• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.4C
• /
• pp.232-240
• /
• 2011

While the water-filling algorithm is an efficient power allocation method that maximizes the ergodic capacity of adaptive MIMO systems, its excessive residual power causes spectrum loss in real systems employing discrete modulation indices. In this paper we propose new power allocation algorithms that improve the spectral efficiency of MIMO systems by efficiently reallocating the residual power of the water-filling algorithm. We apply the proposed algorithms to the adaptive turbo-coded MIMO system to verify their performance through computer simulation in various environments. Simulation results show that the spectral efficiency of the proposed algorithms is better than that of the water-filling algorithm by about 8.9% at SNR of 20dB in Rayleigh fading environments.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.2
• /
• pp.90-94
• /
• 2008

It is well-blown that the convergence rate gets worse when an input signal to an adaptive filter is correlated. In this paper we propose a new adaptive filtering algorithm that makes the convergence rate much improved even for highly correlated input signals. By introducing an orthogonal constraint between successive input signal vectors we overcome the slow convergence problem of the LMS algorithm with the correlated input signal. Simulation results show that the proposed algerian yields fast convergence speed and excellent tracking capability under both time-invariant and time-varying environments, while keeping both computation and implementation simple.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.04a
• /
• pp.437-440
• /
• 2011

본 논문은 이동식 자동영상속도측정기로 과속 단속된 영상자료 중 역광 원인으로 자동차 번호판을 인식할 수 없어 폐기되는 영상자료에 대한 번호판 인식률을 향상시키는 알고리즘을 제안하였다. 명암값 분포가 불규칙한 자동차 번호판 이미지나 영상 자체에 손상이 많은 자동차 번호판 이미지를 지역적 적응 이진화 알고리즘을 사용함으로써, 오츠 전역적 이진화 알고리즘보다 뛰어난 자동차 번호판 인식률을 얻었다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.9C
• /
• pp.922-928
• /
• 2007

This paper presents the fast adaptive algorithm which applies an approximation scheme into RLS algorithm. The proposed algorithm(D-RLS) derives a QRD RLS algorithm derivation process from RLS algorithm recursively. D-RLS has the similar pattern as the algorithm having the approximation that input signals are separated respectively. Computational complexity of D-RLS is O(N), fewer than $O(N^2)$. To evaluate performance of proposed algorithm, we use the system identification method of FIR and Volterra system. And, finally, we can show D-RLS has an excellent performance.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.11
• /
• pp.1714-1725
• /
• 1993

This paper analyzes the properties of such algorithm that corresponds to the nonlinear adaptive algorithm with additional update terms, parameterized by the scalar factors a1, a2, a3 and Presents its structure. The analysis of convergence leads to eigenvalues of the transition matrix for the mean weight vector. Regions in which the algorithm becomes stable are demonstrated. The time constant is derived and the computational complexities of MLMS algorithms are compared with those of the conventional LMS, sign, LFG, and QFG algorithms. The properties of convergence in the mean square are analyzed and the expressions of the mean square recursion and the excess mean square error are derived. The necessary condition for the CFG algorithm to be stable is attained. In the computer simulation applied to the system identification the CFG algorithm has the more computation complexities but the faster convergence speed than LMS, LFG and QFG algorithms.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.1
• /
• pp.18-23
• /
• 2010

The RLS (Recursive Least Squares) method is a broadly used adaptive algorithm for signal processing in electronic engineering. The RLS algorithm shows a good performance and a fast adaptation within a stationary environment, but it shows a Poor performance within a non-stationary environment because the method has a fixed forgetting factor. In order to enhance 'tracking' performances, BLS methods with an adaptive forgetting factor had been developed. This method shows a good tracking performance, however, it suffers from heavy computational loads. Therefore, we propose a modified AFF-RLS which has relatively low complexity m this paper.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.110-113
• /
• 2019

