• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.58-67
• /
• 2020

In this paper, we propose a data-driven-based beam selection scheme for massive multiple-input and multiple-output (MIMO) systems in ultra-dense networks (UDN), which is capable of addressing the problem of high computational cost of conventional coordinated beamforming approaches. We consider highly dense small-cell scenarios with more small cells than mobile stations, in the millimetre-wave band. The analog beam selection for hybrid beamforming is a key issue in realizing millimetre-wave UDN MIMO systems. To reduce the computation complexity for the analog beam selection, in this paper, two deep neural network models are used. The channel samples, channel gains, and radio frequency beamforming vectors between the access points and mobile stations are collected at the central/cloud unit that is connected to all the small-cell access points, and are used to train the networks. The proposed machine-learning-based scheme provides an approach for the effective implementation of massive MIMO system in UDN environment.

• Proceedings of the Korean Institute of IIIuminating and Electrical Installation Engineers Conference
• /
• 1999.11a
• /
• pp.256-261
• /
• 1999

Many control techniques have been proposed in order to improve the control performance of discrete-time domain control system. In the position control system using a DC servo motor as control system, the response-characteristic of system is controlled by the I-PD controller. In the I-PD longer if gains of I-PD controller are unsuitable. In this paper, therefore, a expanded I-PD control system is constructed by inserting a pre-compensator at out terminal of I-PD controller. It is implemented by neural network with two hidden layers. From the result of computer simulation in the proposed control algorithm, its usefulness is verified.

• Proceedings of the KIEE Conference
• /
• 1998.07b
• /
• pp.781-786
• /
• 1998

Since a gas turbine is still a significant contributor to peak time, it is very important to tune the gains of P. I. D to get a maximum power and stability within permissible limits. In the gas turbine, the main control loop must adjust the fuel flow to ensure the correct output power and frequency. but it is not easy, because the control loop is composed of many subsystems. In this paper we acquire a transfer function based on the operations data of Gun-san gas turbine and study to apply a loop compensation type 2-DOF PID controller tuning by neural-network to control loop of gas turbine to reduce phenomena caused by integral and derivative actions through simulation. We obtained satisfactory results to disturbances of subcontrol loop such as, fuel flow, air flow, turbine extraction temperature.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2000.08a
• /
• pp.105-108
• /
• 2000

Many control techniques have been proposed in order to improve the control performance of discrete-time domain control system. In the position control system using a DC servo motor as a driver, the response-characteristic of system is controlled by the PID controller. In the PID control system, the transient response characteristic is more increased and settling time gets longer if gains of PID controller are unsuitable. In this paper, therefore, a expended PID control system is constructed by inserting a pre-compensator at output terminal of PID controller. It is implemented by using the double layers neural network. Form the results of computer simulation in the proposed control algorithm, its usefulness is verified.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.38 no.8
• /
• pp.19-29
• /
• 2001

The recent explosion in the Internet applications, Internet. host number and the traffic in the IP backbone network is posing new challenges for transport network. This requires a high-speed IP backbone network that has a substantially higher bandwidth than the one offered by current networks, which prompts the development of all-optical network. To obtain optical network utilization gains, we need a software which establishs logical topology to make possible the efficient use or physical topology, and control the optical network in combination with the IP layer routing protocols. Finally, the logical topology is required higher efficient than physical topology. For this an efficient algorithm for the routing and wavelength assignment(RWA) in the WDM all-optical network is necessary. In this paper, two kinds of heuristic algorithms to establish logical topology for WDM networks and arc applied to the design of logical topology of domestic backbone network. These algorithms are found to work quite well and they arc compared with each other in terms of blocking rate, etc.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2000.04a
• /
• pp.56-63
• /
• 2000

Abstract - In this paper, an adaptive mechanism based on immune algorithm is designed and it is applied for the autonomous guided vehicle(AGV) driving. When the immune algorithm is applied to the PID controller, there exists the case that the plant is damaged due to the abrupt change of PID parameters since the parameters are adjusted almost randomly. To solve this problem, a neural network is used to model the plant and the parameter tuning of the model is performed by the immune algorithm. After the PID parameters are determined in this off-line manner, these gains are then applied to the plant for the on-line control using immune adaptive algorithm. Moreover, even though the neural network model may not be accurate enough intially, the weighting parameters are adjusted to be accurate through the on-line fine tuning. The computer simulation for the control of steering and speed of AGV is performed. The results show that the proposed controller has better performances than other conventional controllers.

