A METAHEURISTIC FOR DYNAMIC LOAD BALANCING IN CLOUD ENVIRONMENTS

In the past years, cloud computing has become a popular paradigm that provides computingover the internet. With the internet facility, resources are provided dynamically to the end usersin an on-demand fashion. Users could be billions in number accessing the cloud. Their need forservices have been increasing at an alarming rate. To enhance the performance of the system,resources should be used efficiently. Cloud environments have been defined in order to providethe effective services and have different advantages but they encounter some challenges suchas load balancing. Load balancing is one of the main issues in cloud computing, in whichworkload is distributed dynamically to all the nodes. Also, with the rapid development andwide application of cloud computing, how to reduce the energy consumption of cloud datacenter has become a major issue of green cloud data center. The present study provides ametaheuristic Firefly Algorithm (FA), which aims to shorten the distance with optimumsthrough multiple iterations and reduce the cost on the process of task allocated to achieve theefficient search for the optimal virtual machines. The results show that FA achieves the highspeedsearch for the virtual machines which are more eligible to perform tasks and save energyconsumption of cloud data center relatively. The obtained results show better performance forFA in comparison with particle swarm optimization algorithm (PSO).

Keywords:

Dynamic load balancing , firefly algorithm , cloud environment , virtual machines