• The KIPS Transactions:PartC
• /
• v.12C no.1 s.97
• /
• pp.69-76
• /
• 2005

To effectively use limited resources in wireless cellular networks it is necessary to predict exactly the amount of resources required by handoff calls at a future time. In this paper we propose a method which predicts the amount of resources needed by handoff calls more accurately than the existing method based on Wiener processes. The existing method uses the current demands to predict future demands. Although this method is much simpler than using traffic information from neighbor cells, its prediction error increases as time elapses, leading to waste of wireless resources. By using an exponential parameter to decrease the magnitude of error over time, we show in simulation how to outperform the existing method in resource utilization as well as in prediction of resource demands.

• Journal of Advanced Navigation Technology
• /
• v.13 no.3
• /
• pp.372-377
• /
• 2009

This paper evaluates the performance of three methods for predicting resources requested by mobile's calls and a call admission algorithm based on these predicting methods respectively in wireless networks. The first method is based on Wiener prediction model and the second method is based on the MMOSPRED algorithm and the third method is based on the neural network. The proposed call admission algorithm is based on prioritized handoff call in resource allocation. The resources for future handoff calls are therefore predicted and reserved in advance, and then new calls are admitted as long as the remaining resources are sufficient. We compare their performances in terms of prediction error, new call blocking and handoff dropping probabilities by simulation. Results show that the CAC based on Wiener prediction model performs favorably due to exact resources prediction.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.2A
• /
• pp.26-33
• /
• 2005

Exact prediction of resource demands for future calls enhances the efficiency of the limited resource utilization in resource reservation methods for potential calls in wireless IP networks. In this paper, we propose a LMS-Wiener resource(bandwidth) prediction for future handoff calls, and then an the proposed method is compared with an existing Wiener-based method in terms of prediction error through our simulations. In our simulations, we assume that handoff call arrivals follow a non-Poisson process and each handoff call has an non-exponentially distributed channel holdingtime in the cell, considering that handoff call arrival pattern is not Poisson distribution but non-Poisson for long periods of time in wireless picocellular IP networks. Simulation results show that the prediction error in the proposed method converges to the lower value while in an existing method increase as time is passed. Therefore we may conclude that the proposed method improves the efficiency of resource utilization by more exactly predicting resource demands for future handoff calls than an existing method.

• Proceedings of the Korean Society of Fisheries Technology Conference
• /
• 2001.10a
• /
• pp.360-361
• /
• 2001

수산자원의 안정되고 지속적인 이용을 위해서는 철저한 수산자원의 관리가 요구되어지며, 이를 위해서는 무엇보다도 자원의 동향을 정확도 높게 예측할 수 있는 능력이 요구되어진다. 이를 위해서는 현재의 수산자원학적인 연구측면과 함께 수행되어져야 하는 것이 자원 동향의 예측법에 관한 기술 축적이다. (중략)

• Journal of KIISE:Computing Practices and Letters
• /
• v.12 no.5
• /
• pp.275-285
• /
• 2006

This paper proposes an efficient market mechanism-based resource transaction model for grid computing. This model predicts the next resource demand of users and suggests reasonable resource price for both of customers and resource providers. This model increases resource transactions between customers and resource providers and reduces the average of transaction response times from resource providers. For prediction accuracy improvement of resource demands and suggestion of reasonable resource price, this model introduces a statistics-based prediction model and a price decision model of microeconomics. For performance evaluating, this paper measures resource demand prediction accuracy rate of users, response time of resource transaction, the number of resource transactions, and resource utilization. With 87.45% of reliable prediction accuracy, this model works on the less 72.39% of response time than existing resource transaction models in a grid computing environment. The number of transactions and the resource utilization increase up to 162.56% and up to 230%, respectively.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11a
• /
• pp.988-990
• /
• 2005

