• Journal of the Korea Computer Graphics Society
• /
• v.22 no.4
• /
• pp.21-30
• /
• 2016

Risk management system should be able to support a decision making within a short time to analyze stream data in real time. Many analytical systems consist of CPU computation and disk based database. However, it is more problematic when existing system analyzes stream data in real time. Stream data has various production periods from 1ms to 1 hour, 1day. One sensor generates small data but tens of thousands sensors generate huge amount of data. If hundreds of thousands sensors generate 1GB data per second, CPU based system cannot analyze the data in real time. For this reason, it requires fast processing speed and scalability for analyze stream data. In this paper, we present a fast visualization technique that consists of hybrid database and GPU computation. In order to evaluate our technique, we demonstrate a visual analytics system that analyzes pipeline leak using sensor and tweet data.

• Journal of Korea Multimedia Society
• /
• v.10 no.8
• /
• pp.1017-1028
• /
• 2007

In this paper we propose an effective method to represent, store, and retrieve video streams efficiently from a video database. We extract features from each video frame, normalize the feature values, and represent them as values in the range [0,1]. In this way a video frame with f features can be represented by a point in the f-dimensional space $[0,1]^f$, and thus the video stream is represented by a trail of points in the multidimensional space. The video stream is partitioned into video segments based on camera shots, each of which is represented by a trend vector which encapsulates the moving trend of points in a segment. The video stream query is processed depending on the comparison of those trend vectors. We examine our method using a collection of video streams that are composed of sports, news, documentary, and educational videos. Experimental results show that our trend vector representation reduces a reconstruction error remarkably (average 37%) and the retrieval using a trend vector achieves the high precision (average 2.1 times) while maintaining the similar response time and recall rate as existing methods.

• The KIPS Transactions:PartD
• /
• v.14D no.6
• /
• pp.597-604
• /
• 2007

Stream data is a continuous flow of information that mostly arrives as the form of an infinite rapid stream. Recently researchers show a great deal of interests in analyzing such data to obtain value added information. Here, we propose an efficient cube computation algorithm for multidimensional analysis of stream data. The fact that stream data arrives in an unsorted fashion and aggregation results can only be obtained after the last data item has been read. cube computation requires a tremendous amount of memory. In order to resolve such difficulties, we compute user selected aggregation fables only, and use a combination of an way and AVL trees as a temporary storage for aggregation tables. The proposed cube computation algorithm works even when main memory is not large enough to store all the aggregation tables during the computation. We showed that the proposed algorithm is practically fast enough by theoretical analysis and performance evaluation.

• The KIPS Transactions:PartD
• /
• v.14D no.7
• /
• pp.733-742
• /
• 2007

A data stream is a massive unbounded sequence of data elements continuously generated at a rapid rate. Due to this reason, memory usage for data stream analysis should be confined finitely although new data elements are continuously generated in a data stream. To satisfy this requirement, data stream processing sacrifices the correctness of its analysis result by allowing some errors. The old distribution statistics are diminished by a predefined decay rate as time goes by, so that the effect of the obsolete information on the current result of clustering can be eliminated without maintaining any data element physically. This paper proposes a grid based clustering algorithm for a data stream. Given a set of initial grid cells, the dense range of a grid cell is recursively partitioned into a smaller cell based on the distribution statistics of data elements by a top down manner until the smallest cell, called a unit cell, is identified. Since only the distribution statistics of data elements are maintained by dynamically partitioned grid cells, the clusters of a data stream can be effectively found without maintaining the data elements physically. Furthermore, the memory usage of the proposed algorithm is adjusted adaptively to the size of confined memory space by flexibly resizing the size of a unit cell. As a result, the confined memory space can be fully utilized to generate the result of clustering as accurately as possible. The proposed algorithm is analyzed by a series of experiments to identify its various characteristics

• Journal of KIISE
• /
• v.42 no.12
• /
• pp.1480-1485
• /
• 2015

As SSDs are increasingly being used as high-performance storage or caches, attention is increasingly paid to the provision of SSDs with Quality-of-Service for I/O request streams of various applications in server systems. Since most SSDs are using the AHCI controller interface on a SATA bus, it is not possible to provide a differentiated service by distinguishing each I/O stream from others within the SSD. However, since a new SSD interface, the NVME controller interface on a PCI Express bus, has been proposed, it is now possible to recognize each I/O stream and schedule I/O requests within the SSD for differentiated services. This paper proposes Flash Operation Group Scheduling within NVME-based flash storage devices, and demonstrates through QEMU-based simulation that we can achieve a proportional bandwidth share for each I/O stream.

