Article Preview
Top1. Introduction
The work, life, and learning style of people are progressively altered and changed because of the development of cloud systems (Li, et al., 2020). The cloud computing model is the most familiar computing concept. Cloud computing is an internet-based structure in which applications and files are hosted. Generally, thousands of network models were utilized for sharing and accessing the resources (Basu, et al., 2019). Cloud computing system is extremely familiar between application developer, owner, and managers (Ibrahim and Zainal, 2019). The major cause of the cloud computing need and familiarity is simple to handle application services with less cost maintainability of hardware and software stacks (Kalyampudi, et al., 2019). In recent years, load balancing in the cloud data center (Pushpalatha and Ramesh, 2020a) is the main concern, thus optimum physical host is effectively chosen in order to overcome the above problem. Moreover, many previous approaches were utilized to resolve direct load balancing issues (Zhao, et al., 2015). In addition, different studies are developed to evaluate and devise load balancing models in the cloud-based application, although handling information flow is a difficult process (Chen, et al., 2017). Nowadays, cloud computing has the main attention of the user in both industrial and academic surroundings, because of common commercial techniques. Therefore, the fast evolution of complicated cloud computing applications permits different enterprises and individuals to contract out imperative data to control local data centers (Ning, et al., 2017). Besides, cloud computing Pushpalatha and Ramesh, 2022) offers an easy and flexible technique for making files and data available for huge scale users (Huang, et al., 2018).
Edge-cloud computing models generally devise load prediction models (Pushpalatha and Ramesh, 2020b; Chanidini, et al., 2016) to execute realistic resource scheduling using load prediction values, when avoiding rental VM from the central cloud for increasing resource utilization and managing resources (Guo, et al., 2020). Meanwhile, the computing model attains temporary resource load data in edge cloud surroundings. Besides, load identification techniques devise measurable relationships among future and recent resource loads on period axis using historic load data, Moreover, the load prediction model offers a realistic source for edge-cloud resource arrangement and distribution as well as enhanced performance optimization, and decreases the whole rental price of users. Hence, the load prediction approach is the most important efficient system to optimize energy-saving and resource distribution in edge cloud computing (Li, et al., 2020). The major concept of the cloud is virtualization, and it is very expensive as well as it is normal technology in the cloud. Generally, VM is an excessive device for administrators in order to discrete software and hardware, as well as the flexibility to migrate VM from one Physical Machine (PM) to another. Moreover, VM migration also permits’ the load balancing to accumulate the energy and improve resource exploitation (Kapil, et al., 2013). VM is migrated from one to another data center for sustaining data centers deprived of disturbing VMs performance.