A Study on Transfer Functions of Binary Particle Swarm Optimization for Energy-Efficient VM Placement

Cloud computing has emerged as the backbone of the IT industry for infrastructural support. In cloud computing, resources are virtualized in form of virtual machines which eventually is mapped to physical infrastructure. Energy-efficient virtual machine placement is an important problem in cloud computing and has attracted the attention of researchers in recent. As virtual machine placement is an NP-hard problem, meta-heuristics have often been applied vastly for this. In one of the authors' earlier works, modified binary particle swarm optimization algorithm has been applied for the VM placement. It was observed therein that the transfer function, which plays an important role to obtain an optima, does not completely avoid the problem of local optima. Therefore, in this work, they have studied the behavior of eight different transfer functions towards this property. For this, energy-efficient VM placement problem is modelled as multi-objective optimization problem and binary particle swarm optimization is applied. The study is done by simulation and their statistical analysis.