Fahimeh Ramezani et al., 2015 proposed the TBPSO approach builds a multi-objective optimization framework for the migration of tasks. The tasks are transferred from overloaded VMs to reduce task execution time and task transfer time by implementing the “Particle Swarm Optimization (PSO) algorithm”. It develops a multi-objective task scheduling optimization framework to transfer tasks from overloaded VMs which reduces both task execution time and transfer time and builds up a “Multi-Objective Particle Swarm Optimization (MOPSO)” based algorithm to resolve the configuration model. This collects data on the Task-Based blackboard as incoming data to the task migration framework. It helps to find a suitable VM host for the tasks focused on the VM power ratio. This determines a task size based on the combination of VM resources and the total number of assignments for each VM. The overloaded VMs and the candidate VMs list are considered for migrating the task. It also uses the PSO algorithm to solve the problem of finding the target VM for each migration task. The research was done on PSO algorithm by H. Xu et.al, 2016 and also, by Li-Der Chou et.al, 2018. Through PSO it measures the fitness for candidate VM based on the transfer time and the optimum allocation of tasks. Transfer time relies on VM bandwidth and task file size. The number of optical task assignments depends on the number of tasks allocated from PSO solution and the task limit.
It is costly and time-consuming.
It considers only the homogeneous VMs.
It migrates VMs from overloaded hosts alone, so that it eliminates the power consumption of overloaded hosts.
The Load balancing in cloud computing using Ant colony optimization was discussed by Wen et al. 2016. The load balancing with VM technology in cloud computing targets at assigning VMs to appropriate servers and also managing the usage of resources across all the servers. Dynamic input demands are requested in infrastructure-as-a-service environment, the different types of tasks that operate on them are responsible for creating VMs. To reduce the computation time and minimize virtual machine migration, a novel approach is proposed known as GESFLA.