본 논문에서는 임베디드 시스템에서의 양자화 기계학습을 수행할 경우 발생하는 양자화 오차를 효과적으로 보상하기 위한 방법론을 제안한다. 경사 도함수(Gradient)를 사용하는 기계학습이나 비선형 신호처리 알고리즘에서 양자화 오차는 경사 도함수의 조기 소산(Early Vanishing Gradient)을 야기하여 전체적인 알고리즘의 성능 하락을 가져온다. 이를 보상하기 위하여 경사 도함수의 최대 성분에 대하여 직교하는 방향의 보상 탐색 벡터를 유도하여 양자화 오차로 인한 성능 하락을 보상하도록 한다. 또한, 기존의 고정 학습률 대신, 내부 순환(Inner Loop) 없는 비선형 최적화 알고리즘에 기반한 적응형 학습률 결정 알고리즘을 제안한다. 실험결과 제안한 방식의 알고리즘을 비선형 최적화 문제에 적용할 시 양자화 오차로 인한 성능 하락을 최소화시킬 수 있음을 확인하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06a
• /
• pp.130-133
• /
• 2010

본 논문에서는 유전 알고리즘의 성능 향상을 위해 교배(Crossover) 기법의 중요 매개변수인 교배 교차점(Crossover Point)의 수를 개체군(Population)의 진화 과정 중에 적응적으로 변화 할 수 있는 자기-적응형(Self-Adaptive) 교배 기법을 제안한다. 이를 위해 제안 교배 기법은 전체 개체군을 다수개의 작은 개체군들로 군집화(Grouping)하여 일차적으로 서로 다른 교차점을 갖는 교배 기법을 적용시키고, 그 후 각 군집의 개체(Individual)들의 선택률을 기반으로 군집들간의 경쟁을 수행한다. 이는 유전 알고리즘이 개체군의 진화 과정 중에 문제에 적합한 교차점을 갖는 교배 기법을 적응적으로 사용할 수 있도록 한다. 또한 제안 교배 기법은 진화 과정 중에 교차점이 지속적으로 변화되므로 알고리즘 초반에는 높은 탐색 능력을 보유하게 되고 후반에는 높은 부분-해(Building-Block) 보존 능력을 지니게 되어, 최적 해(Optimal Solution)로의 수렴 능력이 향상된다. Deceptive 문제를 통해 제안 자기-적응형 교배 기법과 기존 (고정 교차점) 교배 기법의 성능을 비교 하였으며, 실험 결과로부터 제안 교배의 성능 우위를 확인하였다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.2
• /
• pp.67-72
• /
• 2021

This paper compare the adative equalization performance of the DSE-MMA (Dithered Signed-Error MMA) and QE-MMA (Quantized-Error MMA) which has a simplifies the computational operation of currently used MMA algorithm. The DSE-MMA possible to improve the rubustness to noise by using the dithered signal consider the polarity of error signal in the multiplication part. In QE-MMA, it use the polarity of error signal after performing the nonlinear power-of-two quantizing operation for easiness of H/W implementation. The same channel environment was applied, and it's performance of the output signal constellation, the residual isi and maximum distortion and MSE that means the convergence characteristics, the SER that means the robustness of external noise of algorithm were compared and evaluated. As a result of computer simulation, the QE-MMA has more good in constellation, residual isi, maximum distortion, MSE performanc than DSE-MMA. In SER, the DSE-MMA has more robust due to dither signal than QE-MMA.

• Journal of Korea Multimedia Society
• /
• v.2 no.1
• /
• pp.99-108
• /
• 1999

In this paper, a new robust adaptive algorithm is proposed to improve the performance of AEC without computational burden. The proposed adaptive algorithm is based on NLMS algorithm, and its step-size is varied with the reference input signal power and the desired signal power. Its step-size is normalized by the sum of the powers of the reference input signal and the desired signal. When the near-end speaker's speech and noise are applied into the microphone, the step-size becomes small and the misalignment of coefficients are reduced. The convergence speed is comparable to NLMS algorithm at AEC application because the echo signals are attenuated about 10∼20 dBSPL. The characteristics of this algorithm is also analyzed and compared with conventional ones in this paper.