• Journal of KIISE:Computer Systems and Theory
• /
• v.27 no.12
• /
• pp.951-959
• /
• 2000

본 논문의 병렬음성인식 모델은 연속 은닉 마코프 모델(HMM; hidden Markov model)에 기반한 병렬 음소인식모듈과 계층구조의 지식베이스에 기반한 병렬 문장인식모듈로 구성된다. 병렬 음소인식 모듈은 수천개의 HMM을 병렬 프로세서에 분산시킨 수, 할당된 HMM에 대한 출력확률 계산과 Viterbi 알고리즘을 담당한다. 지식베이스 기반 병렬 문장인식모듈은 음소모듈에서 공급되는 음소열과 지안하는 병렬 음성인식 알고리즘은 분산메모리 MIMD 구조의 다중 트랜스퓨터와 Parsytec CC 상에 구현되었다. 실험결과, 병렬 음소인식모듈을 통한 실행시간 향상과 병렬 문장인식모듈을 통한 인식률 향상을 얻을 수 있었으며 병렬 음성인식 시스템의 실시간 구현 가능성을 확인하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.3
• /
• pp.360-371
• /
• 1993

Flood search algorithm is known to be an effective routing mechanism for tactical application, since it provides high degree of survivability and robustness. But it is known that it has significant drawbacks with respect to the network efficiency [1]. We consider a tactical circuit-switched grid network with a maximum of four links and two priority classes of voice traffic, Using the minimum first-derivative length (MFDL) path, we improve the blocking probability performance of the circuit-switched network without increasing the call set-up time and processor loading of the algorithm.

• Journal of KIISE:Computer Systems and Theory
• /
• v.28 no.3
• /
• pp.134-143
• /
• 2001

NUMA 구조 다중 프로세서 시스템에서는 상호 연결망으로서의 버스의 제약을 극복하기 위해 지점간 링크를 이용한 링 구조가 제안되었다. 링 구조 NUMA 시스템에서, 전송이 페이지 단위(2K 바이트 이상)로 이루어지는 디스크 입출력 요구는 지점간 링크에 대한 오랜 접근을 요구하여 지점간 링크의 트래픽을 증가시키는 원인으로 작용한다. 본 논문에서는 지점간 링크의 트래픽을 줄이기 위한 방안으로 입출력 전용 채널을 사용하는 세 가지 디스크 입출력 구조를 제시한다. 제시되는 디스크 입출력 구조를 사용하는 시스템은 디스크 입출력 요구로 인한 지점간 링크의 접근을 없앰으로써 기존 시스템에 비해 트랜잭션의 수행시간을 줄일 수 있다는 장점이 있다. 중앙집중형, 병렬형, 분산형의 세 가지 디스크 입출력 구조를 설계하기 위한 고려 사항과 구현 비용 등을 생각해 본 후, 노드수, 원격 디스크 접근 확률, 디스크 입출력 전송 데이터 크기 등을 달리한 여러 시스템 환경에서의 각 구조의 성능을 비교, 분석해 본다.

• Proceedings of the KIEE Conference
• /
• 2003.11a
• /
• pp.110-112
• /
• 2003

본 논문에서는 확률적인 전역 최적화 알고리즘인 유전 알고리즘과 경험적인 최적화 알고리즘인 타부 탐색법을 병렬화함으로써 최적해 탐색성능 및 탐색속도를 개선하는 방안을 개발하였다. 배전계통의 주변압기에서 고장이 발생한 경우에는 여러 정전구역에 대한 복구문제가 되어 매우 복잡하고 많은 연산량을 수반한다. 따라서 제안한 고장복구 알고리즘은 PC 클러스트링을 이용하여 각 프로세서별로 유전 알고리즘 또는 타부 탐색법을 사용하여 최적해를 탐색한 후 일정 기간 이후에 해를 교환함으로써 배전계통에서의 주변압기 고장발생시 최적해 탐색에 소요되는 시간을 단축하였으며 고장복구 지원시스템의 성능개선을 도모하였다. 제안한 알고리즘의 유용성을 입증하기 위하여 한전의 실 배전계통 주변압기 고장복구 문제에 적용함으로써 제안한 알고리즘이 해의 탐색속도 및 해의 성능면에서 우수함을 입증하였다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.37 no.6
• /
• pp.1-9
• /
• 2000

