Cloud computing is a modern paradigm for delivering services over the Internet. Load balancing is an important aspect of cloud computing, and it avoids being idle while some nodes become overloaded. Good load balance mechanisms make the entire system run more efficiently. Therefore in the existing method, the efficient load-balancing algorithm is used by a hybrid optimization algorithm. Although this hybrid approach produces better load balancing compared to the traditional method, it has some drawbacks. The exiting load-balancing algorithm takes more operational time for load balancing to overcome those shortcomings proposed method is used here. Initially, the available tasks are compiled by the clustering algorithm. Clustering the tasks can be done by balancing the load and reducing the virtual machine (VM). If there are no clusters, then single jobs are assigned to VM, and the need to use these virtual machines with dynamic creation increases in real-time. For clustering the available task, the proposed modified K means clustering algorithm is used. The proposed technique for load balancing used the optimization algorithm after the clustering procedure. In our recommended technique, the opposition fruit fly algorithm (OFFA) is used to balance the load. The performance of the proposed algorithm will be analyzed in terms of processing and convergence time, migration cost value and load utilization. Cloud SIM simulator can be implemented using Java.
Volume 12 | 07-Special Issue
Pages: 2777-2790
DOI: 10.5373/JARDCS/V12SP7/20202418