• The KIPS Transactions:PartD
• /
• v.15D no.6
• /
• pp.741-754
• /
• 2008

Due to the characteristics of a data stream, it is very important to confine the memory usage of a data mining process regardless of the amount of information generated in the data stream. For this purpose, this paper proposes the Prime pattern tree(PPT) for finding frequent itemsets over data streams with using the confined memory space. Unlike a prefix tree, a node of a PPT can maintain the information necessary to estimate the current supports of several itemsets together. The length of items in a prime pattern can be reduced the total number of nodes and controlled by split_delta $S_{\delta}$. The size and the accuracy of the PPT is determined by $S_{\delta}$. The accuracy is better as the value of $S_{\delta}$ is smaller since the value of $S_{\delta}$ is large, many itemsets are estimated their frequencies. So it is important to consider trade-off between the size of a PPT and the accuracy of the mining result. Based on this characteristic, the size and the accuracy of the PPT can be flexibly controlled by merging or splitting nodes in a mining process. For finding all frequent itemsets over the data stream, this paper proposes a PPT to replace the role of a prefix tree in the estDec method which was proposed as a previous work. It is efficient to optimize the memory usage for finding frequent itemsets over a data stream in confined memory space. Finally, the performance of the proposed method is analyzed by a series of experiments to identify its various characteristics.

• Proceedings of the IEEK Conference
• /
• 2000.09a
• /
• pp.709-712
• /
• 2000

본 논문에서는 인터 프레임(P-프레임) 압축으로 인해 발생되는 블록화 현상을 모델링 하고 기존에 정의된 제한요소를 효율적으로(reasonable) 축소하고 축소된 제한요소를 이용한 후처리를 통해 블록경계는 물론 블록 내부의 불연속을 효율적으로 제거하는 방법을 제안한다. 다음 프레임의 예측영상을 블록화가 제거된 영상을 사용함으로써 프레임간 영상 스트림에서도 주관적으로나 객관적으로나 현저하게 블록화가 제거된 영상을 얻을 수 있다. 제안된 알고리즘은 HDTV나 영상통신과 같은 영상 스트림의 후처리 과정에 적합하다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.04a
• /
• pp.861-864
• /
• 2017

본 논문은 Exponential histogram을 사용하여 데이터스트림에서 개념 변화를 검출 하는 기법을 제안한다. 스트림 데이터와 같이 빠르게 증가하는 데이터에 대한 개념 변화를 찾는 것은 중요 문제이다. 기존에 사용하던 슬라이딩 윈도우 기반의 방법들은 과거의 데이터를 버렸지만, 제안하는 방법은 과거의 데이터를 효율적으로 저장하며, 윈도우의 크기를 변경 할 수 있는 방법을 제안한다. 실험을 통해 제안하는 방법에 대한 효율성과 정확성을 보인다.

• Proceedings of the Korea Institutes of Information Security and Cryptology Conference
• /
• 1998.12a
• /
• pp.421-441
• /
• 1998

본 논문에서는 스트림 암호 알고리즘과 블록 암호 알고리즘과 같은 비밀키 암호 시스템의 통계적 특성을 측정하기 위하여 사용된 여러 가지 randomness 테스트 방법들을 구현하여 그 성능을 randomness test의 power측면에서 서로 비교하였다. 여기서, power란 randomness test가 nonrandom한 비트 스트림을 얼마나 정확하게 검정할 수 있는 가를 나타내는 적도이다. 그리고, 스트림 암호 알고리즘과 블록 암호 알고리즘의 통계적 특성을 측정하기 위해 가장 효율적이라고 생각되는 테스트 방법들을 찾아 이 방법들을 사용할 것을 제안한다. 끝으로 제안된 방법들을 이용하여, DES, AES후보 알고리즘, SEED의 통계적 특성을 분석하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.10a
• /
• pp.805-808
• /
• 2000

통신망의 급격한 발전과 통신 속도의 향상에 따라 암호 알고리듬의 고속화 필요성이 절실하다. 본 논문에서는 LFSR을 고속화하기 위하여 한 클럭에 m번의 이동이 이루어지는 고속형 HS-LFSR을 제안하였고, 이를 기본으로 다수의 키 수열 발생기를 병렬 연결하여 속도를 개선시킨 병렬형 스트림암호를 제안하였다. 그리고 병렬형 스트림 암호 예로서 m-병렬 합산 수열 발생기(m-parallel SUM-BSG)를 제안하여 m = 8인 병렬 발생기를 세부 설계 예시하였으며, 제안된 발생기는 기존의 비도 수준을 유지하면서 처리 속도를 m배 높일 수 있음을 확인하였다.