;董伟辉;;高速铁路宽带无线信道测量方法研究[J];铁道学报;2011年05期
7 王曰芬;宋爽;熊铭辉;;基于共现分析的文本知识挖掘方法研究[J];图书情报工作;2007年04期
中国博士学位论文全文数据库 前2条
1 何睿斯;车载网络复杂场景下无线信道测量与建模研究[D];北京交通大学;2015年
2 赵敏;高速铁路无线宽带信道测量、数据分析与建模研究[D];北京邮电大学;2014年
中国硕士学位论文全文数据库 前6条
1 党永亮;大数据分析在移动通信网络优化中的应用研究[D];华中师范大学;2015年
2 刘旭;高速铁路环境下无线信道特性与建模研究[D];北京交通大学;2015年
3 张晓佳;阴影衰落信道的建模与分析[D];内蒙古大学;2014年
4 朱鹏;短波信道传播特性和测量方法研究[D];电子科技大学;2008年
5 李轶华;GSM-R无线信道模型分析[D];西南交通大学;2007年
6 梅琼;基于射线跟踪法的微蜂窝小区电波传播特性的研究[D];浙江工业大学;2004年
，本文编号：2124872
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2124872.html
