Efficient resource allocation through Virtual machine placement in a cloud datacenter is an ever-growing demand. Different Virtual Machine optimization techniques are constructed for different optimization problems. Particle Swam Optimization (PSO) Algorithm is one of the optimization techniques to solve the multidimensional virtual machine placement problem. In the algorithm being proposed we use the combination of Modified First Fit Decreasing Algorithm (MFFD) with Particle Swarm Optimization Algorithm, used to solve the best Virtual Machine packing in active Physical Machines to reduce energy consumption; we first screen all Physical Machines for possible accommodation in each Physical Machine and then the Modified Particle Swam Optimization (MPSO) Algorithm is used to get the best fit solution.. In our paper, we discuss how to improve the efficiency of Particle Swarm Intelligence by adapting the efficient mechanism being proposed. The obtained result shows that the proposed algorithm provides an optimized solution compared to the existing algorithms.

No abstract

The new developments in the field of information technology offered the people enjoyment, comforts and convenience. Cloud computing is one of the latest developments in the IT industry also known as on-demand computing. It provides the full scalability, reliability, high performance and relatively low cost feasible solution as compared to dedicated infrastructures. It is the application provided in the form of service over the internet and system hardware in the data centers that gives these services. This technology has the capacity to admittance a common collection of resources on request. It is proving extremely striking to cash-strapped IT departments that are wanted to deliver better services under pressure. When this cloud is made available for the general customer on pay per use basis, then it is called public cloud. When customer develops their own applications and run their own internal infrastructure then is called private cloud. Integration and consolidation of public and private cloud is called hybrid cloud.

The piller of cloud computing is the virtualization technology. Virtualization is a technique to share a physical machines resources into many virtual machines. With the growth in the usage of VMs, VM placement and optimization is the need of the hour. It reduces the operating cost and increases physical machine utilization in a data center. The prime requirement is to reduce the number of active physical machines and increase the number of virtual machines per physical machine without compromising the data quality and data security. This significantly reduces power consumption and cloud operation costs. This paper provides a comprehensive survey on the algorithms proposed for optimal VM placement and selection in a cloud environment. The algorithms include Ant colony optimizition(ACO), Particle Swarm Optimization(PSO), Genetic Algorithm(GA), Flower Pollination Algorithm(FPA), Fruit Fly Algorithm(FOA), Honey Bee Algorithm9BA). These are a class of nature inspired algorithms which are found useful in selecting and allocating VM resources in a cloud computing environment which requires highly efficient and scalable solutions for dynamically allocating computing resources in the cloud

To meet the ever-growing demand for computational resources, it is mandatory to have the best resource allocation algorithm. In this paper, Particle Swarm Optimization (PSO) algorithm is used to address the resource optimization problem. Particle Swarm Optimization is suitable for continuous data optimization, to use in discrete data as in the case of Virtual Machine placement we need to fine-tune some of the parameters in Particle Swarm Optimization. The Virtual Machine placement problem is addressed by our proposed model called Improved Particle Swarm Optimization (IM-PSO), where the main aim is to maximize the utilization of resources in the cloud datacenter. The obtained results show that the proposed algorithm provides an optimized solution when compared to the existing algorithms.