【摘要】：隨機(jī)共振指的是利用輸入信號(hào)、噪聲和非線性系統(tǒng)間的某種匹配關(guān)系,令噪聲能量向信號(hào)轉(zhuǎn)移,使得噪聲降低的同時(shí),信號(hào)能量增強(qiáng),進(jìn)而使得輸出信噪比增強(qiáng)并達(dá)到極大值點(diǎn)。利用隨機(jī)共振可以將噪聲視為有利因素,這對(duì)非線性系統(tǒng)中信號(hào)處理領(lǐng)域的研究具有顯著的意義。本文研究了隨機(jī)共振在信道譯碼方面的應(yīng)用,主要工作內(nèi)容如下:首先,以誤比特率(Bit Error Ratio,BER)為測(cè)度,研究高斯噪聲作用下閾值陣列系統(tǒng)中的隨機(jī)共振現(xiàn)象及噪聲改善LDPC碼譯碼性能的情況。特別研究了閾值單元個(gè)數(shù)及系統(tǒng)閾值大小的變化對(duì)隨機(jī)共振現(xiàn)象的影響,并發(fā)現(xiàn)當(dāng)信號(hào)是閾下時(shí),噪聲強(qiáng)度增加,誤碼率先降后增,隨機(jī)共振發(fā)生,且在閾值單元個(gè)數(shù)較多時(shí),隨機(jī)共振現(xiàn)象越明顯;當(dāng)信號(hào)是閾上時(shí),噪聲總是惡化譯碼的性能,隨機(jī)共振不存在。其次,研究了兩種圖像傳輸中常見(jiàn)的噪聲——瑞利噪聲和萊斯噪聲作用下,閾值陣列系統(tǒng)中LDPC碼的譯碼性能改善效果。通過(guò)對(duì)比不同噪聲作用的仿真結(jié)果,可以發(fā)現(xiàn)不同噪聲之間的效果差異。特別研究了閾值單元個(gè)數(shù)及閾值大小的改變對(duì)隨機(jī)共振現(xiàn)象的影響,并發(fā)現(xiàn)萊斯噪聲作用下閾上隨機(jī)共振的存在性。最后,研究了圖像在LDPC編譯碼系統(tǒng)中的傳輸。以峰值信噪比(Peak Signal-to-noise Ratio,PSNR)為衡量標(biāo)準(zhǔn),展示了噪聲對(duì)其傳輸效果的改善作用,以及不同噪聲分布對(duì)接收端圖像質(zhì)量的影響,證實(shí)了噪聲對(duì)LDPC譯碼效果改善作用的可應(yīng)用性。
[Abstract]:Stochastic resonance refers to the use of a certain matching relationship between input signal, noise and nonlinear system to transfer the noise energy to the signal, so as to reduce the noise and increase the signal energy. Then the output signal-to-noise ratio is enhanced and the maximum value is reached. Noise can be regarded as a favorable factor by using stochastic resonance, which is of great significance to the research of signal processing in nonlinear systems. In this paper, the application of stochastic resonance in channel decoding is studied. The main work is as follows: firstly, the bit error rate (Bit Error Ratio,BER) is taken as the measure. The stochastic resonance phenomenon in threshold array system with Gao Si noise and the improvement of decoding performance of LDPC codes by noise are studied. The effects of the number of threshold units and the threshold value of the system on the stochastic resonance are studied. It is found that when the signal is subthreshold, the noise intensity increases, the bit error rate decreases first and then increases, and the stochastic resonance occurs when the number of threshold units is large. The more obvious the phenomenon of stochastic resonance; When the signal is on the threshold, the noise always deteriorates the decoding performance, and the stochastic resonance does not exist. Secondly, the improvement of decoding performance of LDPC codes in threshold array system is studied under the influence of Rayleigh noise and Rice noise, which are common in image transmission. By comparing the simulation results of different noises, we can find the difference between different noises. In particular, the effects of the number of threshold units and the threshold size on the stochastic resonance phenomenon are studied, and the existence of suprathreshold stochastic resonance is found under the influence of Rice noise. Finally, the transmission of image in LDPC codec and decode system is studied. The peak signal-to-noise ratio (Peak Signal-to-noise Ratio,PSNR) is taken as the standard to show the improvement of the transmission effect of noise and the influence of different noise distribution on the image quality of the receiver. It is proved that noise can improve the effect of LDPC decoding.
【學(xué)位授予單位】：南京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN911.22

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 王友國(guó);董洪程;劉健;;最佳匹配陣列隨機(jī)共振系統(tǒng)中利用噪聲改善信息傳輸[J];計(jì)算機(jī)應(yīng)用;2016年08期

2 李歡;王友國(guó);;一類非線性神經(jīng)網(wǎng)絡(luò)中噪聲改善信息傳輸[J];物理學(xué)報(bào);2014年12期

3 郭永峰;譚建國(guó);;一類非線性神經(jīng)網(wǎng)絡(luò)系統(tǒng)的超閾值隨機(jī)共振現(xiàn)象[J];物理學(xué)報(bào);2012年17期

4 王友國(guó);吳樂(lè)南;;極大并聯(lián)閾值網(wǎng)絡(luò)中噪聲改善信號(hào)的相關(guān)性[J];數(shù)據(jù)采集與處理;2006年04期

5 徐好民,王煜;隨機(jī)共振新實(shí)例——衛(wèi)星熱紅外溫度異常點(diǎn)的發(fā)現(xiàn)及其意義[J];國(guó)際地震動(dòng)態(tài);2003年12期

6 周寶晶,楊靈法,辛厚文;Pt(100)/NO+CO體系隨機(jī)共振的理論研究[J];高等學(xué)校化學(xué)學(xué)報(bào);1999年03期

7 鄭文衡,張國(guó)安,陳俊華;用隨機(jī)共振理論解釋地震孕育過(guò)程與地震發(fā)生現(xiàn)象中的一個(gè)顯著特征[J];地震研究;1998年03期

8 祝恒江,吳錫田;隨機(jī)共振研究進(jìn)展[J];大學(xué)物理;1997年07期

，

本文編號(hào)：2300056