Analyzing Performance Parameters for Cloud Service Selection using AHP

As originally published there is an error in this document. The third author's name was given as "Aulick Shah ", on the document submitted for publication. It was intended to be "Aulick Regina Nenuphar Sah", as noted here. The original PDF remains unchanged.
In this era of massive computation there are many numerous cloud vendors who wish to offer services to customers by providing enormous computing power and storage system. Cloud service selection is often a tedious process that involves efficient decision making. When a user wants to host an application on to the cloud there is always a dilemma to choose the optimal service vendor at a less cost. There are many parameters to be considered to choose the optimal service provider. Since there are diverse number of services available and the vast number of options available, which usually complicates the cloud service selection process. This gives rise to the need for the design of an efficient methodology for the selection of cloud service provider. The existing research literature in the area of cloud service selection process involves hosting a single web application on the different types of instances within a single service provider, comparing the parameters of interest captured between different instances and letting the users know which instance type better suites their requirements. There is a need to know how the application behaves when it is run on different instance types of different vendors. In this paper we have categorized different types applications and done extensive experimentation on different cloud platforms. The aim this work is to come up with a systematic approach to rank the service providers based on cost, error rate, and throughput and response time.