Load Balancing Based on Statistical Model in Expert Cloud

Expert cloud is a new class of cloud computing which enables the users to achieve their requirements from a collection containing experts and skills created by the human resources (hrs). the acquisition of these skills and experts from this collection is possible by using the internet and cloud computing concepts without consideration of the hrs location. the load balancing in cloud computing means equal load distribution among resources, virtual machines (vms) and servers. the effective load distribution in a heterogeneous environment such as cloud is an important challenge. the increase in the number of users, the differences of request types and also different resources capabilities and capacities cause that some resources become overload and some others become idle. this paper presents a dynamic load balanced task scheduling algorithm in expert cloud. in this method, we utilize the genetic algorithm (ga) as a ranking for making distinction among the hrs capabilities. in our proposed method, we use interval estimation and specification matrix to allocate the hrs and also to determine the service rate. we model the load balancing and mapping process based on simple exponential smoothing and probability theory. this statistical load balancing model allows us to allocate the hrs based on service rate and poisson model. so, each task is delivered to the hr which is capable to execute it. the simulation results show that the expert cloud can reduce the execution and tardiness time and improve hr utilization. the cost of using resources as an effective factor is also observed.