• 전자공학회논문지S
• /
• 제34S권9호
• /
• pp.23-30
• /
• 1997

In this paper, a decision feedback recurrent neural equalization (DFRNE) scheme is proposed for adaptive equalization problems. The proposed equalizer models a nonlinear infinite impulse response (IIR) filter. The modified Real-Time recurrent Learning Algorithm (RTRL) is used to train the DFRNE. The DFRNE is applied to both linear channels with only intersymbol interference and nonlinear channels for digital video cassette recording (DVCR) system. And the performance of the DFRNE is compared to those of the conventional equalizaion schemes, such as a linear equalizer, a decision feedback equalizer, and neural equalizers based on multi-layer perceptron (MLP), in view of both bit error rate performance and mean squared error (MSE) convergence. It is shown that the DFRNE with a reasonable size not only gives improvement of compensating for the channel introduced distortions, but also makes the MSE converge fast and stable.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 추계학술대회 논문집 학회본부 B
• /
• pp.624-626
• /
• 1998

Adaptive channel equalization accomplished without resorting to a training sequence is known as blind equalization. In this paper, we present a auto-switching blind, equalization for 8-VSB HDTV receiver. The scheme operate in two mode : blind equalization mode and decision-directed equalization mode. This proposed scheme changes from the blind equalization mode at high error levels to the decision-directed equalization mode at lower error levels smoothly and automatically. Manual switch from the blind equalization mode to the decision-directed mode is not necessary.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 추계학술대회 학술발표 논문집
• /
• pp.474-478
• /
• 1998

A new approach for the decision feedback equalizer(DFE) based on the back-propagation neural networks is described. We propose the method of optimal structure for back-propagation neural networks model. In order to construct an the optimal structure, we first prescribe the bounds of learning procedure, and the, we employ the method of incrementing the number of input neuron by utilizing the derivative of the error with respect to an hidden neuron weights. The structure is applied to the problem of adaptive equalization in the presence of inter symbol interference(ISI), additive white Gaussian noise. From the simulation results, it is observed that the performance of the propose neural networks based decision feedback equalizer outperforms the other two in terms of bit-error rate(BER) and attainable MSE level over a signal ratio and channel nonlinearities.

• 한국통신학회논문지
• /
• 제33권5A호
• /
• pp.500-508
• /
• 2008

Direct-sequence ultra-wideband(DS-UWB) system is being considered as one of promising transmission technologies for wireless personal area networks(WPANs). Due to relatively low spreading factors and huge bandwidth of transmit signal, a DS-UWB receiver needs to be equipped not only with a rake receiver but also with an equalizer, of which the equalizer is not required for traditional direct-sequence code division multiple access(DS-CDMA) systems. The number of rake fingers is limited in practice, influencing the performance of the subsequent equalizer. In this paper, we derive a decision feedback equalizer(DFE) for DS-UWB systems based on the minimum mean square error(MMSE) criterion, and investigate the impact of various parameters on the DFE performance in realistic scenarios. In particular, we propose an approach to improving the performance of the DFE using additional channel estimates for multipaths not combined in the rake receiver, and discuss how the accuracy of channel estimation affects desirable DFE configuration. Moreover, we present simulation results that show the impact of turbo equalization on the DFE performance.

• 한국통신학회논문지
• /
• 제36권8C호
• /
• pp.516-521
• /
• 2011

이 논문에서는 심각한 다중경로 페이딩 왜곡과 충격성 잡음을 극복하기 위해 결정궤환 구조를 갖추고 오차의 영확률을 최대화하는 비선형 알고리듬(MZEP-DF)를 제안하였다. 제안된 MZEP-DF 알고리듬은 충격성 잡음에 강인성을 보였고 궤환구조 영역에서 잔여 심볼간 간섭을 제거하는 능력을 보였다. 기존 LMS 알고리듬이 정상상태 오차 전력에 대해 -3 dB 이하로 수렴하지 못했던 심각한 다중경로 페이딩과 충격성 잡음 환경에 대해 제안한 알고리듬은 선형 MZEP 알고리듬과 비교하였을 때, 10 dB 이상 성능향상을 보였다.

