• Journal of Communications and Networks
• /
• v.9 no.2
• /
• pp.112-117
• /
• 2007

In this paper, we focus on the total transmission power minimization problem for downlink beamforming multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems while ensuring each user's QoS requirement. Although the linear integer programming (LIP) solution we formulate provides the performance upper bound of the margin adaptive (MA) optimization problem, it is hard to be implemented in practice due to its high computational complexity. By regarding each user's equivalent channel gain as approximate independent values and using iterative descent method, we present a heuristic MA resource allocation algorithm. Simulation results show that the proposed algorithm efficiently converges to the local optimum, which is very close to the performance of the optimal LIP solution. Compared with existing space division multiple access (SDMA) OFDM systems with or without adaptive resource allocation, the proposed algorithm achieves significant performance improvement by exploiting the frequency diversity and multi-user diversity in downlink multiple-input single-output (MISO) OFDM systems.

• ETRI Journal
• /
• v.20 no.3
• /
• pp.251-271
• /
• 1998

It is well known that if usage parameter control/network parameter control (UPC/NPC) functions are used together with a cell loss priority control scheme in ATM networks, the measurement phasing problem can occur. This makes it difficult for a network provider to define and commit the cell loss ratio as a QoS parameter. To solve the problem, we propose a new UPC/NPC algorithm. By using the proposed UPC/NPC algorithm, we can define the cell loss ratios for CLP = 0 and CLP = 0+1 cell streams without the measurement phasing problem under any conditions. We analyzed the performance of the proposed UPC/NPC algorithm. Using a discrete time model for the UPC/NPC architecture with a discrete-time semi-Markov process (DSMP) input model, we obtained the cell discarding probabilities of CLP = 0 and CLP = 0+1 cells streams and showed that more CLP = 0 cells are accepted compared to what was proposed in ITU-T.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.2
• /
• pp.477-483
• /
• 2004

One of the major paradigms for supervised learning in neural network community is back-propagation learning. The standard implementations of back-propagation learning are optimal under the assumptions of identical and independent Gaussian noise. In this paper, for regression function estimation, we introduce $\varepsilon-insensitive$ back-propagation learning algorithm, which corresponds to minimizing the least absolute error. We compare this algorithm with support vector machine(SVM), which is another $\varepsilon-insensitive$ supervised learning algorithm and has been very successful in pattern recognition and function estimation problems. For comparison, we consider a more realistic model would allow the noise variance itself to depend on the input variables.

• The Journal of Information Technology
• /
• v.9 no.4
• /
• pp.65-74
• /
• 2006

The back propagation of neural networks has the problems of falling into local minimum and delay of the speed by the iterative learning. An algorithm to solve the problem and improve the speed of the learning was already proposed in[8], which updates the learning parameter related with the connection weight. In this paper, we propose the algorithm generating initial weight to improve the efficiency of the algorithm by offering the difference between the input vector and the target signal to the generating function of initial weight. The algorithm proposed here can classify more than 98.75% of the handwritten digits and this rate shows 30% more effective than the other previous methods.

• Journal of the Korean Data and Information Science Society
• /
• v.8 no.2
• /
• pp.163-171
• /
• 1997

Function approximation from a set of input-output pairs has numerous applications in scientific and engineering areas. Multilayer feedforward neural networks have been proposed as a good approximator of nonlinear function. The back propagation(BP) algorithm allows multilayer feedforward neural networks to learn input-output mappings from training samples. It iteratively adjusts the network parameters(weights) to minimize the sum of squared approximation errors using a gradient descent technique. However, the mapping acquired through the BP algorithm may be corrupt when errorneous training data we employed. When errorneous traning data are employed, the learned mapping can oscillate badly between data points. In this paper we propose a robust BP learning algorithm that is resistant to the errorneous data and is capable of rejecting gross errors during the approximation process, that is stable under small noise perturbation and robust against gross errors.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.4
• /
• pp.1954-1971
• /
• 2017

In this work, first, a multiuser detection (MUD) algorithm based on component-level soft interference cancellation and linear minimum mean square error (CLSIC-LMMSE) is proposed, which can enhance the bit error ratio (BER) performance of the traditional SIC-LMMSE-based MUD by mitigating error propagation. Second, for non-binary low density parity check (NB-LDPC) coded high-order modulation systems, when the proposed algorithm is integrated with partial mapping, the receiver with iterative detection and decoding (IDD) achieves not only better BER performance but also significantly computational complexity reduction over the traditional SIC-LMMSE-based IDD scheme. Extrinsic information transfer chart (EXIT) analysis and numerical simulations are both used to support the conclusions.

• Proceedings of the IEEK Conference
• /
• 2000.11d
• /
• pp.243-246
• /
• 2000

Recently, PC monitor users have been replacing cathode ray tubes (CRT) with liquid crystal displays (LCD). But the chromaticity of the primaries are dependent on RGB input signals. And the colorimetry of LCD changes with gray scale and has a poor peformance in color reproduction. In this paper we propose the enhanced algorithm of color reproduction considering color leakage error and black subpixel error in LCD. In order to test peformance of this algorithm we use the colors of Macbeth colorcheck. As a result of experiments, it was confirmed that the color difference of the LCD using the proposed algorithm was considerably reduced.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.29B no.2
• /
• pp.1-8
• /
• 1992

This paper is proposed a Modified WLD (Weighted Levenshtein Distance) algorithm for processor desihn of code-string matching. A proposed MWLD (Modified Weighted Levenshtein Distance) algorithm is consist of 1-dimension bit-serial array processor to pattern matching using a Hamming Distance. The proposed processor is applied to recognition of character with real time input. The recognition rate of Hangul strokes is resulted to 98.65$\%$

• Proceedings of the IEEK Conference
• /
• 2000.11b
• /
• pp.75-78
• /
• 2000

This paper presents a new algorithm for allocating the data path to achieve the minimum power consumption under the constraints of minimum hardware resources. In order to minimize the power consumption, the proposed algorithm tries to minimize the input transitions of functional units, unnecessary computations, and size of interconnects in a greedy manner during a]location. Experimental results using benchmarks indicate the proposed algorithm achieves 17.5% power reduction on average, when compared with the genesis-lp［1］high-level synthesis system.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.579-582
• /
• 1999

In this paper, we propose a human face detection algorithm using adaptive skin color model and neural networks. To attain robustness in the changes of illumination and variability of human skin color, we perform a color segmentation of input image by thresholding adaptively in modified hue-saturation color space (TSV). In order to distinguish faces from other segmented objects, we calculate invariant moments for each face candidate and use the multilayer perceptron neural network of backpropagation algorithm. The simulation results show superior performance for a variety of poses and relatively complex backgrounds, when compared to other existing algorithm.