• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.8
• /
• pp.15-26
• /
• 2011

In this paper, we present a new beamforming scheme for a downlink of multiuser multiple-input multipleoutput (MIMO) communication systems. Recently, a block-diagonalization (BD) algorithm has been proposed for the multiuser MIMO downlink where both a base station and each user have multiple antennas. However, the BD algorithm is not efficient when the number of supported streams per user is smaller than that of receive antennas. Since the BD method utilizes the space based on the channel matrix without considering the receive combining, the degree of freedom for beamforming cannot be fully exploited at the transmitter. In this paper, we optimize the receive beamforming vector under a zero forcing (ZF) constraint, where all inter-user interference is driven to zero. We propose an efficient algorithm to find the optimum receive vector by an iterative procedure. The proposed algorithm requires two phase values feedforward information for the receive combining vector. Also, we present another algorithm which needs only one phase value by using a decomposition of the complex general unitary matrix. Simulation results show that the proposed beamforming scheme outperforms the conventional BD algorithm in terms of error probability and obtains the diversity enhancement by utilizing the degree of freedom at the base station.

• Journal of Broadcast Engineering
• /
• v.25 no.2
• /
• pp.200-207
• /
• 2020

Recently, deep neural networks such as CNN are showing excellent performance in various fields such as image classification, object recognition, visual quality enhancement, etc. However, as the model size and computational complexity of deep learning models for most applications increases, it is hard to apply neural networks to IoT and mobile environments. Therefore, neural network compression algorithms for reducing the model size while keeping the performance have been being studied. In this paper, we apply few compression methods to CNN models and evaluate their performances in the embedded environment. For evaluate the performance, the classification performance and inference time of the original CNN models and the compressed CNN models on the image inputted by the camera are evaluated in the embedded board equipped with QCS605, which is a customized AI chip. In this paper, a few CNN models of MobileNetV2, ResNet50, and VGG-16 are compressed by applying the methods of pruning and matrix decomposition. The experimental results show that the compressed models give not only the model size reduction of 1.3~11.2 times at a classification performance loss of less than 2% compared to the original model, but also the inference time reduction of 1.2~2.21 times, and the memory reduction of 1.2~3.8 times in the embedded board.

• Journal of KIISE:Software and Applications
• /
• v.36 no.10
• /
• pp.777-785
• /
• 2009

Collaborative filtering is a technique used to predict whether a particular user will like a particular item. User-based or item-based collaborative techniques have been used extensively in many commercial recommender systems. In this paper, a hybrid collaborative filtering method that combines user-based and item-based methods using a low-dimensional linear model is proposed. The proposed method solves the problems of sparsity and a large database by using NMF among the low-dimensional linear models. In collaborative filtering systems the methods using the NMF are useful in expressing users as semantic relations. However, they are model-based methods and the process of computation is complex, so they can not recommend items dynamically. In order to complement the shortcomings, the proposed method clusters users into groups by using NMF and selects features of groups by using TF-IDF. Mutual information is then used to compute similarities between items. The proposed method clusters users into groups and extracts features of groups on offline and determines the most suitable group for an active user using the features of groups on online. Finally, the proposed method reduces the time required to classify an active user into a group and outperforms previous methods by combining user-based and item-based collaborative filtering methods.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.6
• /
• pp.69-78
• /
• 2017

Shannon of the 5G smartphone and Fourier of the signal processing meet in the sampling theorem (2 times the highest frequency 1). In this paper, the initial Shannon Theorem finds the Shannon capacity at the point-to-point, but the 5G shows on the Relay channel that the technology has evolved into Multi Point MIMO. Fourier transforms are signal processing with fixed parameters. We analyzed the performance by proposing a 2N-1 multivariate Fourier-Jacket transform in the multimedia age. In this study, the authors tackle this signal processing complexity issue by proposing a Jacket-based fast method for reducing the precoding/decoding complexity in terms of time computation. Jacket transforms have shown to find applications in signal processing and coding theory. Jacket transforms are defined to be $n{\times}n$ matrices $A=(a_{jk})$ over a field F with the property $AA^{\dot{+}}=nl_n$, where $A^{\dot{+}}$ is the transpose matrix of the element-wise inverse of A, that is, $A^{\dot{+}}=(a^{-1}_{kj})$, which generalise Hadamard transforms and centre weighted Hadamard transforms. In particular, exploiting the Jacket transform properties, the authors propose a new eigenvalue decomposition (EVD) method with application in precoding and decoding of distributive multi-input multi-output channels in relay-based DF cooperative wireless networks in which the transmission is based on using single-symbol decodable space-time block codes. The authors show that the proposed Jacket-based method of EVD has significant reduction in its computational time as compared to the conventional-based EVD method. Performance in terms of computational time reduction is evaluated quantitatively through mathematical analysis and numerical results.

