An Adaptive Data Redundancy Strategy in Cloud Storage

Abstract

In the network of poor conditions and high error ratio, the stored data may be lost. If the lost data cannot be recovered effectively, it will make serious effects. Therefore, guaranteeing the availability and integrity of data is very important in any storage system, especially in cloud storage. The existing data redundancy strategies in cloud storage are unable to adapt to dynamic changes in the network environment. ADRS(Adaptive Data Redundancy Strategy), which combines fragment replication and LT(Luby Transform) code, is proposed. ADRS can adjust its parameters according to the current network state to optimize the performance of cloud storage networks. When the amount of source packets sent by the server nodes is smaller than the threshold, fragment replication will be the main storage mode and supplemented by LT code. When the amount of source packets is larger than the threshold, LT code is the main storage mode and supplemented by fragment replication. The simulation results show that ADRS fully integrates the advantages of fragment replication and LT code, which can reduce average delay and improve the reliability and stability of the system at the cost of increasing some storage space.