Customer-satisfaction-aware and deadline-constrained profit maximization problem in cloud computing

As a new and modern service method that can meet the requirements of customers effectively, cloud computing is becoming more and more popular. For cloud service providers, they play an important role in constructing the cloud computing platform, which is convenient for customers to enjoy the services without paying attention to the execution process of the services. For cloud service providers and customers, they focus on profit and satisfaction respectively in service supply-demand relationship, and the pursuit of their objectives both have an influence on each other. On one hand, the configuration of the cloud computing platform is the main solution for cloud service providers to earn profit, which affects customer satisfaction based on specific constraints. On the other hand, customer satisfaction affects the arrival rate of service requests, which will affect the profit in turn. In this paper, we devote ourselves to analyzing a deadline constrained profit maximization problem in a cyclic cascade queuing system, in which the maximum tolerance of customers towards task waiting time is taken into consideration. On this basis, the definition of customer satisfaction is given. And then, a mathematical model is formulated in detail on how customer satisfaction affects the revenues of cloud service providers, and further affects their profits. However, the exact solutions of such model can hardly be calculated due to its complexity. Hence, we propose a heuristic algorithm to find the high-quality solution, then the optimal configuration of cloud computing platform can be obtained to achieve the maximization of the profit for cloud service providers on the basis of promoting customer satisfaction. At last, a series of numerical simulations are conducted to validate the performance of the proposed algorithm. The results show that the proposed algorithm can not only optimize profit and customer satisfaction simultaneously but also maintain them at a high level in the long run of the system. (C)& nbsp;2022 Elsevier Inc. All rights reserved.