基于環(huán)境信息的無線信道指紋研究
發(fā)布時間：2018-07-15 17:46
【摘要】：無線信道與周圍的環(huán)境密切相關,不同環(huán)境下的無線信道具有一些差異化的特征。分析提取這些差異化的特征并將其應用,是當前的一個研究熱點。類比人類指紋,本文將上述無線信道的差異化特征稱為無線信道"指紋",在分析無線信道傳播特性的基礎上,通過對不同場景下的實測數據進行多維度的分析和處理,提取出相應場景的信道沖激響應作為特征研究無線信道指紋,表征不同無線傳播環(huán)境下信道指紋的差異性,建立無線信道"指紋"特征模型,并展開相應研究分析。具體工作如下:1、針對根據測試數據求解信道沖激響應問題,基于無線傳播環(huán)境中信道稀疏的重要特征,提出了稀疏正則最小二乘模型,在利用接收信號求解信道系數時兼顧求解精準度和信道稀疏特點,并給出了其基于二階錐規(guī)劃的求解方法。仿真實驗證明,該方法準確重構了原始信號,同時在一定程度上減少了信道造成的失真影響。2、在對無線信道"指紋"特征模型的分析中,提出了用于表述不同場景傳輸特性差異的"指紋"特征模型,該模型能刻畫不同場景中主要信道數目以及信道系數幅度的變化規(guī)律;提出了基于"指紋"特征模型的的場景識別分類器,該模型通過已知場景下的"指紋"特征訓練BP神經網絡建立映射關系,用以判別不同場景間的差異性進而用于場景的分類;提出了基于"指紋"特征鄰段聚類的連續(xù)路段場景聚類劃分模型,通過"指紋"特征模型實現連續(xù)區(qū)域路段下復雜場景的劃分,從而建立連續(xù)路段指紋庫并可為精準定位服務;提出了基于"指紋"特征的精準定位模型,將定位的過程被簡化為將一個未知位置的"指紋"特征與指紋庫中的信息進行比對匹配的過程,該模型的定位精度較高且可控。并對提出的模型進行仿真分析。
[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.
【學位授予單位】：北京交通大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TN92
本文編號：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.
【學位授予單位】：北京交通大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TN92
【參考文獻】
中國期刊全文數據庫 前7條
1 張劍云;夏杰;周青松;毛云祥;;基于基追蹤算法的寬帶LFM信號信道化偵察[J];數據采集與處理;2016年04期
2 李琰;;基于移動通信中的無線信道“指紋”特征模型研究[J];信息通信;2015年11期
3 劉留;陶成;陳后金;周濤;孫溶辰;邱佳慧;;高速鐵路無線傳播信道測量與建模綜述[J];通信學報;2014年01期
4 孟小峰;慈祥;;大數據管理:概念、技術與挑戰(zhàn)[J];計算機研究與發(fā)展;2013年01期
5 劉留;陶成;邱佳慧;陳后金;;高速鐵路平原場景無線信道小尺度衰落特征的研究[J];鐵道學報;2012年05期
6 劉留;陶成;邱佳慧;余立;董偉輝;;高速鐵路寬帶無線信道測量方法研究[J];鐵道學報;2011年05期
7 王曰芬;宋爽;熊銘輝;;基于共現分析的文本知識挖掘方法研究[J];圖書情報工作;2007年04期
中國博士學位論文全文數據庫 前2條
1 何睿斯;車載網絡復雜場景下無線信道測量與建模研究[D];北京交通大學;2015年
2 趙敏;高速鐵路無線寬帶信道測量、數據分析與建模研究[D];北京郵電大學;2014年
中國碩士學位論文全文數據庫 前6條
1 黨永亮;大數據分析在移動通信網絡優(yōu)化中的應用研究[D];華中師范大學;2015年
2 劉旭;高速鐵路環(huán)境下無線信道特性與建模研究[D];北京交通大學;2015年
3 張曉佳;陰影衰落信道的建模與分析[D];內蒙古大學;2014年
4 朱鵬;短波信道傳播特性和測量方法研究[D];電子科技大學;2008年
5 李軼華;GSM-R無線信道模型分析[D];西南交通大學;2007年
6 梅瓊;基于射線跟蹤法的微蜂窩小區(qū)電波傳播特性的研究[D];浙江工業(yè)大學;2004年
，本文編號：2124872
本文鏈接：http://sikaile.net/kejilunwen/xinxigongchenglunwen/2124872.html
