Robust multiple description coding — Joint Coding for source and storage
暂无分享,去创建一个
We propose a framework for robust content distribution in networks such that each network node stores a description of the source for users to access and it is robust against any single node failure. The fundamental problem is to identify the tradeoff among various parameters such as storage size, repair bandwidth and the level of distortion in the reconstructed estimate. We show that when we design a robust multiple description code, it is usually favourable that the descriptions should be as correlated as possible to reduce the amount of repair bandwidth in our network.
[1] Yunnan Wu,et al. Network coding for distributed storage systems , 2010, IEEE Trans. Inf. Theory.
[2] Vivek K. Goyal,et al. Multiple description coding with many channels , 2003, IEEE Trans. Inf. Theory.
[3] Abbas El Gamal,et al. Achievable rates for multiple descriptions , 1982, IEEE Trans. Inf. Theory.
[4] Vivek K. Goyal,et al. Multiple description coding: compression meets the network , 2001, IEEE Signal Process. Mag..