• Geophysics and Geophysical Exploration
• /
• v.21 no.1
• /
• pp.15-25
• /
• 2018

Seismic inversion is a high-resolution tool to delineate the subsurface structures which may contain oil or gas. On the other hand, marine controlled-source electromagnetic (mCSEM) inversion can be a direct tool to indicate hydrocarbon. Thus, the joint inversion using both EM and seismic data together not only reduces the uncertainties but also takes advantage of both data simultaneously. In this paper, we have developed a simultaneous joint inversion approach for the direct estimation of reservoir petrophysical parameters, by linking electromagnetic and seismic data through rock physics model. A cross-gradient constraint is used to enhance the resolution of the inversion image and the maximum likelihood principle is applied to the relative weighting factor which controls the balance between two disparate data. By applying the developed algorithm to the synthetic model simulating the simplified gas field, we could confirm that the high-resolution images of petrophysical parameters can be obtained. However, from the other test using the synthetic model simulating an anticline reservoir, we noticed that the joint inversion produced different images depending on the model constraint used. Therefore, we modified the algorithm which has different model weighting matrix depending on the type of model parameters. Smoothness constraint and Marquardt-Levenberg constraint were applied to the water-saturation and porosity, respectively. When the improved algorithm is applied to the anticline model again, reliable porosity and water-saturation of reservoir were obtained. The inversion results indicate that the developed joint inversion algorithm can be contributed to the calculation of the accurate oil and gas reserves directly.

• Journal of the Korean Society of Radiology
• /
• v.3 no.1
• /
• pp.29-39
• /
• 2009

In this paper has proposed to the recognition of the disease on medical images using neural network. The neural network is constructed as three-layers of the input-layer, the hidden-layer and the output-layer. The training method applied for the recognition of disease region is adaptive error back-propagation. The low-frequency region analyzed by DWT are expressed by matrix. The coefficient-values of the characteristic polynomial applied are n+1. The normalized maximum value +1 and minimum value -1 in the range of tangent-sigmoid transfer function are applied to be use as the input vector of the neural network. To prove the validity of the proposed methods used in the experiment with a simulation experiment, the input medical image recognition rate the evaluation of areas of disease. As a result of the experiment, the characteristic polynomial coefficient of low-frequency area matrix, conversed to 4 level DWT, was proved to be optimum to be applied to the feature parameter. As for the number of training, it was marked fewest in 0.01 of learning coefficient and 0.95 of momentum, when the adaptive error back-propagation was learned by inputting standardized feature parameter into organized neural network. As to the training result when the learning coefficient was 0.01, and momentum was 0.95, it was 100% recognized in fifty-five times of the stomach image, fifty-five times of the chest image, forty-six times of the CT image, fifty-five times of ultrasonogram, and one hundred fifty-seven times of angiogram.

• Geophysics and Geophysical Exploration
• /
• v.27 no.2
• /
• pp.119-128
• /
• 2024

Marine magnetic surveys provide a rapid and cost-effective method for pioneer geophysical survey for many purposes. Sea-surface magnetometers offer high accuracy but are limited to measuring the scalar total magnetic field and require dedicated cruise missions. Shipboard three-component magnetometers, on the other hand, can collect vector three components and applicable to any cruise missions. However, correcting for the ship's magnetic field, particularly viscous magnetization, still remains a challenge. This study proposes a new additional correction method for ship's viscous magnetization effect in vector data acquired by shipboard three-component magnetometer. This method utilizes magnetic data collected simultaneously with a sea-surface magnetometer providing total magnetic field measurements. Our method significantly reduces deviations between the two datasets, resulting in corrected vector anomalies with errors as low as 7-25 nT. These tiny errors are possibly caused by the vector magnetic anomaly and its related viscous magnetization. This method is expected to significantly improve the accuracy of shipborne magnetic surveys by providing corrected vector components. This will enhance magnetic interpretations and might be useful for understanding plate tectonics, geological structures, hydrothermal deposits, and more.

• Journal of Broadcast Engineering
• /
• v.23 no.4
• /
• pp.519-525
• /
• 2018

This paper deals with noise suppression of multi-channel data captured by microphone array using multi-channel Wiener filter. Multi-channel Wiener filter does not rely on information about the direction of the target speech and can be partitioned into an MVDR (Minimum Variance Distortionless Response) spatial filter and a single channel spectral filter. The acoustic transfer function between the single speech source and microphones can be estimated by subspace decomposition of multi-channel Wiener filter. The errors are incurred in the estimation of the acoustic transfer function due to the errors in the estimation of correlation matrices, which in turn results in speech distortion in the MVDR filter. To alleviate the speech distortion in the MVDR filter, diagonal loading is applied. In the experiments, database with seven microphones was used and MFCC distance was measured to demonstrate the effectiveness of the diagonal loading.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.3
• /
• pp.1853-1862
• /
• 2015

Naval maintenance system carries out repairs of battle ships. Korean Navy has four maintenance stations to maximize the readiness of the battle ships. Since each station can provide different services according to characteristics(specific size of ships, type of maintenances) and the maintenance ability of stations is predetermined, it has been one of complex problems for the Korean Navy to find the optimal resource allocation. We investigate the operation of the stations from the perspective of the human resource allocation which plays crucial role in the performance of the maintenance stations. Using a queueing model and optimization technique, we present a way to derive the optimal personnel allocation which minimize the waiting number of battle ships at each station, leading to the improvement of the military readiness in the Korean Navy.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.5
• /
• pp.89-99
• /
• 2010

A unified method for separating musical sources, singing voice for example, from stereophonic mixtures is provided. We usually have two observed signals in stereophonic music contents, where more than two instruments are played together. If we regard each instrument as source, this problem becomes an underdetermined source separation problem and cannot be solved by conventional methods, which infers the spatial environment of the downmixing process happens. Instead, source-specific information has been exploited to recover a particular instrumental source. This paper provides a unifying structure consists of heterogenious ad-hoc separate algorithms, which are designed for separating vocal sources using stereophonic channel information and dominant pitch information of the sources, respectively. Experiments on real world music contents show that the proposed unification can neutralize the drawbacks of the two ad-hoc separation algorithms and finally enhance the separation results.