基于Tornado码的存储冗余算法研究  

Research in Redundancy Storage Algorithm Based on Tornado Code

作　　者：孙伟平[1] 汤毅凡[1] 

机构地区：[1]华中科技大学计算机学院,武汉430074

出　　处：《微处理机》2008年第2期71-74,共4页Microprocessors

基　　金：国家自然科学基金资助项目(60373088)

摘　　要：纠删码是提供数据高可用性的重要方法。与传统纠删码相比较,Tornado码是基于稀疏矩阵的级连纠删码,通过异或操作完成编解码,实现纠错功能。经实验分析,在存储系统中采用Tornado码进行数据冗余,可以有效的加快编解码速度和减少存储空间的消耗,并提供较高的纠错能力。Erasure code is an important method for providing high availability in data storage. Comparing with the existing kinds of erasure codes,the Tornado code is the one based on a cascade of sparse bipartite graphs, encoded and decoded through the operation of the exclusive - or （ XOR）, implement the function of correcting. By simulations, adopting Tornado code for data redundancy in the storage system can accelerate the speed of encoding and decoding, reduce the cost of storage, and providing the good ability of correcting.

关 键 词：纠删码 高可用性 编解码 纠错 冗余 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]