擴(kuò)譜通信抗干擾的現(xiàn)代信號(hào)處理應(yīng)用研究
發(fā)布時(shí)間：2018-10-30 13:25
【摘要】：憑借優(yōu)異的信息隱藏能力和易于組網(wǎng)等優(yōu)點(diǎn),擴(kuò)譜通信技術(shù)已成為抗干擾通信中的主流方法。單純依賴擴(kuò)譜抗干擾技術(shù)難以適應(yīng)未來戰(zhàn)場(chǎng)復(fù)雜電磁環(huán)境下的軍事通信需求。為了進(jìn)一步提高現(xiàn)有通信裝備的抗干擾能力,本文圍繞擴(kuò)譜通信抗干擾中的現(xiàn)代信號(hào)處理方法及相關(guān)擴(kuò)展應(yīng)用展開研究。 本文首先探討了利用盲源分離混合矩陣逆問題求解方法進(jìn)行擴(kuò)譜通信抗干擾的可行性,設(shè)計(jì)了點(diǎn)對(duì)點(diǎn)和組網(wǎng)方式下基于盲源分離的擴(kuò)譜通信抗干擾系統(tǒng)結(jié)構(gòu),提出了結(jié)合負(fù)熵對(duì)照函數(shù)的盲源分離直接序列擴(kuò)譜抗相關(guān)干擾算法。為提高強(qiáng)噪聲環(huán)境下的抗干擾效果,充分利用實(shí)際采樣速率較高的系統(tǒng)特性,設(shè)計(jì)了基于均值濾波的盲源分離抗相關(guān)干擾算法。分析了非協(xié)作盲接收環(huán)境下用戶異步時(shí)延的構(gòu)成,從前后符號(hào)在當(dāng)前接收窗口的有效能量角度推導(dǎo)了用戶異步時(shí)延對(duì)抗干擾算法收斂性的影響,并利用三角不等式進(jìn)行了論證。 考慮跳頻載頻的跳變對(duì)混合信號(hào)統(tǒng)計(jì)獨(dú)立性的影響,設(shè)計(jì)了基于盲源分離的跳頻通信系統(tǒng)抗干擾算法,并對(duì)各種類型干擾進(jìn)行了算法性能仿真及硬件測(cè)試。根據(jù)跳頻信號(hào)與各種干擾信號(hào)之間時(shí)頻分布均不相同的特性,提出了基于時(shí)頻聯(lián)合的盲源分離抗干擾算法,從矩陣的聯(lián)合對(duì)角化出發(fā)實(shí)現(xiàn)了跳頻信號(hào)與干擾信號(hào)的有效分離�；谔l圖案等先驗(yàn)信息,設(shè)計(jì)了基于跳頻圖案的半盲分離跳頻通信抗干擾算法。 根據(jù)掃頻干擾信號(hào)的特點(diǎn),設(shè)計(jì)了匹配掃頻信號(hào)的非線性變換,提出實(shí)時(shí)掃頻信號(hào)特征檢測(cè)方法,通過自學(xué)習(xí)獲得了能量聚集特性優(yōu)異的掃頻干擾壓縮感知字典。根據(jù)多音干擾與跳頻信號(hào)在時(shí)間、頻率分布上不一致,分別提出了自學(xué)習(xí)多音干擾壓縮感知字典和自學(xué)習(xí)跳頻信號(hào)壓縮感知字典構(gòu)造方法,使得低信噪比時(shí)能夠識(shí)別多音干擾和正常的跳頻信號(hào),為實(shí)時(shí)可靠地解調(diào)創(chuàng)造了有利條件。 將基于形態(tài)結(jié)構(gòu)特征學(xué)習(xí)的字典構(gòu)造與壓縮感知方法結(jié)合,提出多形態(tài)自學(xué)習(xí)壓縮感知跳頻通信抗干擾算法。利用自學(xué)習(xí)壓縮感知字典的稀疏表達(dá)能力,設(shè)計(jì)了單通道自學(xué)習(xí)壓縮感知跳頻信號(hào)抗干擾算法,在自學(xué)習(xí)過程中將檢測(cè)到的強(qiáng)干擾分量去除以減小干擾影響。理論分析和仿真驗(yàn)證表明本文算法在低信噪比和強(qiáng)干擾條件下具有優(yōu)異而高效的抗干擾能力。 最后,總結(jié)了信號(hào)自身內(nèi)在形態(tài)結(jié)構(gòu)特征的廣泛存在性,提出匹配信號(hào)形態(tài)特征的稀疏表達(dá)域自學(xué)習(xí)構(gòu)造方法,從而可以在較小的字典規(guī)模下高效地進(jìn)行信號(hào)的重構(gòu),為壓縮感知方法實(shí)用化進(jìn)行了有益的探索,并拓展應(yīng)用于圖像超分辨率重構(gòu),獲得了優(yōu)異的重構(gòu)效果。
[Abstract]:With the advantages of excellent information hiding ability and easy networking, the spread spectrum communication technology has become the mainstream method in anti-interference communication. It is difficult to meet the requirements of military communication in the complex electromagnetic environment of the future battlefield by relying solely on the spread spectrum anti-jamming technology. In order to further improve the anti-jamming capability of the existing communication equipment, this paper focuses on the modern signal processing methods and related extended applications in the spread spectrum communication. In this paper, we first discuss the feasibility of solving the inverse problem of blind source separation hybrid matrix for anti-jamming of spread spectrum communication, and design the anti-jamming system structure of spread spectrum communication based on blind source separation under point-to-point and networking mode. A blind source separation (BSS) direct sequence spread spectrum (DSS) anti-correlation interference algorithm combining negative entropy contrast function (NSE) is proposed. In order to improve the anti-interference effect in strong noise environment and make full use of the high sampling rate of the system, a blind source separation anti-correlation interference algorithm based on mean filter is designed. This paper analyzes the structure of user asynchronous delay in non-cooperative blind reception environment, deduces the influence of user asynchronous delay on the convergence of anti-jamming algorithm from the point of view of the effective energy of front and rear symbols in the current receiving window, and proves it by using triangular inequality. Considering the influence of hopping frequency hopping on the statistical independence of mixed signals, a frequency hopping communication system based on blind source separation is designed, and the algorithm performance simulation and hardware test are carried out. According to the different time-frequency distribution