• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10c
• /
• pp.841-843
• /
• 2001

인터넷의 사용자가 급증하여 고가용성과 확장성을 지닌 고성능의 인터넷 서버들이 요구된다. 클러스터 시스템은 이러한 요구사항을 만족시킬 수 있는 서버이다. 본 연구에서는 N-node heartbeat 을 구현하였고, 이것을 기반으로 하는 고가용 부하분산 클러스터, PersistentCluster를 구현하였다. PersistentCluster는 로드 벨런서가 사용자의 요구를 서버들에게 분산시켜주는 LVS 시스템에서 로드벨런서가 고장나면 나머지 서버중에 하나가 그 역할을 인계 받아 계속 수행하는 고가용성 클러스터링 솔루tus이다. PersistentCluster는 로드벨런서만 heartbeat 메시지를 전송하는 비대칭 heartbeat을 채택하여 시스템의 메시지 전송 및 처리 오버헤드를 감소시켰다. 비대칭 heartbeat을 적용할 경우에 나타나는 각 노드의 부하 감소량을 실측하며 비대칭 heartbeat 의 성 능을 평가하였다.

• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.4
• /
• pp.246-252
• /
• 2017

Non-invasive and non-restrictive methods for measuring the physiological functions of the human body are useful for health care, security, and surveillance. In this paper, a new method that extracts human heartbeat information by utilizing ultra-wideband (UWB) impulse radar is proposed. The amplitude spectra of received radar pulses reflected from the human body are accumulated at specific time intervals, and chirp z-transform (CZT) is used to extract the heartbeat frequency from the amplitude spectra. The heartbeat frequency can be extracted with high-frequency resolution in the frequency band of the heartbeat of interest using CZT. Experimental results to verify the performance of the proposed method show that a highly accurate extraction of the heartbeat frequency is possible using this method.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04a
• /
• pp.4-6
• /
• 2001

일반 시스템에서는 결함이 발생하였을 때, 즉 어떠한 작업을 수행하는 프로세스 또는 하드웨어에 결함이 발생하였을 때 작업이 중단되거나 처음부터 다시 수행하여야 한다. 그러나, 고가용성 시스템은 일반 다른 시스템과 달리 어떠한 결함이 발생했을 때에도 지속적으로 서비스를 수행할 수 있는 하드웨어나 소프트웨어 시스템이 구축되어 있다. 고가용성 시스템에서 Heartbeat을 이용하여 시스템에서 발생하는 결함이 발견하여 필요한 조치를 위할 수 있도록 한다. 또한, 체크포인트(Checkpoint)f와 롤백(Roll-Back) 기법을 사용하여 컴퓨팅의 손실을 최소화하기 위하여 컴퓨팅 작업을 처음부터 다시 시작하는 것이 아니라 최근의 상태 저장 순간으로 되돌아가 다시 시작한다. 본 논문에서는 고가용성 시스템에서 체크포인트와 Heartbeat을 사용할 때 체크포인트 간격과 Heartbeat 간격에 따른 평균 수행시간을 구하고, 최적의 체크포인트 간격을 적용한 경우에 Heartbeat 간격에 따른 시스템의 성능을 분석하고 비교하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.11
• /
• pp.5252-5268
• /
• 2018

To support a large-scale Hadoop cluster, Hadoop heartbeat messages are designed to deliver the significant messages, including task scheduling and completion messages, via piggybacking to reduce the number of messages received by the NameNode. Although Hadoop is designed and optimized for high-throughput computing via batch processing, the real-time processing of large amounts of data in Hadoop is increasingly important. This paper evaluates Hadoop's performance and costs when the heartbeat period is controlled to support latency sensitive applications. Through an empirical study based on Hadoop 2.0 (YARN) architecture, we improve Hadoop's I/O performance as well as application performance by up to 13 percent compared to the default configuration. We offer a guideline that predicts the performance, costs and limitations of the total system by controlling the heartbeat period using simple equations. We show that Hive performance can be improved by tuning Hadoop's heartbeat periods through extensive experiments.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.9
• /
• pp.200-207
• /
• 2014

