Scalable and Cost-Efficient Algorithms for Reliable and Distributed Cloud Storage

This paper focuses on minimizing jointly data storage and networking costs in a distributed cloud storage environment. We present two new efficient algorithms to place encrypted data chunks and enhance data availability when guaranteeing a minimum cost of storage and communication in the same time. The proposed underlying solutions, based on linear programming approach lead to an exact formulation with convergence times feasible for small and medium network sizes. A new polynomial time algorithm is presented and shown to scale to much larger network sizes. Performance assessment results, using simulations, show the scalability and cost-efficiency of the proposed distributed cloud storage solutions.