• Proceedings of the Korea Contents Association Conference
• /
• 2004.11a
• /
• pp.275-279
• /
• 2004

In this paper, a large-scale, wide-area storage system based on IP is proposed which is under development. OASIS is a storage system enforced with very high-scalability up to hundreds and thousands of clients over IP network, and it is able to extend the service to the wide area network. For this purpose, we adopt an storage interconnection technology based on IP, the object based storage technology and a clustered server architecture which provides high-scalability and availability to our system. This system can be utilized to provide network storage service which gains more reality with the incoming FTTH and WiBro services.

• ETRI Journal
• /
• v.38 no.6
• /
• pp.1052-1063
• /
• 2016

We propose a millimeter wave (MMW)-based mobile hotspot network (MHN) system for application in high-speed railways that is capable of supporting a peak backhaul link throughput of 1 Gbps per train at 400 km/h. The MHN system can be implemented in subways and high-speed trains to support passengers with smart devices and provide access to the Internet. The proposed system can overcome the inherent high path loss in MMW through system designs and high antenna gains. We present a simulation of the system performance that shows that a fixed beamforming strategy can provide high signal-to-interference-plus-noise-ratio similar to those of an adaptive beamforming strategy, with the exception of 15% of the train path in which the network can use link adaptation with low-order modulation formats or trigger a handover to maintain the connection. We also demonstrate the feasibility of the MHN system using a test bed deployed in Seoul subway line 8. The backhaul link throughput varies instantaneously between 200 Mbps and 500 Mbps depending on the SNR variations while the train is running. During the field trial, the smartphones used could make connections through offloading.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.4
• /
• pp.201-212
• /
• 2000

In this paper, an adaptive mechanism based on immune algorithm is designed and it is applied for the autonomous guided vehicle(AGV) driving. When the immune algorithm is applied to the PID controller, there exists the cast that the plant is damaged due to the abrupt change of PID parameters since the parameters are adjusted almost randomly. To solve this problem, a neural network is used to model the plant and the parameter tuning of the model is performed by the immune algorithm. After the PID parameters are determined in this off-line manner, these gains are then applied to the plant for the on-line control using immune adaptive algorithm. Moreover, even though the neural network model may not be accurate enough intially, the weighting parameters are adjusted to be accurate through the on-line fine tuning. The computer simulation for the control of steering and speed of AGV is performed. The results show that the proposed controller has better performances than other conventional controllers.

• Journal of Korean Institute of Industrial Engineers
• /
• v.25 no.4
• /
• pp.500-509
• /
• 1999

In this paper, we consider an optimal routing problem when point-to-point and point-to-multipoint connection traffics are offered in an ATM network. We propose a mathematical model for cost-minimizing configuration of a logical network for a given ATM-based BISDN. Our model is essentially identical to the previous one proposed by Kim(Kim, 1996) which finds a virtual-path configuration where the relevant gains obtainable from the ATM technology such as the statistical multiplexing gain and the switching/control cost-saving gain are optimally traded-off. Unlike the Kim's model, however, ours explicitly considers the VP's QoS(Quality of Service) for more efficient utilization of bandwidth. The problem is a large-scale, nonlinear, and mixed-integer problem. The proposed algorithm is based on the local linearization of equivalent-capacity functions and the relaxation of link capacity constraints. As a result, the problem can be decomposed into moderate-sized shortest path problems, Steiner arborescence problems, and LPs. This fact renders our algorithm a lot faster than the previous nonlinear programming algorithm while the solution quality is maintained, hence application to large-scale network problems.