Computer Science And Technology - Master Dissertation

Permanent URI for this collectionhttps://kr.cup.edu.in/handle/32116/80

Browse

Now showing 1 - 3 of 3

A fault tolerable load balancing technique in cloud computing
(Central University of Punjab, 2014) Shiny; Khurana, Surinder.K
As the IT industry is growing day by day, the need of computing and data storage is increasing rapidly. The process of this increasing mass of data requires more computer equipment to meet the various needs of the organizations. To better capitalize their investment, the over-equipped organizations open their infrastructures to others by exploiting the Internet and other important technologies such as virtualization by creating a new computing model: the cloud computing. Cloud computing is one of the significant milestones in recent times in the history of computers. However, there are number of technical challenges that need to be tackled which include reliability, resource provisioning, fault tolerance, load balancing and efficient mechanism to increase the service level agreement (SLA) and better use of the resources. The main purpose of this dissertation report is to provide a preface of the topic and to work on the various issues involved in the field of load balancing and fault tolerance i.e. load computation and the distribution of load. Load balancing and fault tolerance in cloud computing have a great impact on the performance of the system. Good load balancing makes cloud computing more efficient by provisioning of resources to cloud users on demand basis in pay-as-you-say manner. Load balancers are used for assigning load to different virtual machines in such a way that none of the nodes gets loaded heavily or lightly. When many clients request the server simultaneously, server is overloaded which causes fault. There are many load balancing algorithms and fault tolerance techniques in order to settle down these issues, but these techniques further had some drawbacks. Das and Khilar (2013) discussed a load balancing technique for virtualization and fault tolerance in cloud computing (LBVFT) to assign the task to the virtual nodes depending upon the success rates (SR) and the previous load history. This v technique tolerates not only the faults but also reduce the chance of future faults by not assigning tasks to virtual nodes of physical servers whose success rates are very low and loads are very high. But there is still a need to provide an efficient load balancing, load migration, load calculation and fault handling technique to make the VFT model more effective.
Improved and rule scheme to increase lifetime of wireless sensor networks
(Central University of Punjab, 2014) Saini, Swati; Khurana, Surinder.K
Wireless sensor network (WSN) is a network of small light weight wireless nodes which are highly distributed and deployed in large numbers. The sensor nodes scan the environment and provide useful information to the user through sink node. One of the major issues in wireless sensor networks is developing an energy-efficient routing technique which has a significant impact on the overall lifetime of the sensor network. Our dissertation work focuses on increasing lifetime of wireless sensor networks. This is done by improving the existing Divide and Rule Scheme which is based on dividing network area into logical divisions to reduce distance for intra cluster and inter cluster communications. Improvement of Divide and Rule Scheme is done by developing a new scheme called Segregated Receive and Relay Scheme in which addition of relay nodes is done in inner and middle regions. The Proposed Scheme results in around 50% increase in lifetime of inner region nodes and leading to overall increase in lifetime of the network. Other parameters such as Delay, Throughput and Packet loss also shows better results in new Scheme.
Service level agreement aware energy efficient approach for virtual machine placement in clous data center
(Central University of Punjab, 2014) Kaur, Nirmaljeet; Khurana, Surinder.K
Generally, the growth of computing systems has been concentrate on performance improvement or fast processing of applications from the view point of customer, scientific and business perspective. However, this type of computing increases energy consumption. It is continue to consume large amount of energy in the form of electrical power. In cloud computing data center hosting of cloud applications consume huge amount of energy. Optimization of the energy requirements of data center is a very challenging task. This dissertation report describes the energy efficiency of minimization of migration policy in IaaS environment. In this policy VM migration, energy consumption, SLA violations are calculated. VM migration directly influences resource utilization and Quality of Service (QoS) parameters provided by the system. The effect of disturbance cause server overloads, resource shortage, performance degradation etc. So, one of the challenging task is, the placement of VMs, like where, when, which VM could be migrated for saving energy and less SLA violations, with minimum number of migration. In this research work a new approach has been proposed to select VMs to migrate from the over utilized host. Approach migrates those VMs that belongs to the higher category so that that the resources allocated to these VMs have less cost penalties for SLA violations. Proposed approach results are compared to the simulated Minimization of Migration Policy (Beloglazov et al., 2012) and compared its energy requirements, virtual machine migration and SLA violation with Minimum Migration Time and Random Selection policies.