• Journal of Communications and Networks
• /
• 제9권3호
• /
• pp.254-264
• /
• 2007

A promising approach to improve the performance of mobile location system is the use of antenna arrays in both transmitter and receiver sides. Using advanced array signal processing techniques, such multiple-input multiple-output (MIMO) communication systems can offer more mobile location information by exploiting the spatial properties of the multipath channel. In this paper, we propose a novel approach to determine the position of mobile terminal based on estimated multipath signal parameters using only one base station in MIMO communication systems. This approach intends to minimize the error occurring from the estimation of multiple paths and gives an optimal estimation of the position of mobile terminal by simultaneously calculating a set of nonlinear location equations. This solution breaks the bottleneck of conventional mobile location systems which have to require multilateration of at least three base stations.

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.803-816
• /
• 2023

Explainable artificial intelligence is a method that explains how a complex model (e.g., a deep neural network) yields its output from a given input. Recently, graph-type data have been widely used in various fields, and diverse graph neural networks (GNNs) have been developed for graph-type data. However, methods to explain the behavior of GNNs have not been studied much, and only a limited understanding of GNNs is currently available. Therefore, in this paper, we propose an explanation method for node classification using graph convolutional networks (GCNs), which is a representative type of GNN. The proposed method finds out which features of each node have the greatest influence on the classification of that node using GCN. The proposed method identifies influential features by backtracking the layers of the GCN from the output layer to the input layer using the gradients. The experimental results on both synthetic and real datasets demonstrate that the proposed explanation method accurately identifies the features of each node that have the greatest influence on its classification.

• Journal of Power Electronics
• /
• 제17권3호
• /
• pp.821-834
• /
• 2017

This paper reports on a single-inductor multiple-output step-up converter with digital control. A systematic analog-to-digital-controller design is explained. The number of digital blocks in the feedback path of the proposed converter has been decreased. The simpler digital pulse-width modulation (DPWM) architecture is then utilized to reduce the power consumption. This architecture has several advantages because counters and a complex digital design are not required. An initially designed unit-delay cell is adopted recursively for the construction of coarse, intermediate, and fine delay blocks. A digital limiter is then designed to allow only useful code for the DPWM. The input voltage is 1.8 V, whereas output voltages are 2 V and 2.2 V. A co-simulation was also conducted utilizing PowerSim and Matlab/Simulink, whereby the 55 nm process was employed in the experimental results to evaluate the performance of the architecture.

• Journal of Communications and Networks
• /
• 제16권6호
• /
• pp.639-644
• /
• 2014

In low-density parity-check (LDPC) coded multiple-input multiple-output (MIMO) communication systems, probabilistic information are exchanged between an LDPC decoder and a MIMO detector. TheMIMO detector has to calculate probabilistic values for each bit which can be very complex. In [1], the authors presented a class of linear block codes named low-density MIMO codes (LDMC) which can reduce the complexity of MIMO detector. However, this code only supports the outer-iterations between the MIMO detector and decoder, but does not support the inner-iterations inside the LDPC decoder. In this paper, a new approach to construct LDMC codes is introduced. The new LDMC codes can be encoded efficiently at the transmitter side and support both of the inner-iterations and outer-iterations at the receiver side. Furthermore they can achieve the design rates and perform very well over MIMO channels.

• 한국정보전자통신기술학회논문지
• /
• 제4권1호
• /
• pp.9-16
• /
• 2011

본 논문은 다수의 비트레이트에서 변환하는 멀티레이트 변환 기능을 효율적으로 실현할 수 있는 비디오 트랜스코더를 제안하고자 한다. 이것을 구현하기 위해서는 부호 및 역 양자화기를 하나의 처리기로 공유해야 할 뿐만 아니라 유사한 재양자화 스텝값을 가지고 있는 출력 사이에 재양자화-재부호화 처리기를 공유시킴으로서 멀티레이트 변환시에 처리량을 감쇄시킬 수 있는 구조를 제안하고자 한다. 또, 이론적 평가 및 평가에 따른 실험 결과로부터 제안한 방식인 재양자화-재부호화기의 공유 방법이 출력 수에 대한 처리량을 효율적으로 감쇄시킴으로서 단일 레이트의 출력의 경우보다 약 3~6배의 처리량으로도 임의의 출력 수에 대응할 수 있는 결과를 얻을 수 있었다.

• 한국통신학회논문지
• /
• 제30권12A호
• /
• pp.1168-1173
• /
• 2005

본 논문에서는 군위성통신의 차량용단말에 사용되는 X-대역 TQTA용 HVPS를 설계 및 제작하고 그 성능을 측정하였다. 본 HVPS는 X-대역에서 600 W의 RF 신호 출력을 발생시키는 TWT를 구동시키기 위하여 제작되었다. HVPS의 스위칭 주파수는 80 kHz로 설계되었고 입력전원은 380 V의 직류전원을 사용한다. HVPS의 출력전압은 Cathode 전극에 -10.95 kV의 최대전압을 발생시키고 두개의 Collector 전극에 각각 -4.27 kV, -6.57 kV의 고압 전원을 제공하며 총 출력 전력은 1.6 kW를 발생시킨다. 한편, ripple 전압의 peak-to-peak 크기는 Cathode 전극의 경우 6 V, Collectorl은 12 V, Collector 2는 6 V로서 출력전압에 비해 각각 $0.055\%,\;0.281\%,\;0.091\%$ 수준으로 상당히 안정된 전원특성을 얻었다.

