基于DCT-CS和ATC的联合信源信道编码

Joint Source and Channel Coding Based on DCT-CS and ATC

提出了一种基于离散余弦变换的压缩感知(DCT-CS)编码与非对称Turbo码(ATC)的联合信源信道编码方法.该法将分块图像依据信号的自然属性和视觉特性先进行DCT变换,再按Zigzag扫描,并用压缩感知进行压缩,最后与非对称Turbo码结合实现了自适应图像压缩传输的联合信源信道编码-实验仿真结果表明:根据不同的无线信道环境自动采取不同的图像分块大小和压缩感知测量数以及非对称Turbo码类型,能够达到图像的传输速率和重建质量的优化平衡.

This paper proposed a joint source and channel coding scheme based on compressed sensing with discrete cosine transform （DCT-CS） coding and the asymmetric Turbo codes （ATC）. The block image is transformed by DCT according to the natural properties and visual characteristics, scanned by zigzag and compressed by compressed sensing, and then encoded with the asymmetric Turbo codes. Experimental results show that the image reconstructed quality and transmission efficiency can achieve the optimal balance by using different image block, different measure data and different types of asymmetric Turbo codes according to different channel state information.