• 제목/요약/키워드: randomizing

검색결과 32건 처리시간 0.02초

일라이어스와 페레즈의 방식에 기반한 하이브리드 무작위화 함수 (A Hybrid Randomizing Function Based on Elias and Peres Method)

  • 배성일;김민수
    • 한국컴퓨터정보학회논문지
    • /
    • 제17권12호
    • /
    • pp.149-158
    • /
    • 2012
  • 본 논문에서는 점근적으로 최적인 두가지의 무작위화 함수인 일라이어스(Elias) 함수와 페레즈(Peres) 함수의 장단점을 고려한 하이브리드 무작위화 함수를 제안한다. 무작위화 함수는 편향성이 있는 무작위수의 공급원으로부터 균등한 무작위수를 생성하는데 쓰이는 알고리즘을 수학적으로 추상화한 것이다. 일라이어스 함수와 페레즈 함수는 입력의 길이가 무한으로 증가함에 따라 그 출력효율성이 정보론적 한계치에 다가간다. 특히, 일라이어스 함수는 주어진 (유한의) 입력길이에 대해 최적인 무작위화 함수이다. 그러나 그 계산은 간단하지 않고, 주어진 입력길이에 의존한다. 반면, 페레즈 함수는 정해진 입력의 길이에 대해 출력효율이 최적이지는 않으나, 점근적으로는 최적이고, 간단한 재귀식에 의해 정의되어서 그 계산이 매우 간단하고 적은 메모리를 필요로 한다. 이러한 계산복잡도와 출력효율에 대한 두가지 무작위화 함수의 장단점에 주목하여, 각각의 장점을 고려한 하이브리드 무작위화 함수를 제안하고 이를 분석한다.

개선된 Randomizing 알고리즘을 이용한 Job Shop 일정계획에 관한 연구 (A Study on the Job Shop Scheduling Using Improved Randomizing Algorithm)

  • 이화기;김민석;이승우
    • 대한안전경영과학회지
    • /
    • 제6권2호
    • /
    • pp.141-154
    • /
    • 2004
  • The objective of this paper is to develop the efficient heuristic method for solving the minimum makespan problem of the job shop scheduling. The proposed heuristic method is based on a constraint satisfaction problem technique and a improved randomizing search algorithm. In this paper, ILOG programming libraries are used to embody the job shop model, and a constraint satisfaction problem technique is developed for this model to generate the initial solution. Then, a improved randomizing search algorithm is employed to overcome the increased search time of constrained satisfaction problem technique on the increased problem size and to find a improved solution. Computational experiments on well known MT and LA problem instances show that this approach yields better results than the other procedures.

A Suggestion for Randomized Response Technique using Fuzzy Logic

  • Choi, Kyung-Ho
    • Communications for Statistical Applications and Methods
    • /
    • 제8권2호
    • /
    • pp.465-471
    • /
    • 2001
  • Randomized response technique is a survey technique for eliminating evasive answer bias. But this technique has a problem. This procedure suffers from linguistic expression in randomizing device. Thus for solving the problem, in this paper, we suggested the randomized response technique using fuzzy logic.

  • PDF

무선암호시스템에서 FDI와 GDI를 이용한 전송성능 분석 (Analysis of Transmission Using Frame Domain Interleaving and Group Domain Interleaving on a Radio Encryption System)

  • 홍진근
    • 한국통신학회논문지
    • /
    • 제25권10B호
    • /
    • pp.1744-1759
    • /
    • 2000
  • In this paper a synchronized stream cryptosystem for secure link layer communication on a radio channel is designed. We have proposed new interleaving schemes to randomize a burst error and experimented with different types of interleaving schemes. The proposed techniques of interleaving schemes are : (1) interleaving scheme based on frame(2) interleaving scheme based on group. The proposed schemes are very robust in randomizing

  • PDF

성요한초제제 임상시험 고찰을 통한 한약기원 항우울제 임상시험 방법론 연구 (A Study on the Methodology for a Clinical Trial of an Antidepressive Korean Herb by Comparison with St. John's Wort Trials)

  • 정희;이지형;장보형;고성규
    • 대한한방내과학회지
    • /
    • 제28권2호
    • /
    • pp.363-376
    • /
    • 2007
  • Objectives : This study was undertaken to learn what should be considered in a good clinical trial investigating a herbal medicine as an antidepressant. Methods : Five well-designed clinical trials published from 2000 to 2006 investigating SJWE in depressive disorder were selected. The trials were reviewed and compared in terms of methodology such as trial design, patient selection, efficacy & safety evaluation, and so on. On the basis of this review of the trials and the regulations and guidelines of KFDA, we suggest some points to be considered for a good clinical trial of a herb for depression. Results : Although every trial had its own unique design, procedure, objectives and so on, all trials used randomizing and double blinding methods. If there is no ethical problem, a placebo-controlled design should be considered in a herbal antidepressant clinical trial for depression. Conclusions : Some points to be considered in an optimal & good clinical trial for an antidepressive herbal medicine were suggested as follows: 1) randomizing and double blinding manner is essential, 2) if there is no ethical problem, placebo control design should be considered, 3) the trial period should be 6 weeks, 4) out-patients will be recruited as subjects, 5) investigators will be well-trained psychiatrists or medical doctors, 6) the number of subjects should be calculated by statistical methods, 7) subjects should be diagnosed by DSM-IV criteria, 8) subjects who have current risk of committing suicide should be excluded, etc.

