Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
9 results
Search Results
Item Improving the efficiency of genetic algorithm approach to virtual machine allocation(Institute of Electrical and Electronics Engineers Inc., 2014) Joseph, C.T.; Chandrasekaran, K.; Cyriac, R.Virtual machine (VM) allocation is the process of allocating virtual machines to suitable hosts. This problem is an NP-Hard problem. It can be considered as a variation of the bin-packing problem. Among various solutions that attempt to solve this problem, several approaches that apply Genetic Algorithm have been proposed. This paper proposes a method to improve the efficiency of such approaches. Implementation of the proposed approach shows significant improvements in the runtime, memory used, energy efficiency and SLA violations. © 2014 IEEE.Item A perspective study of virtual machine migration(Institute of Electrical and Electronics Engineers Inc., 2014) Joseph, C.T.; Chandrasekaran, K.; Cyriac, R.Cloud Computing is one of the leading technologies. As a solution to many of the challenges faced by Cloud providers, virtualization is employed in Cloud. Virtual machine migration is a tool to utilize virtualization well. This paper gives an overview of the different works in literature that consider virtual machine migration. The different works related to virtual migration are classified into different categories. Some of the works that consider less explored areas of virtual machine migration are discussed in detail. © 2014 IEEE.Item Multimedia Streaming Using Cloud-Based P2P Systems(Elsevier, 2015) Thomas, N.; Thomas, M.; Chandrasekaran, K.The past one and a half decades have seen great strides in the field of commercially available distributed computing implementations. The two most popular architectures in the modern world are Peer-to-Peer Systems and Cloud Systems. Peer-to-Peer Systems (P2P) have become very popular in recent times, mainly being used to facilitate file sharing among disparate systems. Another recent trend in modern computing has been the wide scale utilization of Cloud Computing architectures. These systems are used to allow multiple systems to pool their resources and allow other tertiary systems to use these shared resources in bulk for tasks such as data storage, complex calculations, and file sharing. This entails the conceptual outsourcing of various data processing tasks to an external cloud system. Given the two nearly independent functionalities of P2P and Cloud architectures, it is interesting to consider the possibility of fusing these two concepts and researching the applications of the resultant amalgamation. In this research paper, we discuss the theory and application of Cloud Based Peer-to-Peer Systems and their potential application in multimedia streaming services. While the value of P2P Systems and Cloud Computing Systems have been extolled individually, the hybrid of both concepts shows great promise. In this paper we provide an introduction to Cloud Computing Systems, P2P Systems, and the advantages as well as the limitations of both configurations. We then describe the concept of a Cloud-Based P2P System, its basic architecture, and its possible implementations. We also describe the possible application of a Cloud-Based P2P System as a platform for a multimedia streaming service. A proposed algorithm to facilitate streaming in such an application is also described, along with a proposed system model and its advantages. © 2015 The Authors. Published by Elsevier B.V.Item Energy aware SLA and green cloud federations(Institute of Electrical and Electronics Engineers Inc., 2016) Joy, N.; Chandrasekaran, K.; Binu, A.Cloud computing has been an emerging research topic since 2007. The online services are delivered as pay-as-you-use in Cloud computing. The customers need not be in a long term contract with Cloud Service Providers (CSPs). Service level agreements (SLAs) are agreements signed between a cloud service provider and customers. Energy aware SLAs extend the existing SLA agreements in order to include energy and carbon aware parameters. The efficiency and availability of the system are not disturbed when certain jobs are relaxed in order to obtain high energy consumption. In order to increase the energy consumption further we incorporated the concept of cloud federations in which a group of CSPs, mutually agree to make use of their resources to execute the VMs of other CSPs. An algorithm based on the amount of energy consumed by different CSPs is being readily implemented. This algorithm grant a group of CSPs to jointly set the federations in a manner to distribute their work equally so as to reduce the energy consumed. The results from the above approach show that it can be considered as a hopeful solution to the problem of reducing the energy cost even though there are a few challenges during implementation. © 2016 IEEE.Item Determination of task scheduling mechanism using computational intelligence in Cloud Computing(Institute of Electrical and Electronics Engineers Inc., 2016) George, N.; Chandrasekaran, K.; Binu, A.Cloud Computing delivers computational services through the internet. The services availed can vary according to the user requirements. The services are basically provided using virtualization technique. One of the services that are provided is the computational services for the client tasks. The client provides the Service Provider with various sized tasks that need to be executed. The execution of tasks is done using the resources present within the Provider of the services. The service provider checks for available resources, and allocates the jobs to these resources in such a manner as to