Efficient Load Balancing Techniques through Multi Objective Approaches in Cloud Data Center

Abstract

In the cloud, the resource-efficient load balancing is demanding one owing to the newlinesignificant number of user request on the server. With the dynamic balance of workload on Cloud Server (CS), the task scheduling efficiency is highly increased. Many research works have been implemented for achieving load balancing on the cloud. However, resource utilization has not been adequately reduced. For that reason, this research work is designed with three proposed techniques for ensuring the load balancing on CS with minimal time. In the proposed research, the scheduling efficiency and load balancing issues are addressed by the development of optimization techniques for the cloud environment. Besides, the proposed newlineresearch technique predicts the workload on CS by adapting smart workload prediction strategy in the cloud. The proposed Threshold Based Multi-Objective Memetic Optimized Round Robin Scheduling (T-MMORRS) Technique is implemented for task scheduling in cloud. The primary objective of T-MMORRS Technique is to attain better efficiency with less time. There are two significant phases being involved in T-MMORRS, one for workload detection, and another for load balancing in the cloud. The number of users requests are forwarded to the CS. newlineDuring the first phase, the burst detector module could identify workload as normal or bursty. After identifying the workload state, the load balancing is carried out in T-MMORRS Technique. Threshold Multi-Objective Memetic Optimization (TMMO) algorithm is applied in T-MMORRS when the workload is in a normal state. Besides, the Weighted Multi-Objective Memetic Optimized Round Robin Scheduling (WMMORRS) algorithm is employed during the bursty condition in T-MMORRS. With the utilization of TMMO and WMMORRS algorithms, better task scheduling is attained in T-MMORRS Technique with less time and higher efficiency in a cloud newline