This paper presents a Cluster Oriented Multistage Interconnection Network called COMR. COMR can be constructed suitable for the parallel application with localized communication by providing the shortcut path inside the processor-memory cluster which has frequent data communication. We evaluate the performance of COMR with respect to probability of acceptance, bandwidth, cost-effectiveness and average distance under varying degrees of localized communication. According to the result of analysis for performance evaluation, COMR shows higher performance than the regular MINs of the same network size in the highly localized communication. In the worst case, the diameter of an N$\times$N COMR is only n+1 which has only one stage more as compared the MIN with the same network size. Therefore COMR can be used as an attractive interconnection network for parallel applications with not only the localized communication distribution but also the uniform distribution in shared-memory multiprocessor system.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.40 no.12
• /
• pp.87-94
• /
• 2003

In this paper, we propose an IFFT/FFT design method to minimize quantization error in IEEE 802.11a WLAN. In the proposed algorithm, the twiddle coefficient of IFFT/FFT processor is manipulated by the statistics distribution of the input data at each stage. We applies this algorithm to radix-2/$^2$ SDF architecture. Both IFFT and FFT processor shares the circuit blocks cause to the symmetric architecture. The maximum quantization error with the 10 bits length of the input and output data is 0.0021 in IFFT and FFT that has a self-loop structure with the proposed method. As a result, the proposed architecture saves 3bits for the data to keep the same resolution compared with the conventional method.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.4
• /
• pp.1058-1070
• /
• 1999

In this paper, we propose a control model which can control the burst input messages of the BSC(Base Station controller) in mobile communication systems more efficiently and reliably, by dividing the input messages characteristically and using multiprocessor system. Using M/M/c/K queueing model, we briefly analyze proposed model to get characteristic parameters which are required to performance improvement. On the base of the results, we compare our proposed model with the conventional one by using SLAM II with regard to the following factors : the call blocking rate of the input message, the distribution of average queue length, the utilization of process controller(server), and the distribution of average waiting time in queue. In addition, we modified our model which has overload control function for burst input messages, and analyzed its performance.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.9
• /
• pp.40-49
• /
• 2011

Integrating massive components and low-power policies have been actively investigated for system-on-chip designs. But in recent years, finding the optimal interconnection structure among heterogeneous components has emerged as a critical system design issue. Therefore, various simulation tools to model interconnection designs are being developed and performance evaluation of simulation is reflected in the real design. But most of the simulation environments employ traffic generation based on the mathematical probability functions, and such traffic generation cannot fully cover for various situations that may be occurred in the real system. Therefore, the demand for traffic pattern generation based on real applications is increasing. However, there have been few simulators that adopt application-specific traffic generators. This paper proposes a novel traffic generation method in simulating various interconnection structures for multi-processor system-on-chip design. The proposed traffic generation method can generate traffic patterns that can reflect the actual characteristics of the application and evaluate the performance of an interconnection structure under more realistic circumstance than traffic patterns using mathematical probability functions. By comparing the differences between the proposed method and the one based on mathematical probability functions, this paper shows advantages of the proposed traffic generation method.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.3
• /
• pp.95-103
• /
• 2012

Modern society is experiencing a variety of crimes, and to prevent crime is being studied. Existing studies related to the crime of crimes that occur on spatial analysis and geographic information, or to analyze the type of criminal offense of studies have been conducted, However the existing studies of the geographical and psychological crime that occurs throughout the study area and by analyzing the motives for the crime prevention research is the most. In this paper, we introduce Markov processor model for predicting the crime is present. Of several crimes, murder, government official crimes, the incidence of violent crime has occurred over time by using the predicted incidence of crime. Presented in this paper, predictive modeling is used in a crime occurred in the average duration of the overall average number of crimes that occurred in the one-year average, which recently labeled as the average prediction was compared to if you can increase the likelihood, recent average to apply to increase the probability of the prediction that crime have been investigated.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.6
• /
• pp.153-162
• /
• 2009

This paper suggests the novel light-weight signal processing algorithm for wireless sensor network applications which needs low computing complexity and power consumption. Exponential average method (EA) is utilized by real time, to process the magnetometer signal which is analyzed to understand the own physical characteristic in time domain. EA provides the robustness about noise, magnetic drift by temperature and interference, furthermore, causes low memory consumption and computing complexity for embedded processor. Hence, optimal parameter of proposal algorithm is extracted by statistical analysis. Using general and precision magnetometer, detection probability over 90% is obtained which restricted by 5% false alarm rate in simulation and using own developed magnetometer H/W, detection probability over 60~70% is obtained under 1~5% false alarm rate in simulation and experiment.