An adaptive blind signal separation filter is proposed using a risk-sensitive criterion framework. This criterion adopts an exponential type function. Hence, the proposed criterion varies the consideration weight of an adaptation quantity depending on errors in the estimates: the adaptation is accelerated when the estimation error is large, and unnecessary acceleration of the adaptation does not occur close to convergence. In addition, since the algorithm derivation process relates to an H filtering, the derived algorithm has robustness to perturbations or estimation errors. Hence, this method converges faster than conventional least squares methods. Such effectiveness of the new algorithm is demonstrated by simulation.

In this paper, we propose a medium access control (MAC) protocol for multimedia code division multiple access (CDMA) communications. In the proposed protocol, a base station (BS) estimates the instantaneous number of simultaneously transmitted packets in the future slots with exploiting a stochastic property of traffic. In order to carry out this estimation, we employ an adaptive algorithm. We evaluate the performance of the proposed protocol by comparing that with two different cases. One is no estimation case and the other is perfect estimation case. From these results, we clarify the advantage of the proposed MAC protocol.

This paper describes the outline of the active noise control system and the adaptive signal processing used in the practical systems. Focus is on the adaptive signal processing and algorithms which are widely used in many applications. Some variations in the algorithms for improving the control effect and for reducing the amount of calculation are also shown. Additionally, the limitations and some design guide are shown with the results of the numerical simulations.

Constant Modulus Algorithm (CMA) is a method that has been widely known as blind adaptive beamforming because it requires no knowledge about the signal except that the transmitted signal waveform has a constant envelope. Although CMA has the merit of this blind operation, it possesses problems in its convergence property. In this paper, problems that are inherent to this algorithm is resolved using a combination of CMA and another major adaptive algorithm SMI (Sample Matrix Inversion). The idea is to use SMI to determine the initial weights for CMA operation. Although the benefit of CMA being a blind algorithm is not fully taken advantage of, good aspects of both SMI and CMA can be introduced. By using this approach, two major problems in convergence properties of CMA can be solved. One of these problems is the reliability and this relates to the convergence performance in certain cases. When the interfering signal is stronger than the desired signal, the algorithm tends to come up with the wrong solution by capturing the interfering signal which has the stronger power. Also, the convergence time of this algorithm is slow, limiting its application in dynamic environment, although the slow convergence time of CMA has been studied previously and several methods have been proposed to overcome this defect. Using the proposed method, the deterioration due to both of these problems can be mitigated. Simulation results are shown to confirm the theory. Furthermore, evaluations are done concerning the fading characteristics. It is also confirmed from the simulation that the tracking performance of this method can be regarded as sufficient in personal mobile communication.

This paper describes a unique motion estimation and compensation (ME/MC) hardware architecture for a scene-adaptive algorithm. By statistically analyzing the characteristics of the scene being encoded and controlling the encoding parameters according to the scene, the quality of the decoded image can be enhanced. The most significant feature of the architecture is that the two modules for ME/MC can work independently. Since a time interval can be inserted between the operations of the two modules, a scene-adaptive algorithm can be implemented in the architecture. The ME/MC architecture is loaded on a single-chip MPEG-2 video encoder.

The multiple access causes an interference problem in the direct-sequence code-division multiple access systems. An efficient adaptive algorithm should be used to suppress this interference for the improvement of system performance. In this paper, the new blind adaptive method is suggested using the constant modulus algorithm for the purpose of interference suppression. Simulation results show that the converged value of signal to interference ratio for the proposed method is approximately 6 [dB] larger than that of a conventional Blind-MOE receiver in the additive white Gaussian noise channel and channel with inter-symbol interference while the signal to interference ratio improvement is almost 4 [dB] better in the Rayleigh fading channel. The suggested method is also robust to the new user interference resulting the nearly 3 [dB] improvement of the SIR value comparing with the conventional receiver. Based on these results, it is shown that the BER of the proposed receiver is lower than that of any other conventional receiver. Therefore, using the newly suggested method, the considerable performance improvement can be obtained for the DS-CDMA systems.

In this paper, we discuss a neural network approach for blind signal extraction of temporally correlated sources. Assuming autoregressive models of source signals, we propose a very simple neural network model and an efficient on-line adaptive algorithm that extract, from linear mixtures, a temporally correlated source with an arbitrary distribution, including a colored Gaussian source and a source with extremely low value (or even zero) of kurtosis. We then combine these extraction processing units with deflation processing units to extract such sources sequentially in a cascade fashion. Theory and simulations show that the proposed neural network successfully extracts all arbitrarily distributed, but temporally correlated source signals from linear mixtures.

Since most digital phase-locked loops (DPLLs) used in digital data transmission receivers require both fast acquisition of input frequency and phase in the beginning and substantial jitter reduction in the steady-state, the DPLL loop bandwidth is preferred to being adjusted accordingly. In this paper, a bandwidth adjusting (adaptive) algorithm is presented, which allow both fast acquisition and significant jitter reduction for each different noise environment and hardware requirement. This algorithm, based on the recursive least squares (RLS) criterion, suggest an optimal sequence of control parameters for a dual-loop DPLL which achieves the fastest initial acquisition time by trying to minimize the jitter variance in any given time instant. The algorithm can be used for carrier recovery or clock recovery in mobile communications, local area networks and disk drivers that require a short initial preamble period.