본 논문에서는 그리드 컴퓨팅 환경에서 사용자와 자원 제공자간의 자원거래 시 사용자의 자원 요구량을 예측하고, 합리적인 가격 결정 알고리즘을 이용하여 기존의 자원 거래 모델에 비해 빠른 응답 속도를 갖는 모델을 제안한다. 본 논문에서 제안하는 모델은 사용자의 자원 요구량를 예측을 위해 통계학의 예측 모델을 적용하였고, 그리드 자원의 거래 가격 결정을 위한 경제학의 이론을 도입하였다. 우리는 실험을 통해 기존의 모델들과 비교하여 그리드 자원 거래를 위한 응답시간을 비교 하였다. 우리는 실험을 통해 기존의 모델들과 비교하여 응답시간이 최소 $72.39\%$ 향상된 결과를 얻었고 우리가 제안한 모델이 기존 모델에 비해 우수하다는 것을 입증하였다.

• Journal of Advanced Navigation Technology
• /
• v.12 no.1
• /
• pp.22-27
• /
• 2008

This paper presents adaptive-MMOSPRED method for prediction of resource demands requested by multimedia calls, and shows the performance of the call admission control based on proposed resource prediction method in multimedia wireless networks. The proposed method determines (I-CDP) random variables of the standard normal distribution by using LMS algorithm that minimize errors of prediction in resource demands, while parameters in an existing method are constant all through the prediction time. Our simulation results show that prediction error in adaptive-MMOSPRED method is much smaller than in fixed-MMOSPRED method. Also we can see via simulation the CAC performance based on the proposed method improves the new call blocking performance compared with the existing method under the desired handoff dropping probability.

• Journal of the Korea Society of Computer and Information
• /
• v.7 no.1
• /
• pp.94-102
• /
• 2002

In this paper, we present a forecasting model for the spectrum demand which will be used for the mid/long-term spectrum forecasting in Korea. In specific, we present the bottom-up model with considering the customer not the previous top-down method. The Proposed model consists of service definition. classification of service characteristics, drawing representative service characteristic , forecasting of service demand, mapping with spectrum resource, verification and spectrum forecasting. The carried actions in each step is described in detail. For the validation of the model an example for the PCS environments is shown. traverse stepping stones for a variety of legitimate reasons.

• Korean Journal of Remote Sensing
• /
• v.23 no.4
• /
• pp.287-295
• /
• 2007

A mineral potential map showing the distributions of potential areas for exploration of undiscovered mineral deposits is a kind of predictive thematic maps. For any predictive thematic maps to show reasonably significant prediction results, validation information on prediction capability should be provided in addition to spatial locations of high potential areas. The objective of this paper is to apply prediction rate curves to the estimation of prediction probability of future discovery. A case study for Au-Ag mineral potential mapping using geochemical data sets is carried out to illustrate procedures for estimating prediction probability and for an interpretation. Through the case study, quantitative information including prediction rates and probability obtained by prediction rate curves was found to be very important for the interpretation of prediction results. It is expected that such quantitative validation information would be effectively used as basic information for cost analysis of exploration and environmental impact assessment.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.10a
• /
• pp.251-258
• /
• 1997

임계 근처에서 반응도 미터로 계단식 반응도 변화를 측정할 때는 중성자원과 감마의 영향 하에서도 정확한 반응도를 결정할 수 있으며, 중성자원과 감마를 측정할 수도 있다. 중성자원과 감마의 영향은 없으나 중성자 분포 함수만 변하는 경우에는 계산으로 예측한 분포 함수의 변화로 측정된 중성자 신호를 보정하여 반응도를 예측할 수 있다. 그러나 중성자원, 감마, 분포 함수가 복합적으로 작용하는 경우에 대하여는 이러한 방법을 적용할 수 없다. 이 매 중성자원과 감마의 영향만 있는 경우에 적용하는 방법을 쓰면 분포 함수의 변화가 측정 결과에 어떤 영향을 미치는지 분석하였다. 그 결과 분포 함수의 변화도 어느 정도 측정이 가능하며, 계산으로 예측하는 분포 함수의 변화로 측정 결과를 단순 보정하여 실제 반응도를 예측할 수 있는 것으로 나타났다.