• 한국항공운항학회지
• /
• 제22권3호
• /
• pp.50-59
• /
• 2014

It has been focused on that the major reasons of aircraft accidents resulted not from human error but from the failure of teamwork or communication in 1980's. Such opinions were suggested in the workshop, so called, "Resource Management on the Flight Deck" by NASA in 1979. The researchers agreed the fact the source of human error was originated from the failure in teamwork, communication or even in leadership of captain. Due to the rapid development of aircraft technologies, the reasons for aircraft accident could be easily found out. According to the analysis results of the technology, most of reasons for the accident might directly be connected not to human error or stick-Rudder skill but to situational awareness, communication, leadership or decision making in the aircraft. This paper has tried to research empirically the satisfaction of flight crews who have gone thorugh CRM training in the commercial airline. Based on the quantitative scale by J. Ford et al. (2014), this paper has proved which characteristics of CRM training has a positive impact on the overall satisfaction of CRM training. It was proven that the teamwork and decision making programs among CRM training have a major effect to the satisfaction level of flight crews.

• 해양환경안전학회지
• /
• 제17권2호
• /
• pp.149-154
• /
• 2011

해상에서 인명과 재산을 보호하기 위해서는 해양사고를 예방하고 긴급상황 시에 신속하게 대처하는 능력이 중요시된다. 해양사고의 80%가 선박의 운항과 관리에 관련된 인적요인으로 발생한다. 그것은 대부분의 해양사고가 한 개인의 순간적인 실수에서 유발되는 것이 아니라 그 배경에는 인적과실을 유발하는 잠재적인 요인들이l 복합적으로 작용한다는 것을 나타낸다. 따라서 본 연구에서는 선박에서 선상의 리더십이 선박에 미치는 영향과 선박의 의사결정 과정을 알아보고 특히, 선박의 긴급상황 시에 발생할 수 있는 행동특성의 이해를 통하여 특정선박에서 인적과실을 줄이기 위한 해결방안 및 대응전략을 제시하였다.

• 한국통신학회논문지
• /
• 제40권2호
• /
• pp.247-252
• /
• 2015

영확률을 성능기준으로 하는 적응 알고리듬은 충격성 잡음에 강인함을 나타내며 그 결정 궤환 알고리듬은 심각한 다경로 채널 왜곡을 효과적으로 보상하는 것으로 알려져 있다. 그러나 이러한 결정 궤환 영확률 알고리듬은 각 필터 구역에 대해 매 샘플시간마다 여러 합산 동작을 계산해야하는데 이것이 실제 구현에 장애가 되고 있다. 이 논문에서는 반복적 기울기 추정 방식을 가진 결정 궤환 영확률 알고리듬을 제안하며 이 알고리듬은 기존 계산량 O(N)을 샘플 사이즈 N에 무관한 상수량으로 줄일 수 있음을 보인다. 또한 초기상태와 안정상태의 가중치 갱신이 연속적인 과정으로 이루어져 결정 궤환에서 어떤 기울기 추정 오류 전파도 일으키지 않음을 보인다.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권8호
• /
• pp.1037-1045
• /
• 1999

We propose two new classes of robust blind equalization algorithms against abrupt changes of channel conditions, which we call a triple-mode algorithm(TMA) and an automatic switch-over algorithm(ASA). The conventional DMGSA exhibits slow convergence rates due to the incorrect equalizer tap-updating process under the severe channel conditions. In order to speed up the convergence process, the TMA operates in triple-mode that is based on the dual-mode of the DMGSA incorporated with the tap-updating control modes of the SGA as well as the MSGA. Without resorting to the decision region for selecting the operation mode in the TMA, the ASA automatically switches the blind mode to the smoother conventional decision-directed mode. The ASA uses the error functional that is the weighted sum of the Generalized Sato error and the decision-directed error, where the weights correspond to the channel conditions. Test results on 16-QAM and 8-VSB datas confirm that the TMA and the ASA perform well under the sudden changes of channel conditions.

• 대한전자공학회논문지TC
• /
• 제43권8호
• /
• pp.69-76
• /
• 2006

이 논문에서는 조정 가능한 소프트 판정기를 이용하여 자력복구 판정궤한 채널등화기의 적응 모드를 포착모드와 추적 모드 사이에서 최적화하는 기법들을 제안한다. 제안된 기법들은 주어진 SNR에 따라 소프트 판정기를 최적화하여 DFE를 위한 궤환 신호를 생성하고 그에 따라 자력복구 IIR 필터 적응모드와 DD-LMS 적응모드를 결합한 적응방식을 적용한다. 제안된 기법들은 포착모드와 추적모드 사이의 최적화된 스위칭을 성취할뿐아니라 DFE 에러 propagation을 최소화하는데도 기여한다.