智能光網(wǎng)絡(luò)物理層調(diào)制模式識別技術(shù)的研究
發(fā)布時間：2018-10-15 12:21
【摘要】：不像無線通信領(lǐng)域,調(diào)制模式識別技術(shù)起步早,發(fā)展相對成熟,在光通信領(lǐng)域,特別是智能光網(wǎng)絡(luò)(Automatic Switch Optical Network,ASON)的調(diào)制模式識別技術(shù)才剛剛發(fā)展。但是對光調(diào)制模式的識別,由于其潛在的研究價值和巨大的應(yīng)用前景,近些年來引起了越來越多的關(guān)注。為了從理論上更好的分析智能光網(wǎng)絡(luò)物理層光傳輸系統(tǒng),在本文的研究中采用ASON+DWDM組網(wǎng)方案,提出了一種基于密集波分復(fù)用(Dense Wavelength Division Multiplexing,DWDM)的非補(bǔ)償光傳輸系統(tǒng)(Uncompensation Transfer,UT)模型,并對該模型的推導(dǎo)進(jìn)行詳細(xì)討論。本文模式識別算法的研究就是基于該DWDM-UT系統(tǒng)模型展開的。研究中共涉及到四類共18種調(diào)制信號的判決,它們分別是強(qiáng)度調(diào)制信號MASK,相位調(diào)制信號MPSK,幅相調(diào)制信號MQAM和MAPSK。文中所提出的調(diào)制模式識別算法(Modulation Format Identification,MFI)主要基于調(diào)制信號高階累積量(High Order Cumulants,HOC)的特征值參數(shù),利用所設(shè)計的分類決策算法對不同的調(diào)制信號進(jìn)行分類判決。對分類判決算法中所使用的閾值,考慮到存在一些調(diào)制信號特征值參數(shù)隨信噪比變化的特性,提出了一種實時訓(xùn)練序列閾值優(yōu)化(Real-time Training Sequence Threshold Optimization,RT-TSTO)算法,通過對比測試,發(fā)現(xiàn)該算法能夠有效的保證閾值的精確性,極大地改善了不同調(diào)制信號的識別效果�？紤]到智能光網(wǎng)絡(luò)物理層DWDM-UT系統(tǒng)受色散(D)、非線性效應(yīng)(γ)和傳輸距離(L)的影響,我們分別研究了這些因素的改變對模式識別性能的影響,并利用仿真工具M(jìn)atlab給出了相應(yīng)的仿真結(jié)果。與此同時,為了進(jìn)一步驗證本文所提出的MFI算法的有效性,我們對當(dāng)前已經(jīng)廣泛投入商用的基于PM-QPSK調(diào)制的高速率Nyquist WDM也進(jìn)行了討論,通過VPI和Matlab的聯(lián)合仿真,結(jié)果表明該系統(tǒng)即使經(jīng)過超長距離的傳輸,PM-QPSK調(diào)制信號的識別率也能達(dá)到96.5%以上。以上仿真結(jié)果對今后智能光網(wǎng)絡(luò)物理鏈路的工程實施提供了一定的理論依據(jù)。
[Abstract]:Unlike in wireless communication, modulation pattern recognition technology starts early and develops relatively mature. Modulation pattern recognition technology in optical communication field, especially in intelligent optical network (Automatic Switch Optical Network,ASON), has just been developed. However, the recognition of optical modulation pattern has attracted more and more attention in recent years because of its potential research value and great application prospect. In order to analyze the physical layer optical transmission system of intelligent optical network better in theory, an uncompensated optical transmission system (Uncompensation Transfer,UT) model based on dense wavelength division multiplexing (Dense Wavelength Division Multiplexing,DWDM) is proposed by using ASON DWDM networking scheme in this paper. The derivation of the model is discussed in detail. The research of pattern recognition algorithm in this paper is based on the DWDM-UT system model. In the study, there are four types of 18 kinds of modulation signals, namely, intensity modulation signal, MASK, phase modulation signal, MPSK, amplitude-phase modulation signal, MQAM and MAPSK., respectively. The proposed modulation pattern recognition algorithm (Modulation Format Identification,MFI) is mainly based on the eigenvalue parameters of the higher-order cumulant (High Order Cumulants,HOC) of the modulation signal. The proposed classification decision algorithm is used to classify different modulated signals. For the threshold used in classifying decision algorithm, considering the fact that there are some characteristic parameters of modulation signal changing with signal-to-noise ratio, a real-time training sequence threshold optimization (Real-time Training Sequence Threshold Optimization,RT-TSTO) algorithm is proposed. It is found that the algorithm can effectively guarantee the accuracy of the threshold and greatly improve the recognition effect of different modulation signals. Considering that the physical layer DWDM-UT system of the intelligent optical network is affected by the dispersion (D), nonlinear effect (緯) and the transmission distance (L), we study the effect of these factors on the pattern recognition performance. The corresponding simulation results are given by using the simulation tool Matlab. At the same time, in order to further verify the effectiveness of the proposed MFI algorithm, we also discuss the high rate Nyquist WDM based on PM-QPSK modulation, which has been widely used in commercial applications, and through the joint simulation of VPI and Matlab. The results show that the recognition rate of PM-QPSK modulation signal can reach more than 96.5% even though the system is transmitted over a long distance. The above simulation results provide a theoretical basis for the implementation of the physical link of the intelligent optical network in the future.