• The KIPS Transactions:PartB
• /
• v.8B no.6
• /
• pp.699-704
• /
• 2001

ABSTRACT In this paper, a new hybrid modular recurrent TDNN (time-delay neural network)-HMM (hidden Markov model) architecture for speech recognition has been studied. In TDNN, the recognition rate could be increased if the signal window is extended. To obtain this effect in the neural network, a high-level memory generated through a feedback within the first hidden layer of the neural network unit has been used. To increase the ability to deal with the temporal structure of phonemic features, the input layer of the network has been divided into multiple states in time sequence and has feature detector for each states. To expand the network from small recognition task to the full speech recognition system, modular construction method has been also used. Furthermore, the neural network and HMM are integrated by feeding output vectors from the neural network to HMM, and a new parameter smoothing method which can be applied to this hybrid system has been suggested.

• Journal of KIISE:Databases
• /
• v.37 no.1
• /
• pp.1-11
• /
• 2010

Recently, there has been a growing interest in the processing of continuous queries over multiple data streams. When the arrival rates of tuples exceed the memory capacity of the system, a load shedding technique is used to avoid the system becoming overloaded by dropping some subset of input tuples. In this paper, we propose an effective load shedding algorithm for multi-way windowed joins over multiple data streams. Most previous load shedding algorithms estimate the productivity of each tuple, i.e., the number of join output tuples produced by the tuple, based on its "join attribute value" and drop tuples with the lowest productivity. However, the productivity of a tuple cannot be accurately estimated from its join attribute value when the join attribute values are unique and do not repeat, or the distribution of the join attribute values changes over time. For these cases, we estimate the productivity of a tuple based on its "arrival order" on data streams, rather than its join attribute value. The proposed method can effectively estimate the productivity of a tuple even when the productivity of a tuple cannot be accurately estimated from its join attribute value. Through extensive experiments and analysis, we show that our proposed method outperforms the previous methods in terms of effectiveness and efficiency.

• Journal of Technology Innovation
• /
• v.17 no.2
• /
• pp.131-158
• /
• 2009

We examine the effects of environmental or organizational factors on the performance of TLOs(technology transfer offices) in the PRIs(Public research institutes) using SFA(Stochastic Frontier Analysis), a technique for estimating the efficiency of DMUs(decision making units). In SFA, independent variables are assumed to determine the efficient production technique(production frontier) or affect the efficiency of DMUs. Previous researchs show that input variables such as number of personnel, R&D expenditure affect the production frontier while environmental or organizational variables affect the efficiency. We tried to estimate various types of models to find out whether environmental or organizational variables affect output variables differently from the previous research. Main empirical findings are as follows. First, R&D expenditure tends to increase all output variables considered. Second, environmental factors such as type of institutions and location of institutions affect the level of outputs. Third, organizational factors such as reward system for technology transfer also appear to affect the output variables. Fourth, environmental or organizational variables affect the production frontier directly rather than affect the efficiency of DMUs. Lastly, the efficiency of each DMU appear to be 1 or near to 1. Since almost all DMUs are equally efficient, it may not be effective to evaluate technology transfer activities of PRIs by efficiency criteria. We believe that this research should be complemented by additional data. More general types of production function need to be considered, and new techniques with concepts like output distance functions need to be developed to analyse multiple outputs simultaneously.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.20 no.9
• /
• pp.843-852
• /
• 2009

In this paper, we present design, fabrication, and measured results for a frequency upconverter for a wireless communication system of UAV(Unmanned Aerial Vehicle). The specifications of such wireless communication system requires the special features of maximum range of communication as well as deployment in UAV and repairing. The frequency upconverter operating at $5.25{\sim}5.45\;GHz$ in C-band was designed and fabricated considering such special features. The AGC function was included because the required output power should be constant for optimal system operation. The fabricated upconverter showed a constant output power of $+2{\pm}0.5\;dBm$ for the $-15{\sim}-10\;dBm$ input. Spuriouses were below -60 dBc and the adjacent leakage power was below -40 dBc. In addition, LO sources in the upconverter was implemented using the frequency synthesizer with step 1 MHz. This is for the application to the situation where multiple UAVs employed and the possible change of the permitted frequency band. The synthesizer showed a phase noise of -100 dBc/Hz at the 100 kHz frequency offset.

• Journal of Korea Society of Industrial Information Systems
• /
• v.9 no.2
• /
• pp.32-43
• /
• 2004

In this paper, a system which can effectively recognize the plate image extracted from camera set on a moving car is proposed. To extract car licence plate from moving vehicles, multiple candidates are maintained based on the strong vertical edges which are found in the region of car licence plate. A candidate region is selected among them based on the ratio of background and characters. We also make a comparative study of recognition performance between support vector machines and modular neural networks. The experimental results lead us to the conclusion that the former is superior to the latter. For a better recognition rate, a simple method combining the support vector machine with modular neural network where the output of the latter is used as the input of the former is suggested and evaluated. As we expected, the hybrid one shows the best result among those three methods we have mentioned.