  • PDF

전극 함몰형 고효율 실리콘 태양전지에서의 texturing 효과 (Texturing Effects on High Efficiency Silicon Buried Contact Solar Cell)

  • 지일환;조영현;이수홍
    • 한국전기전자재료학회:학술대회논문집
    • /
    • 한국전기전자재료학회 1995년도 춘계학술대회 논문집
    • /
    • pp.172-176
    • /
    • 1995
  • Schemes to trap weakly absorbed light into the cell have played an important role in improving the efficiency of both amorphous and crystlline silicon solar cells. One class of scheme relies on randomizing the direction of light within the cell by use of Lambertian(diffuse)surfaces. A second class of scheme relies on the use fo well defined geometrical features to control the direction of light wihin the cell, Widly used geometrical features in crystalline silicon solar cells are the square based pyramids and V-shaped grooves formed in (100) orientated surfaces by intersecting(III) crystallographic planes exposed by anisotropic etching. 18.5% conversion efficiency of Buried Contact Solar Cell with pyramidally textured surface has been achieved. 18.5% efficiency of silicon solar cell is one the highest record in the world The efficieny of cell without textured surface was 16.6%, When adapting textured surface to the Cell, the efficiency has been improved over 12%.

  • PDF

DCT 계수의 통계적 분석을 통한 최적의 워터마크 계수 추출 (Optimal Watermark Coefficient Extraction by Statistical Analysis of DCT Coefficients)

  • 최병철;김용철
    • 대한전자공학회:학술대회논문집
    • /
    • 대한전자공학회 2000년도 추계종합학술대회 논문집(3)
    • /
    • pp.69-72
    • /
    • 2000
  • In this paper, a novel algorithm for digital watermarking is proposed. We use two pattern keys from BCH (15, 7) code and one randomizing key. In the embedding process, optimal watermark coefficients are determined by statistical analysis of the DCT coefficients from the standpoint of HVS. In the detection, watermark coefficients are restored by correlation matching of the possible pattern keys and minimizing the estimation errors. Attacks tested in the experiments ate image enhancement and image compression (JPEG). Performance is evaluated by BER of the logo images and SNR/PSNR of the restored images. Our method has higher performance against JPEG attacks. Analysis for the performance is included.

  • PDF

Simulation Study on the Minimization of Moiré Patterns Caused by Microlens Array Films for Backlight Applications

  • Joo, Byung-Yun;Ko, Jae-Hyeon
    • Journal of the Optical Society of Korea
    • /
    • 제18권5호
    • /
    • pp.538-545
    • /
    • 2014
  • Microlens array (MLA) films have been used as brightness-enhancement components in many illumination systems, such as a backlight, which should meet the industrial needs of low cost and high brightness performance. A hexagonally-arranged MLA (HA-MLA) was chosen due to its high filling factor, which was a prerequisite for achieving high brightness-enhancement capability. The moir$\acute{e}$ fringes produced in a film stack composed of two HA-MLA were investigated, under four different lens configurations, by optical simulation. It was found that randomizing the lens configuration, while keeping the filling factor high, was effective way to develop high-gain MLA films for backlight applications.

Performance Analysis of Coordinated Random Beamforming Technique in Multi-cell Environments

  • Lee, Jong-Min;Jung, Bang-Chul
    • Journal of information and communication convergence engineering
    • /
    • 제8권4호
    • /
    • pp.393-398
    • /
    • 2010
  • For multi-cell environments, coordinated random beamforming technique in multiuser MIMO(multiple-input multiple-output) broadcast channel is considered. In order to mitigate severe interference at receivers, the multi-cell environments might require complex transmitter and receiver design because the scheduler decision based on full channel state information (CSI) in one cell must be intertwined with decision made by other cells' CSI. With limited CSI, however, this paper considers a scheme of randomizing transmitters' beamforming but being coordinated with other cell transmitters. The transmitters in each cell share random beamforming patterns and schedule data transmission within coherent scheduling period. The corandomized beams allow the users to be selected with the highest SINRs even in multi-cell environments. We analyze the performance of the proposed scheme. And numerical results show that the scheme achieves better performance than the conventional random beamforming when applying to multi-cell environments.