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TN929.1
本文編號：2272539
[Abstract]:Unlike in wireless communication, modulation pattern recognition technology starts early and develops relatively mature. Modulation pattern recognition technology in optical communication field, especially in intelligent optical network (Automatic Switch Optical Network,ASON), has just been developed. However, the recognition of optical modulation pattern has attracted more and more attention in recent years because of its potential research value and great application prospect. In order to analyze the physical layer optical transmission system of intelligent optical network better in theory, an uncompensated optical transmission system (Uncompensation Transfer,UT) model based on dense wavelength division multiplexing (Dense Wavelength Division Multiplexing,DWDM) is proposed by using ASON DWDM networking scheme in this paper. The derivation of the model is discussed in detail. The research of pattern recognition algorithm in this paper is based on the DWDM-UT system model. In the study, there are four types of 18 kinds of modulation signals, namely, intensity modulation signal, MASK, phase modulation signal, MPSK, amplitude-phase modulation signal, MQAM and MAPSK., respectively. The proposed modulation pattern recognition algorithm (Modulation Format Identification,MFI) is mainly based on the eigenvalue parameters of the higher-order cumulant (High Order Cumulants,HOC) of the modulation signal. The proposed classification decision algorithm is used to classify different modulated signals. For the threshold used in classifying decision algorithm, considering the fact that there are some characteristic parameters of modulation signal changing with signal-to-noise ratio, a real-time training sequence threshold optimization (Real-time Training Sequence Threshold Optimization,RT-TSTO) algorithm is proposed. It is found that the algorithm can effectively guarantee the accuracy of the threshold and greatly improve the recognition effect of different modulation signals. Considering that the physical layer DWDM-UT system of the intelligent optical network is affected by the dispersion (D), nonlinear effect (緯) and the transmission distance (L), we study the effect of these factors on the pattern recognition performance. The corresponding simulation results are given by using the simulation tool Matlab. At the same time, in order to further verify the effectiveness of the proposed MFI algorithm, we also discuss the high rate Nyquist WDM based on PM-QPSK modulation, which has been widely used in commercial applications, and through the joint simulation of VPI and Matlab. The results show that the recognition rate of PM-QPSK modulation signal can reach more than 96.5% even though the system is transmitted over a long distance. The above simulation results provide a theoretical basis for the implementation of the physical link of the intelligent optical network in the future.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TN929.1
【引證文獻(xiàn)】
相關(guān)期刊論文 前1條
1 余庚;;基于約束的ASON生存性探討[J];光通信技術(shù);2018年01期
，本文編號：2272539
本文鏈接：http://sikaile.net/kejilunwen/xinxigongchenglunwen/2272539.html
最近更新
教材專著
