准循环LDPC码及其在等差错保护图像传输中的应用

发布时间：2018-04-16 18:45

本文选题：信道编码 + 准循环LDPC码　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：目前科学技术发展迅猛,其中高清晰电视、视频会议、远程监控以及遥感成像等技术得到广泛应用,人们生活中的信息交流已经主要依靠于多媒体通信。而通信系统往往会受到噪声干扰,使得无线信道中传输的图像质量大大下降,尤其是深空通信有着超远的传输距离、超复杂的传输环境以及自身设备及能源严格受限等特性,对信道纠错编码等关键技术提出了较高的要求,因此学习研究信道编码的意义深远。随着稀疏图码与译码中迭代理论的研究与实现,LDPC(低密度奇偶校验码)渐渐受到普遍关注,它性能逼近香农限、实现复杂度低、译码算法可并行且描述简单,因此LDPC码已经应用于各种新的通信标准。其中,有着独特奇偶校验矩阵结构的准循环LDPC码,能够实现具有线性复杂度的编码,是一种得到了广泛应用的结构化LDPC码。1)本文对于基于欧式有限几何方法构造的准循环LDPC码进行了深入研究,设计了CCSDS标准的基础(8176,7154)准循环LDPC码,并进行了实验仿真与分析。实验结果说明,与对应的随机LDPC码相比,合理构造的准循环LDPC码在编译码复杂度大大降低的基础上基本没有损失性能。另外,对于迭代次数对译码性能的影响做了基础研究。2)由于准循环LDPC码被广泛应用于各种实际工程,高效译码算法的重要性便体现出来。本文在研究了BP算法、LLR-BP算法、MS算法等基本译码方法的基础上深入探讨并实现了SBP算法以及其改进算法的实验仿真。结果说明,改进的SBP算法不仅在译码性能上略占优势,还降低了运算复杂度,提高了收敛速度。最后分析了译码错误产生的原因并综合考虑了系统性能的需求以及硬件资源,提出了一种改进的多点最小和算法并进行了实验仿真。该算法在很少增加复杂度的前提下,提高了译码算法的性能,减少了平均迭代次数,因此适合于对误码率性能和译码速度均有严格要求的高吞吐量LDPC码译码器的使用。3)最后针对CCSDS库中遥感图像的传输问题,提出了一种基于LDPC码的等差错保护遥感图像传输方案。此方案对原始遥感图像先进行小波变换分析,再根据嵌入式分块的方法进行信源编码。信道编码则运用基于LDPC码的联合编码来对嵌入式分块编码过后的输出码流进行同等程度的保护,使得码流的抗干扰能力得到提高。实验结果说明,相比于传统的RS、Tornado码,遥感图像的传输在此方案中更为可靠。最后本文总结了以上研究内容,并且针对准循环LDPC的编译码均提出了改进的方向。
[Abstract]:With the rapid development of science and technology, high-definition television, video conferencing, remote monitoring and remote sensing imaging are widely used. The communication of information in people's lives has mainly depended on multimedia communication.However, the communication system is often disturbed by noise, which makes the image quality in wireless channel decline greatly. Especially, the deep space communication has the characteristics of ultra-long transmission distance, ultra-complex transmission environment, strict limitation of its own equipment and energy, and so on.The key technologies such as channel error correction coding are required, so it is of great significance to study channel coding.With the research and implementation of iterative theory in sparse graph code and decoding, LDPC (low density parity check code) has been paid more and more attention. Its performance is close to Shannon limit, the complexity of implementation is low, the decoding algorithm can be parallel and simple to describe.Therefore, LDPC codes have been applied to various new communication standards.Among them, quasi-cyclic LDPC codes with unique parity check matrix structure can realize coding with linear complexity.It is a widely used structured LDPC code. 1) in this paper, the quasi-cyclic LDPC codes based on the Euclidean finite geometry method are studied in depth, and the quasi-cyclic LDPC codes based on the CCSDS standard are designed, and the experimental simulation and analysis are carried out.The experimental results show that compared with the corresponding random LDPC codes, the reasonably constructed quasi-cyclic LDPC codes have no loss of performance on the basis of greatly reducing the complexity of encoding and decoding.In addition, the influence of iteration times on decoding performance is studied. 2) because quasi-cyclic LDPC codes are widely used in various practical projects, the importance of efficient decoding algorithm is reflected.Based on the study of the basic decoding methods such as the BP algorithm and the LLR-BP algorithm and MS algorithm, this paper discusses and implements the SBP algorithm and the experimental simulation of its improved algorithm.The results show that the improved SBP algorithm not only has a slight advantage in decoding performance, but also reduces the computational complexity and improves the convergence speed.Finally, the causes of decoding errors are analyzed, and the requirements of system performance and hardware resources are considered synthetically. An improved multi-point minimum sum algorithm is proposed and the experimental simulation is carried out.The algorithm improves the performance of the decoding algorithm and reduces the average number of iterations without increasing the complexity.Therefore, it is suitable for high throughput LDPC decoder with strict performance and decoding speed. Finally, an equal-error protection remote sensing image transmission scheme based on LDPC code is proposed to solve the problem of remote sensing image transmission in CCSDS library.In this scheme, the original remote sensing image is analyzed by wavelet transform, and then the source coding is carried out according to the embedded block method.Channel coding uses joint coding based on LDPC code to protect the output bitstream after embedded block coding to the same degree, so that the anti-jamming ability of bitstream is improved.The experimental results show that the transmission of remote sensing images is more reliable than that of the traditional RSU Tornado code.Finally, this paper summarizes the above research contents, and proposes the direction to improve the encoding and decoding of quasi-cyclic LDPC.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.22

【相似文献】