• Journal of Advanced Navigation Technology
• /
• v.16 no.2
• /
• pp.351-359
• /
• 2012

This paper deals with the design and development of the remote control and measurement systems using CDMA(Code Division Multiple Access) data modem. We propose a bi-directional data communication link without the public IP address in CDMA modem device by the TCP/IP packet and SMS(Short Message Service) communication. The remote control and measurement systems are implemented by a Telit WM-800 modem as DCE(Data Communication Equipment), and Atmel AT89C51 microcontroller as DTE(Data Terminal Equipment). The user application software for the control and measurement system user, and the firmware software of device drivers for peripherals such as a digital input/output device, AD/DAC(Analog to Digital/Digital to Analog Converter), LCD, and temperature/humidity sensor are written in Microsoft C and Keil C language respectively for further various applications. The experimental result of the proposed control and measurement systems implemented in this paper is evaluated via real-time experiments, which works well as designed.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.3 no.4
• /
• pp.365-374
• /
• 1998

The SHE(Selected Harmonic Elimination) PWM scheme which eliminates specific lower order harmonics can generate h high quality output waveforms in 3-level PWM inverters. However. its application has limited since SHE switching a angles cannot be calculated on-line by a microprocessor-implemented control system. Based on off-line optimization. in which multiple SHE solutions were found and analysed for 2 to 5 switching angles per quarter in the 3-level SHE PWM pattern. this paper presents an algebraic algorithm for an ordinary microprocessor to calculate approximate SHE S switching angles on-line with such high resolution that it makes no practical difference between the accurate and the a approximate SHE switching angles. By employing the variable of the dc-link voltage Vdc' the proposed SHE PWM p pattern can ideally compensate the dc input fluctuation together with selected harmonics eliminated.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.7
• /
• pp.1397-1404
• /
• 2015

Recently, high data rate transmission is required in maritime communication. In this paper, we consider multiple input multiple output (MIMO) spatial multiplexing (SM). However, the performance of SM is severly degraded due to spatial channel correlation and line-of-sight (LOS) component. In the maritime communication, the MIMO channel correlation and LOS are critical due to the lack of scatteres around the transmitter and/or the receiver. When the feedback of channel information is available, precoding can enhance the error performance by exploiting the channel information. However, it is difficult to derive closed-form solution considering both the correlation and LOS. In this paper, we present open-loop precoding-based spatial multiplexing transmission method by showing that the effect of performance for the correlation and LOS. It is shown that the open-loop precoding can mitigate the performance degradation due to the LOS as well as the correlation. Consequently, we expect that the proposed open-loop precoding can be adopted to the maritime communication system.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.7
• /
• pp.1557-1564
• /
• 2014

In this paper, we propose a signal detection technique based on the parallel orthogonal matching pursuit (POMP) is proposed for generalized shift space keying (GSSK) systems, which is a modified version of the orthogonal matching pursuit (OMP) that is widely used as a greedy algorithm for sparse signal recovery. The signal recovery problem in the GSSK systems is similar to that in the compressed sensing (CS). In the proposed POMP technique, multiple indexes which have the maximum correlation between the received signal and the channel matrix are selected at the first iteration, while a single index is selected in the OMP algorithm. Finally, the index yielding the minimum residual between the received signal and the M recovered signals is selected as an estimate of the original transmitted signal. POMP with Quantization (POMP-Q) is also proposed, which combines the POMP technique with the signal quantization at each iteration. The proposed POMP technique induces the computational complexity M times, compared with the OMP, but the performance of the signal recovery significantly outperform the conventional OMP algorithm.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.6
• /
• pp.169-180
• /
• 2018

White light-emitting diodes (LEDs) are more economical than fluorescent lights, and provide high brightness, a high lifetime expectancy, and greater durability. As LEDs are closely connected with people's daily lives, dimming control of LED is an important component in providing energy savings and improving quality of life. In visible light communications systems using these LEDs, multiple input multiple output (MIMO) technology has attracted a lot of attention, in that it can attain the channel capacity in proportion to the number of antennas. This paper analyzes the power performance of three kinds of modulation in visible light communications (VLC) systems applied space-time block code (STBC) techniques. The modulation schemes are return-to-zero on-off keying (RZ-OOK), variable pulse position modulation (VPPM), and overlapping pulse position modulation (OPPM), and dimming control was applied. The power requirements and power consumption were used as metrics to compare the power efficiency in $2{\times}2$ STBC-VLC environments under the three kinds of modulation. We confirm that dimming control affects the communications performance of each modulation scheme. VPPM showed greater consumption among the three modulations, and OPPM showed energy savings comparable to VPPM.