Binary Information Press
Rational allocation of resources to achieve high storage efficiency and load balancing is an important problem to be solved in the cloud storage. This paper presents a new method that based on improved adaptive dynamic programming algorithm to solve the problem. In the resource allocation process, virtualized storage nodes and grouped the storage data. And then the stored process of the data is stage decision optimized. Innovative to introduce 'Duplicate back-stepping strategy' to dynamic programming algorithm was improved and the efficiency of stored nodes combines the load balancing to acquire optimal allocation sequences.