• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.5
• /
• pp.27-33
• /
• 1993

This paper describes the hardware implementation of 2-D FIR digital filter for a real-time image processing. Generally, the most time-consuming operation in signal processing is the multiplication operation. To avoid it in digital filter. Pelid and Liu proposed the distributed arithmetic method for the one-dimensional case. The implementation method proposed in this paper is to extend Pelid's method to two-dimensional FIR filter using simple ROM lookup table and to use the technique of pipe lining two main operations of memory access and arithmetic. As a result, the speed of our proposed hardware implementation is two times faster than that of conventional methods and can be close to the real time speed.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.4E
• /
• pp.3-10
• /
• 1998

This paper concerns the dynamical behavior, in probabilistic sense, of a feedforward neural network performing auto association for novelty. Networks of retinotopic topology having a one-to-one correspondence between and output units can be readily trained using back-propagation algorithm, to perform autoassociative mappings. A novelty filter is obtained by subtracting the network output from the input vector. Then the presentation of a "familiar" pattern tends to evoke a null response ; but any anomalous component is enhanced. Such a behavior exhibits a promising feature for enhancement of weak signals in additive noise. As an analysis of the novelty filtering, this paper shows that the probability density function of the weigh converges to Gaussian when the input time series is statistically characterized by nonsymmetrical probability density functions. After output units are locally linearized, the recursive relation for updating the weight of the neural network is converted into a first-order random differential equation. Based on this equation it is shown that the probability density function of the weight satisfies the Fokker-Planck equation. By solving the Fokker-Planck equation, it is found that the weight is Gaussian distributed with time dependent mean and variance.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.298-302
• /
• 2009

This paper proposes a single sign-on scheme in which a mobile user offers his credential information to a home network running the OSGi (Open Service Gateway Initiative) service platform, to obtain user authentication and control a remote device through a mobile device using this authentication scheme, based on SAML (Security Assertion Markup Language). Especially by defining the single sign-on profile to overcome the handicap of the low computing and memory capability of the mobile device, we provide a clue to applying automated user authentication to control a remote device via a mobile device for distributed mobile environments such as a home network based on OSGi.

• Proceedings of the KSME Conference
• /
• 2003.11a
• /
• pp.1286-1290
• /
• 2003

This paper presents a newly implemented parallel finite element procedure for contact-impact problems. Three sub-algorithms are includes in the proposed parallel contact-impact procedure, such as a parallel Belytschko-Lin-Tsay (BLT) shell element generation, a parallel explicit time integration scheme, and a parallel contact search algorithm based on the master slave slide-line algorithm. The underlying focus of the algorithms is on its effectiveness and efficiency for inclusion in future finite element systems on parallel computers. Throughout this research, a prototype code, named GT-PARADYN, is developed on the IBM SP2, a distributed-memory computer. Some numerical examples are provided to demonstrate the timing results of the procedure, discussing the accuracy and efficiency of the code.

• Proceedings of the KIPE Conference
• /
• 2002.07a
• /
• pp.697-700
• /
• 2002

In this paper, we proposed the micro web-server Implementation on Strong ARM processor with embedded Linux. The parallel port connecting parallel I/O is controlled via HPPT protocol and web browser program. HTTP protocol is ported into Linux and the micro web server program and port control program are installed on-board memory using CGI to be accessed by web browser, such as Internet Explore and Netscape. 8bit LED and DIP switches are connected to the processor port and the switch input status is monitored and the LED output is controlled from remote hosts vie internet. The result of the proposed embedded micro-web server can be used in automation systems, remote distributed control via internet using web browser.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.45-46
• /
• 2016

최근 GPU를 비롯한 하드웨어의 성능이 급격이 증가하면서 인공지능, 딥러닝 기술에 대한 관심이 높아지고 있다. 또한 데이터가 더욱 방대해 지면서 대용량 데이터를 처리하고 위한 딥러닝 분산 프레임워크에 대한 필요성이 제기되고 있다. 이에 본 논문에서는 대규모의 분산 환경에서 딥러닝 고속 처리를 위한 분산 프레임워크를 비교 분석하였다. 특히 최근 주목받고 있는 인메모리 기반 분산 프레임워크인 Spark, SparkNet, HeteroSpark의 특징을 비교 분석하였다.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.4
• /
• pp.1141-1149
• /
• 2000

This paper proposes new cost error measurement method and analyzes the optimistic and pessimistic cost errors statistically which is resulted from an asynchronous parallel Simulated annealing (SA) in distributed memory multicomputers. The traditional cost error measurement scheme has inherent problems which are corrected in the new method. At each temperature the new method predicts the amount of cost error that an algorithm will tolerate and still converge by the hill-climbing nature of SA. This method also explains three interesting phenomenon of he cost error analytically. So the new cost error measurement method provides a single mechanism for the occurrence of cost error and its control.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.31-34
• /
• 2002

There are two kinds of network architectures in the IEEE 802.11:[1] distributed (ad-hoc) or centralized (infrastructure) wireless network. Centralized networks have an access point (base station) that can control the wireless medium access of stations in these networks. The 802.11 MAC protocol of an access point is the same as those of other stations in the contention period. This paper propose a novel MAC protocol of an access point to solve these problems. This MAC protocol adds a new contention-free period called EPCF (Extended PCF) to resolve accumulated data in the queue of an access point. Simulation results show that the new protocol performs better throughput than the 802.11 standard MAC with the less queue memory site requirement.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.354-356
• /
• 1998

축약 분산 기억 장치는 적응적 문제 해결 능력과 하드웨어화의 용이성으로 인해 현실성이 있는 신경망의 한 모델로 주목받고 있다. 그러나 다층 인식자의 개별 뉴론이 선형의 결정 함수로 해 공간을 이분하고 그들이 다양하게 결합하므로써 일반적인 문제 해결 능력을 갖는데 비해, 축약 분산 기억 장치의 뉴론은 해 공간에서 자신을 중심으로 한 일정 반경 영역을 안과 밖으로 이분하고 이들을 단순하게 합하므로 해 공간이 크기 관계를 갖는 경우 비효율적인 모델로 된다. 본 논문에서는 이러한 축약 분산 기억 장치의 특성과 그 원인을 규명하고 해결 방안으로써 개선된 축약 분산 기억 장치를 제안한다. 아울러 새로운 모델의 적용 예를 ATM 호 수락 제어 과정을 통해 보인다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.634-636
• /
• 2000

기존의 병렬 볼륨 렌더링 방법들은 프로세서간의 발생하는 많은 통신량 때문에 통신 속도가 매우 빠른 병렬컴퓨터를 이용하였고 통신속도가 느린 분산 환경에서는 구현이 불가능해 보였다. 또한 가시화하려는 볼륨 데이터도 점점 방대해지고 있는 실정이다. 이에 본 논문에서는 통신 속도에 구애받지 앉을뿐더러 매우 큰 볼륨데이터를 다루는 병렬/분산 볼륨 렌더링을 제안한다. 본 방법은 고비용을 필요로 하는 원격 메모리 접근 대신에 압축을 기반으로 하여 필요한 데이터를 지역 메모리에서 빠르게 복원함으로써 좋은 성능향상(speedup)을 나타낸다. 이것은 각 프로세서가 전체 볼륨 데이터를 모두 적재하고 있다는 것을 의미한다. 다라서 렌더링 과정중에 발생하는 프로세서간의 통신을 최소화할 수 있었고, 이런 방식은 높은 통신 비용으로 효율적 병렬/분산 처리가 힘든 분산 메모리 병렬 컴퓨터나 PC/워크스테이션 클러스터상에서 매우 적합하다.