• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.6
• /
• pp.1016-1026
• /
• 1994

In this paper, we have studied load balancing problems in distributed systems. The nodes of distributed systems exchange periodically system state information each other. The information is stored in history. Based on the information, we compute an expected load index(ELI) using a five-degree interpolation polynomial in Newton`s backward difference interpolation formula. A new location policy of dynamic load balancing systems makes use of the ELI. We show that its performance is better than that of the existing load balancing algorithm through a simulation study.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04a
• /
• pp.381-384
• /
• 2002

This paper describes the motivation, design and performance of bi-active Load balancer in Linux Virtual Server. The goal of bi-active Load balancer is to provide a framework to build highly scalable, fault-tolerant services using a large cluster of commodity servers. The TCP/IP stack of Linux Kernel is extended to support three IP load balancing techniques, which can make parallel services of different kinds of server clusters to appear as a service on a single IP address. Scalability is achieved by transparently adding or removing a node in the cluster. and high availability is provided by detecting node or daemon failures and reconfiguring the system appropriately. Extensive simulation reveals that the proposed approach improves the reply rate about 20% compared to earlier design.

• Proceedings of the IEEK Conference
• /
• 2000.06c
• /
• pp.105-108
• /
• 2000

As Internet and WWW grow rapidly, the role of web servers is getting more important, and the number of users in a few popular site is also explosively increasing. Load balancing in clustered web server systems is important task to utilize whole system effectively, so dynamic load balancing is required to overcome the limit of static load balancing. In this paper, we propose two dynamic load balancing schemes, and analyzed load model and Performance improvement and also compare existing load balancing methods and IP filtering method. In case of load balancing with threshold, little extra traffic was required for better performance, but in case of load balancing with load weight, we found that the performance mainly depends on information exchange rate.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.6
• /
• pp.67-73
• /
• 2012

A distributed conference architecture with a new load control method has been suggested in this paper. A new event package in this paper enables to control conference load. Some additional elements for exchanging SIP messages between server and participants, and for distributing the load, have been added to new conference information data format. Furthermore to lessen the load, all conference servers share the processing of conference information data which should be transferred periodically to all participants. The suggested load control event package makes each server can get current load status of the overall servers. When load increases in one server SIP client requests are distributed by selecting a server which has the lowest load value, or new server is created to share the load. The performance of the proposed system has been evaluated by experiments. They shows 21.6% increase in average delay time, and 29.2% increase in average SIP message processing time.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.151-153
• /
• 2000

본 논문은 분산 서버 시스템을 위한 Load Balancing Architecture를 제안한다. 이 구조는 기존에 구현된 클라이언트/서버 시스템을 그대로 사용하면서 규모를 확장할 수 있는 특징을 가지고 있다. Load Balancing Service가 동작하기 위해서는 서버의 성능과 구성상태에 대한 정보를 주고받는 것이 필요하다. 이를 위하여 본 논문에서는 Load Balacing Information Transfer Protocol을 제시한다. 본 논문이 제시하는 구조는 DCOM 및 EJB 등의 Component 서버에서도 사용되어서 확장성을 가지게 할 수 있다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.05a
• /
• pp.908-910
• /
• 2013

Customers of different electricity consumer types have different daily load shapes in the manner of different characteristics. Therefore, maximally capture load shape variability are desirable in load flow analysis. And most of time, such load shape variability can be found in the particular subspace of load diagrams. Therefore, in this paper, we are using subspace projection method to capture the emerging subspaces of load diagrams which maximize the difference between particular load shapes in different group of customers. As the result, subspace projection method can be used in load profiling and the performance is good as traditional approaches.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.5
• /
• pp.848-853
• /
• 2007

This paper presents a methodology of estimating incomplete information in electricity markets for analyzing the gaming behavior of Generating Companies (GENCOs). Each GENCO needs to model its opponents' unknown information of strategic biddings and cost functions. In electricity markets with complete information, each GENCO knows its rivals' payoff functions and tries to maximize its own profit at Nash equilibriurnl Nli) by acknowledging the rivals' cost function. On the other hand, in the incomplete information markets, each GENCO lacks information about its rivals. Load patterns can change continuously due to many factors such as weather, price, contingency, etc. In this paper, we propose the method of the estimation of the opponents' cost function using market price, transaction quantities. and customer load patterns. A numerical example with two GENCOs is illustrated to show the basic idea and effectiveness of the proposed methodology.

• Journal of Digital Convergence
• /
• v.10 no.11
• /
• pp.323-329
• /
• 2012

This study examines the anchoring effect and the adjustment process of two variables(verbal information, cognition load) with snack products. In the results of this study, 1) we found the anchoring effect because the respondents predicted more the number of real units(goraebap) on the packaging painted 25 units than 5 ones. 2) We confirmed the moderating effect of verbal information. The difference of the number of real units predicted between the two packaging was decreased when the visual information was in company with verbal information. And 3) the moderating effect of cognitive load appeared because the more cognitive load was, the less the difference of the number of real units predicted was. This study has shown that we can reduce the errors and biases by adjusting the information frame or the cognitive load. This research provides a theoretical-practical implications to the marketing staffs like packaging designers as well as scholars to study consumer psychology.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.1
• /
• pp.117-135
• /
• 2016

Performance enhancement is one of the most important issues in high performance distributed computing systems. In such computing systems, online users submit their jobs anytime and anywhere to a set of dynamic resources. Jobs arrival and processes execution times are stochastic. The performance of a distributed computing system can be improved by using an effective load balancing strategy to redistribute the user tasks among computing resources for efficient utilization. This paper presents a multi-class load balancing strategy that balances different classes of user tasks on multiple heterogeneous computing nodes to minimize the per-class mean response time. For a wide range of system parameters, the performance of the proposed multi-class load balancing strategy is compared with that of the random distribution load balancing, and uniform distribution load balancing strategies using simulation. The results show that, the proposed strategy outperforms the other two studied strategies in terms of average task response time, and average computing nodes utilization.

• Journal of information and communication convergence engineering
• /
• v.20 no.4
• /
• pp.295-302
• /
• 2022

This paper contains the development of a smart power device designed to collect load power data from industrial manufacturing machines, predict future variations in load power data, and detect abnormal data in advance by applying a machine learning-based prediction algorithm. The proposed load power data prediction model is implemented using a Long Short-Term Memory (LSTM) algorithm with high accuracy and relatively low complexity. The Flask and REST API are used to provide prediction results to users in a graphical interface. In addition, we present the results of experiments conducted to evaluate the performance of the proposed approach, which show that our model exhibited the highest accuracy compared with Multilayer Perceptron (MLP), Random Forest (RF), and Support Vector Machine (SVM) models. Moreover, we expect our method's accuracy could be improved by further optimizing the hyperparameter values and training the model for a longer period of time using a larger amount of data.