本文選題：圖論 + 匹配��； 參考：《海南大學(xué)》2017年碩士論文

【摘要】：本研究旨在發(fā)掘和鑒別所測得的信道數(shù)據(jù)的指紋特點(diǎn),并證實(shí)其在不同場景的所具有的特性。本研究包括五部分內(nèi)容:無線信道特性;建模及特性提取;場景指紋辨識;場景復(fù)合指紋辨識;場景區(qū)域辨識與匹配。1.基于圖論知識介紹無線信道的特征�；谶@項(xiàng)研究工作,明確無線信道特征。2.基于三種場景的真實(shí)信道測量數(shù)據(jù),采用數(shù)字信號處理和主成分分析法對信道的特性進(jìn)行提取和建模。研究結(jié)果說明,該模型體現(xiàn)的信道特征與實(shí)際測量的數(shù)據(jù)有較好的吻合度。3.引入神經(jīng)網(wǎng)絡(luò)重點(diǎn)研究兩個(gè)待測場景的識別。通過對樣本數(shù)據(jù)的離線訓(xùn)練與在線識別匹配,使得待測場景都獲得匹配。實(shí)驗(yàn)結(jié)果表明,該辨識模型有效,學(xué)習(xí)的自適應(yīng)性較好。4.采用聚類分析研究無線信道復(fù)合場景的鑒別。對不同路段的分析結(jié)果進(jìn)行對比,得到結(jié)論:路段可以依據(jù)指紋劃分的區(qū)域數(shù)量進(jìn)行分類。研究結(jié)果表明,算法對信道的區(qū)分、辨識和分類的方法是有效的。5.采用時(shí)間序列分析和決策樹模型對某區(qū)域的場景識別與匹配研究。結(jié)果表明,所提供的兩個(gè)信道樣本誤判概率小。
[Abstract]:The purpose of this study is to explore and identify the fingerprint characteristics of the measured channel data and to verify their characteristics in different scenes. This research includes five parts: wireless channel characteristics; modeling and feature extraction; scene fingerprint identification; scene complex fingerprint identification; scene region identification and matching. 1. This paper introduces the characteristics of wireless channel based on graph theory. Based on this work, the wireless channel features. 2. 2. Based on the real channel measurement data of three scenarios, digital signal processing (DSP) and principal component analysis (PCA) are used to extract and model the channel characteristics. The results show that the channel characteristics of the model are in good agreement with the measured data. Neural network is introduced to study the recognition of two scenes to be tested. By off-line training of sample data and online recognition matching, the scene to be tested can be matched. The experimental results show that the identification model is effective and the learning adaptability is good. 4. 4. Cluster analysis is used to study the discrimination of wireless channel composite scene. The analysis results of different sections are compared, and the conclusion is drawn: the road sections can be classified according to the number of areas divided by fingerprints. The results show that the algorithm is effective for channel differentiation, identification and classification. Time series analysis and decision tree model are used to study the scene recognition and matching in a certain region. The results show that the error probability of the two channel samples is small.
【學(xué)位授予單位】：海南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：O157.5

【相似文獻(xiàn)】

相關(guān)期刊論文 前7條

1 姚朝灼;;圖論算法的可視化操作平臺(tái)設(shè)計(jì)[J];福州大學(xué)學(xué)報(bào)(自然科學(xué)版);2006年01期

2 許承斌,陳希有;無限域平行平面場的圖論算法[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);1986年01期

3 張良震;;圖論算法復(fù)雜性問題(一)[J];安徽大學(xué)學(xué)報(bào)(自然科學(xué)版);1988年03期

4 周巧臨;;實(shí)用的應(yīng)用圖論算法的單元及通道布線[J];科技信息(科學(xué)教研);2007年27期

5 安紹萱;奚建春;;求高階系統(tǒng)頻率響應(yīng)的圖論算法[J];南京郵電學(xué)院學(xué)報(bào);1987年01期

6 楊華康,任國鵬;鎖具裝箱問題中鎖具總數(shù)的圖論算法[J];工科數(shù)學(xué);1999年02期

7 徐亦文,趙永昌;樹的q-最優(yōu)劃分及其應(yīng)用[J];上海機(jī)械學(xué)院學(xué)報(bào);1990年01期

相關(guān)會(huì)議論文 前2條

1 吳軍;劉秀蘭;;兩個(gè)圖論算法的實(shí)現(xiàn)與界面設(shè)計(jì)[A];2007'中國儀器儀表與測控技術(shù)交流大會(huì)論文集（一）[C];2007年

2 史曉紅;劉文斌;;基于圖論算法的蛋白質(zhì)結(jié)構(gòu)的同源模建[A];’2004系統(tǒng)仿真技術(shù)及其應(yīng)用學(xué)術(shù)交流會(huì)論文集[C];2004年

，

本文編號：1784296