between frequency-hopping signals and various interference signals, an anti-jamming algorithm for blind source separation based on time-frequency joint is proposed. The effective separation of frequency-hopping signal and interference signal is realized from the diagonalization of matrix. Based on the prior information such as frequency hopping pattern, a semi-blind anti-jamming algorithm for frequency hopping communication is designed. According to the characteristics of the frequency-sweeping interference signal, the nonlinear transformation of the matched frequency-sweeping signal is designed, and the method of detecting the feature of the real-time frequency-sweeping signal is proposed. Through self-learning, a dictionary of scan interference compression perception with excellent energy aggregation characteristic is obtained. According to the difference of time and frequency distribution between multi-tone interference and frequency-hopping signal, a self-learning multi-tone interference compression perception dictionary and a self-learning frequency-hopping signal compression perception dictionary are proposed, respectively. It makes it possible to recognize multi-tone interference and normal frequency-hopping signal at low signal-to-noise ratio (SNR), which creates favorable conditions for real-time and reliable demodulation. By combining the dictionary construction based on morphological structure feature learning with the compression sensing method, a multi-morphological self-learning compression sensing frequency hopping communication anti-jamming algorithm is proposed. Using the sparse expression ability of self-learning compression perceptual dictionary, a single-channel self-learning compression sensing frequency hopping signal anti-jamming algorithm is designed. In the process of self-learning, the detected strong interference components are removed to reduce the interference effect. Theoretical analysis and simulation results show that the proposed algorithm has excellent and efficient anti-jamming capability under the condition of low SNR and strong interference. Finally, this paper summarizes the extensive existence of the inherent morphological structure features of the signal itself, and proposes a sparse expression domain self-learning construction method to match the morphological features of the signal, so that the signal can be reconstructed efficiently on a small dictionary scale. This paper makes a useful exploration for the practical application of compression sensing, and extends its application to image super-resolution reconstruction, and obtains excellent reconstruction results.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN975
本文編號(hào)：2300127
[Abstract]:With the advantages of excellent information hiding ability and easy networking, the spread spectrum communication technology has become the mainstream method in anti-interference communication. It is difficult to meet the requirements of military communication in the complex electromagnetic environment of the future battlefield by relying solely on the spread spectrum anti-jamming technology. In order to further improve the anti-jamming capability of the existing communication equipment, this paper focuses on the modern signal processing methods and related extended applications in the spread spectrum communication. In this paper, we first discuss the feasibility of solving the inverse problem of blind source separation hybrid matrix for anti-jamming of spread spectrum communication, and design the anti-jamming system structure of spread spectrum communication based on blind source separation under point-to-point and networking mode. A blind source separation (BSS) direct sequence spread spectrum (DSS) anti-correlation interference algorithm combining negative entropy contrast function (NSE) is proposed. In order to improve the anti-interference effect in