• 한국통신학회논문지
• /
• 제21권12호
• /
• pp.3215-3226
• /
• 1996

In this paper a K-band harmonic oscillator competitive to ordinary Push-Push type oscillators is introduced. This oscillator is composed of two-X-band dielectric resonance circuits. To favor its harmonic generation, the load effect and the bias effect are studied to allow the maximum harmonic distortion. As results, the dielectric resonated load and the class A bias are used for the 2nd harmonic generation. analytical study for modelling of voltage controlled dielectric resonator is carried out with theoretical background. The performance of the circuit is evaluated by simulation using harmonic balanced method. The novel structure has ont only a voltage tuning circuit but also an output port at fundamental frequency as the function of prescaler for phase lockede loop application on the just single oscillation structure. In experimentation, the output freqneyc of the 2nd harmonic signal is 20.5GHz and the maximum power level of output is +5.5dBm without additional post amplifiers. the harmonic oscillator exhibits -30dBc of high fundamental frequency rejection without added extra filters. The phase noise of -90dBc/Hz at 100kHz off-carrier has been achieved under free running condition, that satisfies phase noise requirement of IESS 308. The proposed oscillator may be utilized as the clean and stable fixed local oscillator in Transmit Block Upconvertor(TBU) or Low oise Block downconvertor(LNB) for K/Ka-band digital communications and satellite broadcastings.

• 한국산업정보학회논문지
• /
• 제6권4호
• /
• pp.48-53
• /
• 2001

TMT-2000 기지국용 3단 저잡음 증폭기를 설계하고 제작한다. 첫째 단에서의 증폭소자는 잡음특성이 좋은 GaAs HJ-FET를 사용하고, 둘째 및 셋째 단에는 이득과 출력전압이 높은 값을 갖도록 하기 위해 모노리딕(monolithic) 마이크로웨이브 집적회로를 사용한다. 또한 입력 정재파비를 낮추기 위해서 평형증폭기를 사용하는데, 이 평형증폭기의 위상차로 인한 잡음지수를 최소화하기 위해서 첫째 단에만 제한적으로 사용한다. 제작된 증폭기는 동작 주파수에서 이득 39.74$\pm$0.4dB, 최대잡음지수 0.97dB, 입.출력 정재파비 1.2 이하 및 OIP$_3$ 특성은 38.17dBm을 나타낸다.

• Journal of Power Electronics
• /
• 제17권6호
• /
• pp.1433-1444
• /
• 2017

A family of non-isolated DC-DC three-port converters (TPCs) that allows for a more flexible power flow among a renewable energy source, an energy storage device and a current-reversible DC bus is introduced. Most of the reported non-isolated topologies in this area consider only a power consuming load. However, for applications such as hybrid-electric vehicle braking systems and DC microgrids, the load power generating capability should also be considered. The proposed three-port family consists of one unidirectional port and two bi-directional ports. Hence, they are well-suited for photovoltaic (PV)-battery-DC bus systems from the power flow viewpoint. Three-port converters are derived by combining different commonly known power converters in an integrated manner while considering the voltage polarity, voltage levels among the ports and the overall voltage conversion ratio. The derived converter topologies are able to allow for seven different modes of operation among the sources and load. A three-port converter which integrates a boost converter with a buck converter is used as a design example. Extensions of these topologies by combining the soft-switching technique with the proposed design example are also presented. Experiment results are given to verify the proposed three-port converter family and its analysis.

• 대한간호학회지
• /
• 제25권1호
• /
• pp.88-98
• /
• 1995

The Fluid and electrolytes balance in the body is of critical importance in maintaining good health. When the fluid and electrolyte imbalance is present, patients are in great danger. They must be assessed immediately by a nurse so that appropriate treatment can be started as soon as possible. Patients' fluid intake and output records contain highly important information for the diagnosis and treatment of fluid imbalance, but, these records are often inaccurate and the method of recording the fluid intake is not universal for every hospital. Be-cause they are few quantitative measurements of a patient's hydration, the need to improve the accuracy of fluid intake records is very important. However, very few studies have been done to investigate the accuracy of measurements of patients' fluid intake and output. The purpose of this study was to investigate the methods used for calculation of fluid intake which is most similar to fluid output in normal adults and hospitalized patients. This study focused on three different calculation methods for fluid intake and compared these to fluid output and developed suggestions as to the ideal way to record fluid in-take. Data for 43 hospitalized patients and 37 normal adults were analyzed. The findings of this study are as follows ; 1) In normal adults, the daily intake of water which enteres by the oral route was 2415m1 (the first method of calculation). The daily intake of water in the form of pure water or some other beverage was 1365m1 (the third method of calculation) The daily intake of water including fresh fruits and vegetables, rice, porridges, and Me m which have water content more than 80% were 2186m1 (the second method of calculation). 2) The urine output of the normal adults was 1350m1. This apprroximates the amount of fluid an adult takes in the form of pure water. 3) In patient group, the total intake of water was 2550m1 (the first method of calculation). The in-take of water in the form of pure water or as some other beverage and IV fluid was 1661m1 (the third method of calculation). The daily in-take of water including foods which have high water content was 2356m1 (the second method of calculation). 4) The urine output of the patient's group was 1728m1. This approximates the amount of fluid an adult takes in the form of pure water. 5) Investigation of the method of calculation of the patient fluid intake showed that among the 31 hospitals studied, only eight use the third method of calculation which reflects the most close value to urine output. From the results obtained in this study, it was indicated that the amount of fluid taken in the form of pure water reflects the most close value to urine output. Therefore, it can be suggested that the third method of calculation which includes water in-take only in the form of pure water or beverage should be used as patients' fluid intake record.