基于海量存储云调度机制的云网络数据存储算法

针对当前云网络超宽带存储难以实现流量的分流，且存储过程中存在严重的存储效率低下及存储时延过高等问题，提出了一种基于海量存储云调度机制的云网络数据超带宽存储算法。采用周期调度及梯度优化方式，且综合考虑数据存储中使用强度，最小传输粒度等数字特征，对数据存储过程中的指纹梯度进行优化，且将该梯度引入到数据传输过程中，成功地实现了数据的流量分离，提高了数据存储效率。仿真实验表明，与当前广泛使用的超线性存储调度算法（super linear memory scheduling algorithm，SLMS）相比，本文算法的存储效率更高，能够在流量分离的前提下显著降低存储时延。

Research on cloud data storage algorithm based on mass storage cloud scheduling

In order to solve the current cloud storage network to realize ultra wideband traffic diversion, storage efficiency and storage delay serious problems and the storage process, we proposed a mass data storage cloud scheduling mechanism based on super cloud network bandwidth storage algorithm. The periodic scheduling and gradient optimization methods, and considering the data stored in the use of strength, the minimum size of digital transmission characteristics, fingerprint gradient data storage process was optimized, and the gradient is introduced into data transmission process, the successful implementation of the data flow separation, improve the efficiency of data storage. The simulation results show that compared with the widely used linear storage scheduling algorithm, the proposed algorithm has high storage efficiency and can greatly reduce the storage delay under the premise of traffic separation.