strong noise environment and make full use of the high sampling rate of the system, a blind source separation anti-correlation interference algorithm based on mean filter is designed. This paper analyzes the structure of user asynchronous delay in non-cooperative blind reception environment, deduces the influence of user asynchronous delay on the convergence of anti-jamming algorithm from the point of view of the effective energy of front and rear symbols in the current receiving window, and proves it by using triangular inequality. Considering the influence of hopping frequency hopping on the statistical independence of mixed signals, a frequency hopping communication system based on blind source separation is designed, and the algorithm performance simulation and hardware test are carried out. According to the different time-frequency distribution between frequency-hopping signals and various interference signals, an anti-jamming algorithm for blind source separation based on time-frequency joint is proposed. The effective separation of frequency-hopping signal and interference signal is realized from the diagonalization of matrix. Based on the prior information such as frequency hopping pattern, a semi-blind anti-jamming algorithm for frequency hopping communication is designed. According to the characteristics of the frequency-sweeping interference signal, the nonlinear transformation of the matched frequency-sweeping signal is designed, and the method of detecting the feature of the real-time frequency-sweeping signal is proposed. Through self-learning, a dictionary of scan interference compression perception with excellent energy aggregation characteristic is obtained. According to the difference of time and frequency distribution between multi-tone interference and frequency-hopping signal, a self-learning multi-tone interference compression perception dictionary and a self-learning frequency-hopping signal compression perception dictionary are proposed, respectively. It makes it possible to recognize multi-tone interference and normal frequency-hopping signal at low signal-to-noise ratio (SNR), which creates favorable conditions for real-time and reliable demodulation. By combining the dictionary construction based on morphological structure feature learning with the compression sensing method, a multi-morphological self-learning compression sensing frequency hopping communication anti-jamming algorithm is proposed. Using the sparse expression ability of self-learning compression perceptual dictionary, a single-channel self-learning compression sensing frequency hopping signal anti-jamming algorithm is designed. In the process of self-learning, the detected strong interference components are removed to reduce the interference effect. Theoretical analysis and simulation results show that the proposed algorithm has excellent and efficient anti-jamming capability under the condition of low SNR and strong interference. Finally, this paper summarizes the extensive existence of the inherent morphological structure features of the signal itself, and proposes a sparse expression domain self-learning construction method to match the morphological features of the signal, so that the signal can be reconstructed efficiently on a small dictionary scale. This paper makes a useful exploration for the practical application of compression sensing, and extends its application to image super-resolution reconstruction, and obtains excellent reconstruction results.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN975
【參考文獻(xiàn)】
相關(guān)期刊論文 前3條
1 石光明;劉丹華;高大化;劉哲;林杰;王良君;;壓縮感知理論及其研究進(jìn)展[J];電子學(xué)報(bào);2009年05期
2 彭耿;黃知濤;姜文利;周一宇;;單通道盲信號(hào)分離研究進(jìn)展與展望[J];中國電子科學(xué)研究院學(xué)報(bào);2009年03期
3 劉良云,李英才,相里斌;超分辨率圖象重構(gòu)技術(shù)的仿真實(shí)驗(yàn)研究[J];中國圖象圖形學(xué)報(bào);2001年07期
，本文編號(hào)：2300127
本文鏈接：http://sikaile.net/kejilunwen/wltx/2300127.html
最近更新
教材專著
