• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.10
• /
• pp.1712-1732
• /
• 2011

Cloud provides dynamically scalable virtualized computing resources as a service over the Internet. To achieve higher resource utilization over virtualization technology, an optimized strategy that deploys virtual machines on physical machines is needed. That is, the total number of active physical host nodes should be dynamically changed to correspond to their resource usage rate, thereby maintaining optimum utilization of physical machines. In this paper, we propose a pattern-based prediction model for resource provisioning which facilitates best possible resource preparation by analyzing the resource utilization and deriving resource usage patterns. The focus of our work is on predicting future resource requests by optimized dynamic resource management strategy that is applied to a virtualized data center in a Cloud computing environment. To this end, we build a prediction model that is based on user request patterns and make a prediction of system behavior for the near future. As a result, this model can save time for predicting the needed resource amount and reduce the possibility of resource overuse. In addition, we studied the performance of our proposed model comparing with conventional resource provisioning models under various Cloud execution conditions. The experimental results showed that our pattern-based prediction model gives significant benefits over conventional models.

• KDI Journal of Economic Policy
• /
• v.28 no.2
• /
• pp.97-129
• /
• 2006

This study investigates whether dynamic loan loss provisioning is necessary in Korean banking environments. Under the dynamic provisioning rule banks are required to accumulate additional reserves to general and specific provisionings in preparation for expected loan losses until maturity. This provisioning is most effective in the case that banks tend to recognize less loan losses in the business upturns and/or in the periods of increasing profits. The empirical study, however, shows that banks support procyclicality of loan loss privisioning and earning smoothing behavior over profit fluctuations. These findings suggest that Korea would not seriously need the introduction of dynamic loan loss provisioning. But this policy implication does not seem robust in view that the recent experience shows the countercyclicality of loan loss provisioning practices and negative correlation between earnings and provisioning after financial restructuring was completed. This result is partly attributable to vigorous shareholder activism because of high foreign ownership of most commercial banks. Once it is true that bank management is more interested in short-term performances, current loan loss provisioning would have attributes of impairing capital adequacy, hence strengthening loan loss provisiong requirements.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.884-888
• /
• 2009

Vehicular Adhoc Networks (VANET), a Vehicle-to-Infrastucture or Vehicle-to-Vehicle communication technology, is an area that makes more specific use of Mobile Adhoc Networks(MANET). VANET's Quality of Service(QoS) focuses on preventing possible emergencies like car crash from happening by immediately transmitting information to the cars around, while MANET's QoS is being studied for the quality of multimedia data such as Video on Demand(VoD), Video streaming, Voice over IP(VoIP), etc. In this paper, I structure the actual network configuration using Link State Routing(LSR), implement QoS Provisioning Model using Common Open Policy Service(COPS), and suggest more effective k-hop Cluster and inter-domain policy negotiation which fit better to the characteristics of VANET.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.11S
• /
• pp.3633-3650
• /
• 2000

Internet is no longer a network for special communities but became a global means of communication infrastructure for everyday life. People are exchanging their personal messages using e-mails, students are getting their educational aids through the web, people are buying a variety of goods from cyber shopping malls, and companies are conducting their businesses over the Internet. Recently, such an explosive growth of the traffic in the Internet raised a big concern on how to accommodate ever-changing user's needs in terms of an amount of the traffic, characteristics of the traffic, and various service quality requirements, Over provisioning can be a simple solution but it is too expensive and inefficient. Thus many new technologies to solve this very difficult puzzle have bcen introduced recently, Any single solution, however, can be insufficient and a carefully designed architecture, which integrates a group of solutions, is required. In this paper, we propose a policy-based Internet QoS provisioning, traffic engineering and perfonnance management system as our solution to this problem. Our integrated management QoS solution can provide highly responsive flow-through service provisioning, more realistic service and resource policy control based on the real network performance information, and centralized control of traffic engineering for heterogeneous networks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.11B
• /
• pp.707-721
• /
• 2005

