eMRA: an efficient multi-optimization based resource allocation technique for infrastructure cloud

Cloud computing, an emerging internet based computing paradigm, provides resources for on-demand requests from various geographically distributed data centers. The allocation techniques direct the on-demand requests to the suitable data centers for effective resource utilization. Various parameters like virtual machine (VM) cost, data transfer cost; response time and request processing time are responsible for efficient data center allocation. The majority of the works consider single parameter optimization technique with objective of user or cloud service provider. In reality the user and service provider have different objectives. The selection of data center could be conceivable through considering these objectives. In this proposed work, we suggest efficient multi-optimization resource allocation model (eMRA) using optimization techniques to achieve the objectives of users and the data centers. Social group optimization (SGO) is proposed to optimize the user requests considering various related parameters for allocation. Likewise, particle swam optimization (PSO) is applied to optimize data center list that are suitable for the optimized user requests. The eMRA considers distinctive related parameters of user request, data center and network to design the model that separate the design model from other existing works. The eMRA technique is simulated using CloudAnalyst and the performance is studied for ten different scenarios under three existing broker policy of CloudAnalyst. The performance of eMRA is studied and compared with benchmark mechanisms and found better in its class.