http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
플래쉬 메모리 SSD 기반 해쉬 조인 알고리즘의 성능 평가
박장우(JangWoo Park),박상신(SangShin Park),이상원(SangWon Lee),박찬익(ChanIk Park) 한국정보과학회 2010 정보과학회 컴퓨팅의 실제 논문지 Vol.16 No.11
데이터베이스 관리 시스템의 핵심 알고리즘인 해쉬 조인은 해싱을 위한 메모리가 부족한 경우(즉, 해쉬 테이블 오버플로우) 디스크 입출력를 유발하게 된다. 하드디스크를 임시 저장공간으로 사용할 경우, 해쉬 조인의 probing 단계에서 과도한 임의 읽기로 인해 I/O 시간이 성능을 저하시키게 된다. 한편, 플래시메모리 SSD가 저장장치로 각광을 받고 있으며, 머지않아 엔터프라이즈 환경에서 하드디스크를 대체할 것으로 예상 된다. 하드디스크와 달리, 기계적인 동작 장치가 없는 플래시메모리 SSD의 경우 임의 읽기에서 빠른 성능을 보이기 때문에 해쉬 조인의 성능을 크게 향상시킬 수 있다. 본 논문에서는 플래시 메모리 SSD를 해쉬 조인을 위한 임시 저장공간으로 사용할 경우의 몇 가지 중요하고 현실적인 이슈들을 다룬다. 우선, 해쉬 조인의 I/O 패턴을 자세히 설명하고, 하드디스크에 비해 플래시메모리 SSD가 수십 배에 가까운 성능 향상을 보이는 이유를 설명한다. 다음으로, 클러스터 크기(즉, 해쉬 조인 알고리즘에서 사용하는 I/O 단위)가 성능에 미치는 영향을 제시하고 분석한다. 마지막으로, 하드디스크의 경우, DBMS의 질의 최적화기가 산출하는 비용이 실 수행시간과 편차가 클 수 있는데 반해, 플래시메모리 SSD의 경우 비용 산출을 정확히 하게 됨을 실험적으로 보인다. 결론적으로, 플래시메모리 SSD를 해쉬 조인을 위한 임시 저장공간으로 사용할 경우, 빠른 성능과 더불어 질의 최적화기의 비용 산출이 훨씬 더 신뢰할 수 있음을 보인다. Hash join is one of the core algorithms in databases management systems. If a hash join cannot complete in one-pass because the available memory is insufficient (i.e., hash table overflow), however, it may incur a few sequential writes and excessive random reads. With harddisk as the tempoary storage for hash joins, the I/O time would be dominated by slow random reads in its probing phase. Meanwhile, flash memory based SSDs (flash SSDs) are becoming popular, and we will witness in the foreseeable future that flash SSDs replace harddisks in enterprise databases. In contrast to harddisk, flash SSD without any mechanical component has fast latency in random reads, and thus it can boost hash join performance. In this paper, we investigate several important and practical issues when flash SSD is used as tempoary storage for hash join. First, we reveal the I/O patterns of hash join in detail and explain why flash SSD can outperform harddisk by more than an order of magnitude. Second, we present and analyze the impact of cluster size (i.e., I/O unit in hash join) on performance. Finally, we emperically demonstrate that, while a commerical query optimizer is error-prone in predicting the execution time with harddisk as temporary storage, it can precisely estimate the execution time with flash SSD. In summary, we show that, when used as temporary storage for hash join, flash SSD will provide more reliable cost estimation as well as fast performance.
Application of EEG for Multimodal Human-Machine Interface
Jangwoo Park,Il Woo,Shinsuk Park 제어로봇시스템학회 2012 제어로봇시스템학회 국제학술대회 논문집 Vol.2012 No.10
There are many input modalities for human-machine interface (HMI). Brain-signal that is one of biosignal has been studied as an input modality for HMI. Brain-signal based HMI can help disabled people to communicate with a machine using the brain’s electrical activity. this study is focuses on usability of the EEG-based HMI’s for available tools in real life and possibility of the EEG signal as input modality of multimodal interface. This study attempt to explore the electroencephalogram (EEG) signal measurement and analysis methods related to concentration for multimodal Interface. The experiments have been performed with various tasks, such as self-concentration, self-arithmetic (non-display), self-arithmetic (show display) and eye-closing. EEG signals are recorded while subjects perform each task on Fz, Cz, Pz. The receiver operating characteristic (ROC) curve analysis is to determine the threshold on each task. Rate of distinction range is 50.32% ~ 56.77% with the threshold about self-arithmetic and 71.67%~78.33% with the threshold about eye-closing. There are some meaningful results about threshold, self-arithmetic and eye-close activity. It can be used for brain-machine interface and multi-modal interface.
Greenhouse Management Framework based on Localization Using RGPSi and AoA
Jangwoo Park,Hong-geun Kim,Yongyun Cho,Changsun Shin,Kyungryong Cho,Dong-guk Park 보안공학연구지원센터 2015 International Journal of Smart Home Vol.9 No.12
In this paper, the concept of the smart greenhouse control framework is proposed with the help of localization algorithm. The proposed smart greenhouse framework consists of the data aggregator with the database, the environment control part and the crop growth status control part. The data aggregator has been equipped with the various sensors to measure the data for crop growth. The sensors have ability of communication and calculation of location of target. As a localization algorithm, RGPSi is used. RGPSi uses the iteration algorithm similar to the GPS algorithm, but utilize the ratio of signal strengths instead of absolute strengths. To improve the accuracy of the localization, the method of AOA(angle of arrival) of signal will be added. The environment control part has the role to generate the control signals for the greenhouse to operate properly for satisfying the goal. Information for the status of crop growth is generated from the growth control part. With the help of the goal function two parts will be interacting each other and have fed back the sensed data from the greenhouse.
Output Energy Maximization Approach for Carrier-Phase Offset Recovery of 8-VSB Signals
Jangwoo Park,Thinh Nguyen,Wonzoo Chung [Institute of Electrical and Electronics Engineers 2016 IEEE transactions on broadcasting Vol.62 No.1
<P>This paper presents a blind adaptive carrier-phase offset recovery algorithm based on an output energy maximization approach for eight-level vestigial sideband (8-VSB) signals. Unlike conventional quadrature amplitude modulation signals, the 8-VSB signals in practice have an asymmetric energy balance between the in-phase and quadrature components, which can be used for recovering phase offset, but this has been neglected. We investigate this energy imbalance of the VSB signals and propose a blind adaptive phase offset recovery scheme by maximizing the energy of the in-phase component. Due to the maximum energy property, the proposed algorithm results in superior mean square error performance without undesirable local minima in comparison with the existing phase offset recovery algorithm based on dispersion minimization. We verify the performance of the proposed algorithm with a mathematical analysis and simulation.</P>