Recently, there has been a dramatic increase in data traffic, driven primarily by the explosive growth of the Internet. Optical networking is believed as a key solution to keep up with the growth, thus, the most pressing issue is how to manage and control large optical networks. Currently, provisioning end-to-end connections across the transport network has involved the network operator, leading to long provisioning times in an era when customers are demanding shorter provisioning time. To address this critical issue, new control intelligence is being studied for use within optical networks to shorten provisioning time. Both the IETF and the ITU-T have been aggressively defining many aspects of a control plane for the next generation convergence transport network. Basically, they are based on the distributed control scheme. In this Paper, we suey the applicability of the centralized control scheme for the provisioning control of optical transport network to utilize its inherent advantages over the distributed control scheme. We discuss new central control architecture, and control procedure. Also, we examine the applicability of the existing IETF routing and signaling protocols to the new control concepts, and then, we propose the additional routing and signaling information elements.

• KIPS Transactions on Computer and Communication Systems
• /
• v.4 no.3
• /
• pp.81-84
• /
• 2015

Cloud service is one of major technologies in modern IT business. Due to the dynamics of user demands, service providers need VM(Virtual Machine) provisioning mechanism to predict the amount of resources demanded by cloud users for the next service and to prepare the resources. VM provisioning provides the QoS to cloud user and maximize the revenue of a service provider by minimizing the expense. In this paper, we propose a new VM provisioning technique to minimize the total expense of a service provider by minimizing the expected value of the expense based on the predicted demands of users. To evaluate the effectiveness of our prediction technique, we compare the total expense of our technique with these of the other prediction techniques with a series of real trace data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.12
• /
• pp.5286-5306
• /
• 2016

Emerging applications in automation, medical imaging, traffic monitoring and surveillance need real-time data transmission over Wireless Sensor Networks (WSNs). Guaranteeing Quality of Service (QoS) for real-time traffic over WSNs creates new challenges. Rapid penetration of smart devices, standardization of Machine Type Communications (MTC) in next generation 5G wireless networks have added new dimensions in these challenges. In order to satisfy such precise QoS constraints, in this paper, we propose a new cross-layer QoS-provisioning strategy in Wireless Multimedia Sensor Networks (WMSNs). The network layer performs statistical estimation of sensory QoS parameters. Identifying QoS-routing problem with multiple objectives as NP-complete, it discovers near-optimal QoS-routes by using evolutionary genetic algorithms. Subsequently, the Medium Access Control (MAC) layer classifies the packets, automatically adapts the contention window, based on QoS requirements and transmits the data by using routing information obtained by the network layer. Performance analysis is carried out to get an estimate of the overall system. Through the simulation results, it is manifested that the proposed strategy is able to achieve better throughput and significant lower delay, at the expense of negligible energy consumption, in comparison to existing WMSN QoS protocols.

• International journal of advanced smart convergence
• /
• v.1 no.2
• /
• pp.12-15
• /
• 2012

Energy costs for operating and cooling computing resources in Cloud infrastructure have increased significantly up to the point where they would surpass the hardware purchasing costs. Thus, reducing the energy consumption can save a significant amount of management cost. One of major approach is removing hardware over-provisioning. In this paper, we propose a technique that facilitates power saving through reducing resource over provisioning based on virtualization technology. To this end, we use dynamic workload shaping to reschedule and redistribute job requests considering overall power consumption. In this paper, we present our approach to shape workloads dynamically and distribute them on virtual machines and physical machines through virtualization technology. We generated synthetic workload data and evaluated it in simulating and real implementation. Our simulated results demonstrate our approach outperforms to when not using no workload shaping methodology.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 2006.07a
• /
• pp.744-744
• /
• 2006

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.2B
• /
• pp.105-116
• /
• 2003

The differentiated service architecture is based on a simple model by applying a per-class service in the core node of the network. However, due to the simplified network behavior, the network structure and provisioning were more complicated. If a service provider wants the dynamic provisioning or better bandwidth guarantee, signaling protocol with QoS parameters or admission control method should be deployed in DiffServ network. However, these methods increase the complexity. Therefore, we proposed DPM2 mechanism for admission control in the DiffServ network. In this paper, we describe and survey the admission control methods that are applicable to IP networks and propose also the dynamic provisioning mechanism based on the bandwidth broker and distributed measurement based admission control and movable boundary bandwidth management to support heterogeneous QoS services in differentiated service networks. For the performance evaluation for proposed mechanism, we used ns-2 simulator.