minimize execution time, and various other factors that affect the performance of the Cloud. The process of allocating resources to the tasks is known as scheduling, and various scheduling mechanisms are present. A single scheduling strategy may not be always optimal in performing scheduling. In this paper, an improved mechanism for choosing the scheduling strategy is explained, which aims at addressing the problems associated with choosing the right scheduling mechanism according to the previously exhibited performances. Experimental results demonstrate the importance of using such a mechanism in selecting the right scheduling strategy. © 2015 IEEE.Item A study on energy efficient cloud computing(Institute of Electrical and Electronics Engineers Inc., 2016) Joy, N.; Chandrasekaran, K.; Binu, A.The rapid increase in demand for computation, increasing amount of data storage needed for running high performance computing enterprises increases the energy and power consumed by large infrastructure. Cloud computing provides a solution to reduce the adverse environmental impacts and saves energy. Our paper draws the attention on the various methods enforced on the cloud environment to make it more energy efficient. In this paper the different energy and power models are being discussed and the main challenges to build a model for green cloud with the help of SLAs are identified. One of the main objectives considered in cloud computing is to provide reliable QoS. This can be defined in the SLA which describes such characteristics. The different ways to minimize energy and power in the cloud computing services are also being discussed in this paper. Our work surveys the various models which will pave the road map for an energy efficient cloud. © 2015 IEEE.Item An objective study on improvement of task scheduling mechanism using computational intelligence in cloud computing(Institute of Electrical and Electronics Engineers Inc., 2016) George, N.; Chandrasekaran, K.; Binu, A.Cloud Computing facilitates delivery of various types of computational services through the internet. These services can be availed according to the user demand. The resource scarcity problems within the Service Providers are met using Virtualization technique, which allows scalability of resources and thereby helps to meet the client requirements. Allocation of resources to client tasks is an issue that is being addressed for a long time. Due to the increased complexity in the area, there has not yet been a perfect scheduling mechanism. Practices have been done profusely in order to find solutions for scheduling that nears optimality. A single scheduling mechanism may not always give the expected outcome. The task scheduling mechanisms are designed in a manner as to optimize some metrics related to the Cloud. This paper overviews various literature associated with task scheduling and resource scheduling in Cloud Computing. An examination of the techniques is done and a proposal is made, which will allow to further improve the scheduling mechanism. © 2015 IEEE.Item A bio-inspired model to provide data security in cloud storage(Institute of Electrical and Electronics Engineers Inc., 2017) Hitaswi, N.; Chandrasekaran, K.The demand for cloud computing is increasing rapidly because of the advantages it provides to the customers like, pay as you use, self-serving, elastic, sharing of resources, ease of use, and accessibility. Due to the increase in the usage of the technology, there exists a high chance of compromising the security of the data being stored on the cloud. The major hindrance in the usage of the technology is the security concerns which accompany it. This increases the demand for a robust security mechanism to protect the data on the cloud. So as to overcome this drawback of cloud computing, encrypting the data to be stored on the cloud is one of the solutions. As part of this paper, a security mechanism to improve the security of data in cloud storage is suggested. The security mechanism used is inspired by the bio-inspired genetic algorithm. The inspiration behind the proposed security model is an amalgamation of genetic algorithm and attribute based encryption. As per the methodology proposed the data need to be encrypted before being stored on the cloud. This way the cloud service provider is unaware of the data being stored and even if the data is compromised to some third party, there is no information leakage. © 2016 IEEE.Item Fuzzy Reinforcement Learning based Microservice Allocation in Cloud Computing Environments(Institute of Electrical and Electronics Engineers Inc., 2019) Joseph, C.T.; Martin, J.P.; Chandrasekaran, K.; Kandasamy, A.Nowadays the Cloud Computing paradigm has become the defacto platform for deploying and managing user applications. Monolithic Cloud applications pose several challenges in terms of scalability and flexibility. Hence, Cloud applications are designed as microservices. Application scheduling and energy efficiency are key concerns in Cloud computing research. Allocating the microservice containers to the hosts in the datacenter is an NP-hard problem. There is a need for efficient allocation strategies to determine the placement of the microservice containers in Cloud datacenters to minimize Service Level Agreement violations and energy consumption. In this paper, we design a Reinforcement Learning-based Microservice Allocation (RL-MA) approach. The approach is implemented in the ContainerCloudSim simulator. The evaluation is conducted using the real-world Google cluster trace. Results indicate that the proposed method reduces both the SLA violation and energy consumption when compared to the existing policies. © 2019 IEEE.