• 디지털산업정보학회논문지
• /
• 제10권3호
• /
• pp.119-125
• /
• 2014

Echo cancellation is required for a dual-duplex high speed communication such as digital subscriber line(DSL), in order to allow each individual loop to operate in a full duplex fashion. Echo cancellation was one of the most difficult aspects of DSL design, requiring high linearity and total echo return loss in excess of 70 dB. For a long and rapidly changing echo response, if the echo is cancelled by an adaptive echo canceller, the echo canceller needs more taps and its performance is decreased. But if the response is divided into several responses, which response is estimated by a adaptive digital filter and combined, the computation complexities are decreased and the performance is increased. Therefore, the adaptive multi-stage echo canceller is proposed to decrease the computation complexity and increase the performance of echo return loss, in which the echo canceller is composed of several stage echo canceller estimating each divided echo response. Through computer simulations, this multi-stage echo canceller is verified to have merits for high speed communications such as DSL application.

• 한국통신학회논문지
• /
• 제18권11호
• /
• pp.1714-1725
• /
• 1993

본 논문에서는 스칼라 인수 a1, a2, a3를 매개변수화하여 갱신항을 첨가한 비선형 적응 알고리즘의 특성을 해석하고 그 구조를 나타낸다. 수렴 특성의 해석에서 평균 필터계수 벡터에 대하여 전이행열의 값이 기술된다. 그 알고리즘이 안정하기 위한 범위도 증명된다. 또한 본 알고리즘의 시정수도 유도되고, Sign 알고리즘, 기존의 LMS 알고리즘, LFG 알고리즘, QFG 알고리즘의 계산량도 비교해 본다. 평균자승의 수렴특성을 해석하고 평균자승 순환식과 초과 평균자승 오차(excess mean square error) 표현식을 유도하고 본 알고리즘이 안정하기 위한 조선도 정한다. 컴퓨터 모의실험(simulation)에서 CFG 알고리즘이 LMS, LFG 및 QFG 알고리즘보다 계산량이 증가하는 반면 수렴속도에서 현저한 향상을 보여준다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권6호
• /
• pp.1512-1532
• /
• 2013

A convertible undeniable signature requires a verifier to interact with the signer to verify a signature and furthermore allows the signer to convert a valid one to publicly verifiable signature. In 2007, Yuen et al. proposed a convertible undeniable signature without random oracles in pairings. However, it is recently shown that Yuen et al.'s scheme is not invisible for the standard definition of invisibility. In this paper, we propose a new improvement by using extended Euclidean algorithm that can overcome the visibility attack. The proposed scheme has been evaluated based on computation and communication complexities and the performance comparisons of Yuen et al.'s scheme and various convertible undeniable signature schemes are provided. Moreover, it has been observed that the proposed algorithm reduces the computation and communication times significantly.

• 한국멀티미디어학회논문지
• /
• 제7권12호
• /
• pp.1708-1718
• /
• 2004

In this paper, we are concerned with a digital fingerprinting scheme for multi-purchase where a buyer wants to buy more than a digital content. If we apply previous schemes to multi-purchase protocol, the number of execution of registration step and decryption key should be increased in proportion to that of digital contents to be purchased in order to keep unlinkability. More worse, most of fingerprinting schemes in the literature are based on either secure multi-party computation or general zero-knowledge proofs with very high computational complexity. These high complexities complicate materialization of fingerprinting protocol more and more. In this paper, we propose a multi-purchase fingerprinting scheme with lower computational complexity. In the proposed scheme, a buyer executes just one-time registration step regardless of the number of contents to be purchased. The number of decryption key is constant and independent of the number of contents to be purchased. We can also reduce the computational costs of buyers by introducing a concept of proxy-based fingerprinting protocol.

• Journal of applied mathematics & informatics
• /
• 제20권1_2호
• /
• pp.637-645
• /
• 2006

An adaptive algorithm for reducing the hardware complexity is presented. This paper proposes a simplified LMS algorithm for the adaptive system and analyzes its convergence characteristics mathematically. An objective of the proposed algorithm is to reduce the hardware complexity. In order to test the performances, it is applied to the echo canceller, and a program is described. The results from simulations show that the echo canceller adopting the proposed algorithm achieves almost the same performances as one adopting the NLMS algorithm. If an echo canceller is implemented with this algorithm, its computation quantities are reduced to the half as many as the one that is implemented with the LMS algorithm, without so much degradation of performances.

