Selecting the SLA guarantee by evaluating the QOS availability

Abstract

Consumers face confusion when selecting a Service level agreement from a Cloud Service Provider (CSP) due to the wide range of available options, such as Platform-as-a-Service, Infrastructure-as-a-Service, Software-as-a-Service and Network-as-a-service. CSPs provide their services to consumers based on the availability of computing and networking, which define the nature of the service and the corresponding costs. However, strict SLA adherence and achieving quality of service (QoS) can be challenging. In many cases, the availability of services falls short, making it difficult for consumers to choose the most suitable and guaranteed SLA among those offering similar functionality. Therefore, ensuring the availability of QoS becomes crucial for SLA selection and user satisfaction in cloud computing. In this paper, we introduce a fuzzy logic system-based model designed to classify SLA into 11 levels, ranging from 90% to 99.999%, contingent upon the QOS availability of computing (up and downtime) and QOS availability of networking (bandwidth, jitter, round-trip time, and packet loss) metrics. The research had two primary objectives: (i) To develop a versatile SLA model that employs nuanced techniques, diverging from typical CSP offerings, addressing both dominant forms of QoS availability. (ii) To improve the precision of SLA categorization, tailored to each user's specific requirements, enhancing task efficiency and cost-effectiveness. Our research was conducted using the MATLAB program.