Setting up SLAs using a dynamic pricing model and behavior analytics in business and marketing strategies in cloud computing

Abstract Increasing amounts of data are being generated every year. Sustainable computing systems have become capable of extracting and learning information from the underlying data. Edge and AI (artificial intelligence) are expanding into industrial systems requiring new computing and networking infrastructure. Due to this, SLA computing is becoming increasingly challenging to handle in these emerging cloud environments. The cloud is a service that provides virtual resources to users. Qualitative and quantitative findings in market-oriented approaches are one of the most common methods for managing virtual and physical machines in a network. When allocating services, price is an important factor to consider. In this study, we aim to determine the initial price of VMs while considering the dynamic pricing model in a competitive, sustainable computing system. Besides negotiation-based trading, a multifactor architecture is used for trading in the marketplace. Based on the simulation results, it was found that the performance could be improved by categorizing the VMs based on regression. According to the simulation results, the cloud market system provides a better service-level agreement (SLA) and response time when assigning virtual machines to the market. Based on the results, we found that using the regression method for categorizing the VMs to manage the market improved the SLA.