本文關(guān)鍵詞： 跳頻通信 跳頻序列 參數(shù)估計(jì) 序列預(yù)測(cè)　出處：《電子科技大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：跳頻通信因具有抗干擾能力強(qiáng)、截獲率低等特點(diǎn)而被廣泛應(yīng)用,尤其是在軍事領(lǐng)域。與傳統(tǒng)的抗干擾方式不同,它通過偽隨機(jī)碼控制載波的跳變,從而有效躲避干擾信號(hào),克服了定頻通信的缺陷。然而隨著軍事電子對(duì)抗的升級(jí),如何有效干擾、截獲乃至利用對(duì)方跳頻信號(hào)已成為目前研究的一個(gè)熱點(diǎn)問題。目前研究該問題的方法通常是:首先,對(duì)接收到的跳頻信號(hào)進(jìn)行參數(shù)估計(jì),得到跳頻周期、跳頻時(shí)刻、跳頻頻率等相關(guān)參數(shù);其次,利用估計(jì)出來的跳頻頻率序列進(jìn)行建模預(yù)測(cè),估計(jì)出未來時(shí)刻的頻率;最后,借助同步跟蹤算法實(shí)現(xiàn)信號(hào)的同步。本文主要研究前兩個(gè)環(huán)節(jié),即跳頻信號(hào)參數(shù)估計(jì)和跳頻序列預(yù)測(cè)。跳頻信號(hào)參數(shù)估計(jì)作為首要環(huán)節(jié),其估計(jì)性能直接影響整個(gè)系統(tǒng)的性能。在研究基本時(shí)頻分析方法的基礎(chǔ)上,重點(diǎn)介紹了常用參數(shù)估計(jì)方法,綜合分析了基于STFT和SPWVD參數(shù)估計(jì)的性能。在此基礎(chǔ)上,針對(duì)跳頻信號(hào)這種特殊的非平穩(wěn)信號(hào),改進(jìn)了一種基于STFT的參數(shù)估計(jì)方法,該方法直接從STFT窗函數(shù)提取跳頻信號(hào)參數(shù),避開了STFT時(shí)頻分辨率低的限制,提高了參數(shù)估計(jì)的精度,仿真結(jié)果表明該方法可以實(shí)現(xiàn)參數(shù)的有效估計(jì)。跳頻序列預(yù)測(cè)作為中間環(huán)節(jié),其預(yù)測(cè)性能尤為關(guān)鍵。文中主要研究了基于移位寄存器產(chǎn)生的跳頻序列和基于混沌時(shí)間序列生成的跳頻序列的預(yù)測(cè)方法。針對(duì)前者,在研究基于其產(chǎn)生機(jī)理預(yù)測(cè)方法的基礎(chǔ)上,改進(jìn)了一種基于Berlekamp-Massey算法的移位寄存器序列預(yù)測(cè)方法,實(shí)現(xiàn)線性移位寄存器連續(xù)與非連續(xù)抽頭序列的有效預(yù)測(cè);至于后者,混沌跳頻序列因保密性強(qiáng)、數(shù)量巨大而具有巨大優(yōu)勢(shì),目前對(duì)該類序列的預(yù)測(cè)基本上是在相空間進(jìn)行的。因此在研究相空間重構(gòu)的基礎(chǔ)上,綜合分析局域預(yù)測(cè)法性能,重點(diǎn)研究了基于RBF神經(jīng)網(wǎng)絡(luò)、Bernstein多項(xiàng)式、Volterra自適應(yīng)濾波和支持向量機(jī)的混沌跳頻序列預(yù)測(cè)方法,通過理論分析和仿真實(shí)驗(yàn)比較其預(yù)測(cè)性能。同時(shí)考慮到工程實(shí)時(shí)性的需要,探討了直接和間接多步預(yù)測(cè)法,并在仿真基礎(chǔ)上分析其優(yōu)劣。最后,討論了非混沌同步跳對(duì)預(yù)測(cè)的影響,在具體應(yīng)用預(yù)測(cè)模型時(shí)應(yīng)通盤考慮。
[Abstract]:Frequency hopping communication is widely used because of its strong anti-jamming ability and low interception rate, especially in the military field. Different from the traditional anti-jamming mode, it controls the carrier jump by pseudo-random code. In order to effectively avoid interference signals, overcome the defects of fixed-frequency communication. However, with the upgrading of military electronic countermeasures, how to effectively interfere. Interception and even utilization of frequency hopping signals has become a hot issue. At present, the methods to study this problem are as follows: firstly, the parameters of the received frequency hopping signals are estimated and the frequency hopping period is obtained. Frequency hopping time, frequency hopping frequency and other related parameters; Secondly, the estimated frequency hopping sequence is used to model and predict the frequency of the future time. Finally, the synchronization algorithm is used to realize signal synchronization. In this paper, the first two links, namely frequency hopping signal parameter estimation and frequency hopping sequence prediction, are studied. The frequency hopping signal parameter estimation is the first step. The estimation performance directly affects the performance of the whole system. Based on the study of the basic time-frequency analysis methods, the commonly used parameter estimation methods are emphatically introduced. The performance of parameter estimation based on STFT and SPWVD is analyzed synthetically. Based on this, a parameter estimation method based on STFT is improved for frequency hopping signal, which is a special non-stationary signal. This method extracts the parameters of frequency hopping signal directly from the STFT window function, avoids the limitation of low time-frequency resolution of STFT, and improves the precision of parameter estimation. The simulation results show that this method can effectively estimate the parameters, and the frequency hopping sequence prediction is the intermediate link. In this paper, the prediction methods of frequency hopping sequence based on shift register and frequency hopping sequence based on chaotic time series are studied. A shift register sequence prediction method based on Berlekamp-Massey algorithm is improved on the basis of studying the prediction method based on its generation mechanism. The effective prediction of continuous and discontinuous tap sequences of linear shift registers is realized. As for the latter, chaotic frequency hopping sequences have great advantages because of their strong confidentiality and huge quantity. At present, the prediction of chaotic frequency hopping sequences is basically carried out in the phase space, so it is based on the study of phase space reconstruction. The performance of local prediction method is analyzed synthetically, and the Bernstein polynomial based on RBF neural network is studied emphatically. The prediction method of chaotic frequency hopping sequence based on Volterra adaptive filter and support vector machine is compared by theoretical analysis and simulation. The direct and indirect multistep prediction methods are discussed, and their advantages and disadvantages are analyzed on the basis of simulation. Finally, the influence of non-chaotic synchronous hopping on prediction is discussed, which should be taken into account in the application of the prediction model.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN914.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 況愛武,黃中祥;基于RBF神經(jīng)網(wǎng)絡(luò)的短時(shí)交通流預(yù)測(cè)[J];系統(tǒng)工程;2004年02期

2 張家樹,肖先賜;混沌時(shí)間序列的Volterra自適應(yīng)預(yù)測(cè)[J];物理學(xué)報(bào);2000年03期

3 甘建超,肖先賜;基于相空間鄰域的混沌時(shí)間序列自適應(yīng)預(yù)測(cè)濾波器(Ⅱ)非線性自適應(yīng)濾波[J];物理學(xué)報(bào);2003年05期

4 張曦;杜興民;王星;;基于S變換的跳頻信號(hào)特征參數(shù)盲估計(jì)[J];現(xiàn)代雷達(dá);2008年02期

，

本文編號(hào)：1452360