Arrhythmia electrocardiogram signal contains a specific unusual heartbeat with abnormal morphology. Because unusual heartbeat is useful for diagnosis and classification of various diseases, such as arrhythmia, detection of unusual heartbeat from the arrhythmic ECG signal is very important. Amplitude and kurtosis at R-peak point and RR interval are characteristics of ECG signal on R-wave. In this paper, we provide a method for detecting unusual heartbeat based on these. Through the value of the attribute deviates more from the average value if unusual heartbeat is more certainly, the proposed method detects unusual heartbeat in order using the mean and standard deviation. From 15 ECG signals of MIT-BIH arrhythmia database which has R-wave distortion, we compare the result of conventional method which uses the fixed threshold value and the result of proposed method. Throughout the experiment, the sensitivity is significantly increased to 97% from 50% using the proposed method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.4
• /
• pp.1266-1271
• /
• 2010

In this paper, we implemented a heartbeat-signal monitoring system that can measure and check the heartbeat of a human body in any time and anywhere. The implemented prototype is composed of heartbeat measurement terminal, gateway, and server that collect and process data. The terminal compose sensor network using ZigBee protocol. And server acquire heartbeat signal through differentiating collected data two times, and store above data to database and manage it. We can understand the status of examinee's heart without restriction and self-awareness. Through experiment, we confirmed the possibility of ubiquitous healthcare system based on sonsor network using the ZigBee.

• Journal of Environmental Science International
• /
• v.16 no.4
• /
• pp.425-432
• /
• 2007

The water flea has been used as a test organism of toxicity test for surface water. Toxicity test with water flea is categorized into two parts. One is acute toxicity test with observing immobility and mortality and the other is chronic toxicity test determined by survival and reproduction of water flea. Heartbeat measurement of water flea was designed as a short-term toxicity test in this study. Direct measurement of heartbeat under microscope by aid of video camera gives and early diagnosis of mortality in short time. Therefore, the effects of measuring illumination, measuring time, and non-feeding during the test on heartbeat of water flea was evaluated to establish a new test approach. Test organisms used in this study are Daphnia magna, a well standardizes toxicity test organism, and Simocephalus mixtus, a newly refined organism $IC_{50}$ values of these test organism by heartbeat measurement were compared and discussed. It was found that toxicity test by heartbeat measurement was a reproducible, easy and simple method accomplished in a few hours.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.7
• /
• pp.643-649
• /
• 2013

In this paper, human heartbeat detection using signal reflected from the antenna which is varied reflection coefficient by near field variation of the antenna. For detection reflected signal from antenna, 20 dB directional coupler is used because of not affecting transmitting signal. Variance of reflection coefficient of dipole antenna is about 0.07 dB which is too small the distinction between heartbeat and noise. Sensitivity increasing method is applied and heartbeat is clearly detected. Due to phase and magnitude errors come from discrete value components, antenna is located some points in experiments ro find optimum sensitivity position. And providing verification of using communication signal, heartbeat detection when frequency modulated signal which have 4 MHz bandwidth is applied.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.18 no.4
• /
• pp.100-108
• /
• 2019

In this paper, a new stacked element pressure sensor has proposed for heartbeat and respiration measurement. This device can be directly attached to an individual's chest; heartbeat and respiration are detected by the pulsatile vibration and deformation of the chest. A key feature of the device is the simultaneous measurement of heart rate and respiration. The structure of the sensor consists of two stacked elements, in which one element includes one polyvinylidene fluoride (PVDF) thin film bonded on polydimethylsiloxane (PDMS) substrate. In addition, for the measurement and signal processing, the electric circuit and the filter are simply constructed with an OP-amp, resistance, and a capacitor. One element (element1, PDMS) maximizes the respiration signal; the other (element2, PVDF) is used to measure heartbeat. Element1 and element2 had sensitivity of 0.163V/N and 0.209V/N, respectively, and element2 showed improved characteristics compared with element1 in response to force. Thus, element1 and element2 were optimized for measuring respiration heart rate, respectively. Through mechanical and vivo human tests, this sensor shows the great potential to optimize the signals of heartbeat and respiration compared with commercial devices. Moreover, the proposed sensor is flexible, light weight, and low cost. All of these characteristics illustrate an effective piezoelectric pressure sensor for heartbeat and respiration measurements.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.12
• /
• pp.82-87
• /
• 2016

The photoplethysmogram (PPG) signal is one of the mainly considered bio signals along with the electrocardiogram (ECG) signal. PPG signals can be used to estimate the speed of flow of blood in vein, saturation of peripheral oxygen and etc. The heartbeat rate is a common feature in order to evaluate those checkup lists. To estimate the correct heartbeat rate, dynamic noises must be removed in the PPG signal. Conventionally, the acceleration signal is used to remove dynamic noises. This method, however, increases the computational complexity. In this paper, we proposes a solution that uses only PPG signals to calculate the heartbeat rate, and which can be used as a basement in real-time healthcare solution.