• 한국통신학회논문지
• /
• 제21권7호
• /
• pp.1668-1676
• /
• 1996

This paper presents the evaluation of an angular spectrum-based method used to calculate scattering pattern of a three-dimensional object modelled as a collection composed of vertical sectional two-dimensional images. This is done via comparing a proposed method with two existing methods, i.e., a Fresnel hologram method and a ray-tracing method, in terms of computatioal complexities and reconstructed results. Maathematical derivations for each methods are reviewed and implementing procedures are described in detail, along with the amount of computaions required from the implementation point of view, rather than from the mathematical point of view. We show simulation results in which the Fresnel holoram method dose not exhibit promising results although it requires the least computation. Moreover, it is also shown that the proposed method, even with much less computational requirement than the ray-tracing method, produces good performances asmuch as the ray-tracing method does.

• 한국항공운항학회지
• /
• 제23권1호
• /
• pp.56-61
• /
• 2015

Because of kinematic complexities, nonlinear behavior, etc, the performance of oleo-pneumatic landing gear is predicted by qualified commercial softwares. While commercial softwares predict more exactly, it takes a long time to construct or modify a model. At initial design stage, design parameters can be determined quickly and exactly enough with simple 2 degree of freedom model of mass, spring and damping. 2 degree of freedom model can be easily applied to optimization and reliability analysis which takes repetitive computation. In this paper, oleo-pneumatic landing gear is modeled as a nonlinear 2 degree of freedom model. The analysis are compared with landing gear drop test. To determine design parameter, optimization problem is solved with genetic algorithm and 2 degree of freedom model.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제54권6호
• /
• pp.359-365
• /
• 2005

This paper presents an reduced-order near-optimal controller for the congestion control of Available Bit Rate (ABR) service in Asynchronous Transfer Mode (ATM) networks. We introduce the model, of a class of ABR traffic, that can be controlled using a Explicit Rate feedback for congestion control in ATM networks. Since there are great computational complexities in the class of optimal control problem for the ABR model, the near-optimal controller via reduced-order technique is applied to this model. It is implemented by the help of weakly coupling and singular perturbation theory, and we use bilinear transformation because of its computational convenience. Since the bilinear transformation can convert discrete Riccati equation into continuous Riccati equation, the design problems of optimal congestion control can be reduced. Using weakly coupling and singular perturbation theory, the computation time of Riccati equations can be saved, moreover the real-time congestion control for ATM networks can be possible.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2005년도 춘계종합학술대회
• /
• pp.655-658
• /
• 2005

본 논문은 블록 정합에 크게 기여하는 화소들만을 이용함으로 정합 계산량을 줄이는 변형된 블록 정합 알고리즘을 제안한다. 알고리즘 구현과 부가 정보를 고려하여 제안된 알고리즘은 정합 블록을 서브블록들로 나누고, 서브블록의 복잡도를 이용해 일부의 서브블록들만을 선택해 정합을 수행한다. 모의실험 결과는 제안된 알고리즘이 일부의 서브블록만으로 유효한 정합을 수행하는 것을 보여준다.

• 전자공학회논문지
• /
• 제51권2호
• /
• pp.76-84
• /
• 2014

최근 고화질 영상의 증가와 더불어 대용량 영상 데이터의 처리는 높은 연산이 요구되어 병렬 처리 설계가 선택되고 있다. 영상 처리에서 나타나는 많은 단순 연산이 병렬처리 가능한 경우, CPU 기반 병렬처리보다는 GPU 기반 병렬처리를 적용하는 것이 계산문제의 시간과 공간 계산 복잡도를 줄일 수 있다. 본 논문은 영상에서 샷 경계 탐지 알고리즘의 병렬 설계와 구현을 연구하였다. 제안하는 샷 경계 탐지 알고리즘은 프레임 간 지역 화소 밝기 비교와 전역 히스토그램 정보를 이용하는데, 이들 데이터의 계산은 대량의 데이터에 대한 높은 병렬성을 갖는다. 이들 연산의 병렬처리를 최대화하기 위해 화소 밝기와 히스토그램의 계산을 NVIDIA GPU에서 병렬 설계 하였다. GPU 기반 샷 탐지 방법은 국가기록원에서 선택된 10개의 비디오 데이터에 대한 성능 테스트를 수행하였다. 테스트에서 GPU 기반 알고리즘의 탐지율은 CPU 기반 알고리즘과 유사하였으나 약 10배의 연산 속도가 개선되었다.