Providing a Solution for Optimal Management of Resources using the Multi-objective Crow Search Algorithm in Cloud Data Centers

Abstract

Cloud data centers are a model of distributed systems that provide users with services to share data and information through the Internet. These centers also face challenges due to their popularity among users. With the increase in the number of hosts to respond to users' needs, challenges such as increasing power consumption, service level agreement violation, and time are seen. As a result, it is very important to address these challenges in these centers in order to reduce costs and increase profits. task scheduling for hosts is one of the most effective methods to improve productivity and optimal use of hosts' resources. In this process, with proper allocation, we can prevent hosts from becoming overloaded and increasing energy consumption due to inefficient use of hosts' resources. The proposed solution in this paper is to use multiple goals for the allocation process using the Crow search optimization algorithm. The Crow search optimization algorithm is new, fast, and powerful. As a result, in the proposed method by modeling this algorithm and considering the multi-criteria fitness function based on the requested resources of the tasks and the available resources of the host, we tend to manage resources properly. The simulation results show that the proposed method has a 9% reduction in service quality parameters such as power consumption compared to paper [21] and 15% compared to article [16], 11% execution time compared to paper [21], and 14% compared to the paper [16] and the service level agreement violation has improved by 16% compared to the paper [21] and 8% compared to the paper [16] and has been able to reduce the mentioned parameters.