We present a cascade neural network for blind source extraction. We propose a family of unconstrained optimization criteria, from which we derive a learning rule that can extract a single source signal from a linear mixture of source signals. To prevent the newly extracted source signal from being extracted again in the next processing unit, we propose another unconstrained optimization criterion that uses knowledge of this signal. From this criterion, we then derive a learning rule that deflates from the mixture the newly extracted signal. By virtue of blind extraction and deflation processing, the presented cascade neural network can cope with a practical case where the number of mixed signals is equal to or larger than the number of sources, with the number of sources not known in advance. We prove analytically that the proposed criteria both for blind extraction and deflation processing have no spurious equilibria. In addition, the proposed criteria do not require whitening of mixed signals. We also demonstrate the validity and performance of the presented neural network by computer simulation experiments.

In this paper, we propose a new structure of the frequency domain adaptive filter (FDAF). The proposed structure is based on the modified DFT pair which consists of the FIR filters, so that un-delayed output signal can be obtained with stable convergence and without accumulated error which are problems for the conventional FDAFs. The convergence performance of the proposed FDAF is examined through the computer simulations in the adaptive line enhancer (ALE) comparing with the conventional FDAF and the DCT domain adaptive filter. Furthermore, in order to improve the error performance of the FDAF, we propose a composite algorithm which consists of the normalized step size algorithm for fast convergence and the variable step size one for small estimation error. The advantage of the proposed algorithm is also confirmed through simulations in the ALE. Finally, we propose a reduction method of the computational complexity of the proposed FDAF. The proposed method is to utilize a part of the FFT flow-graph, so that the computational complexity is reduced to O(N log N).

An adaptive infinite impulse response (IIR) filter implemented using an allpass and a minimum phase system has an advantage of its poles converging to the poles of the unknown system when the input is a white signal. However, when the input signal is colored, convergence speed deteriorates considerably, even to the point of lack of convergence for certain colored signals. Furthermore with a colored input signal, there is no guarantee that the poles of the adaptive digital filter (ADF) will converge to the poles of the unknown system. In this paper we propose a method which uses a linear predictor filter to whiten the input signal so as to improve the convergence characteristic. Computer simulation results confirm the increase in convergence speed and the convergence of the poles of the ADF to the poles of the unknown system even when the input is a colored signal.

In this paper we develop a new family of on-line adaptive learning algorithms for blind separation of time delayed and convolved sources. The algorithms are derived for feedforward and fully connected feedback (recurrent) neural networks on basis of modified natural gradient approach. The proposed algorithms can be considered as generalization and extension of existing algorithms for instantaneous mixture of unknown source signals. Preliminary computer simulations confirm validity and high performance of the proposed algorithms.

A noble blind equalization algorithm (BEA) using a single/multilevel modulus is proposed. According to the residual intersymbol interference (ISI) level of the equalizer output, the new algorithm adopts relevantly a single modulus or a multilevel modulus to form its cost function. Moreover, since the proposed approach separates complex two-dimensional signal into in-phase and quadrature components, and forms the error signals for each component, it has inherently the capability of phase recovery. Hence, it improves the performances of steady-state and recovers the phase rotation without any degradation of transient property. Simulation results confirm the effectiveness of the new approach.

In the adaptive lattice estimation process, it is well known that the convergence speed of the successive stage is affected by the estimation errors of reflection coefficients in its preceding stages. In this paper, we propose block estimation methods of two-dimensional (2-D) adaptive lattice filter. The convergence speed of the proposed algorithm is significantly enhanced by improving the adaptive performance of preceding stages. Furthermore, this process can be simply realized. The modeling of 2-D AR field and texture image are demonstrated through computer simulations.

An adaptive algorithm for a single sinusoid detection using IIR bandpass filter with parallel block structure has been proposed by Nishimura et al. However, the algorithm has three problems: First, it has several input frequencies being impossible to converge. Secondly, the convergence rate can not be higher than that of the scalar structure. Finally, it has a large amount of computation. In this paper, a new algorithm is proposed to solve these problems. In addition, a new structure is proposed to reduce the amount of computation, in which the adaptive control signal generator is realized by the paralel block structure. Simulation results are given to illustrate the performance of the proposed algorithm.

This paper relates to a novel algorithm for fast estimation of the coefficients of the adaptive FIR filter. The novel algorithm is derived from a first order IIR filter experssion clarifying the estimation process of the NLMS (normalized least mean square) algorithm. The expression shows that the estimation process is equivalent to a procedure extracting the cross-correlation coefficient between the input and the output of an unknown system to be estimated. The interpretation allows to move a subtraction of the echo replica beyond the IIR filter, and the movement gives a construction with the IIR filter coefficient of unity which forms the arithmetic mean. The construction in comparison with the conventional NLMS algorithm, improves the covergence rate extreamly. Moreover, when we use the construction with a simple technique which limits the term of calculating the correlation coefficient in the beginning of a convergence process, the convergence delay becomes negligible. This is a very desirable performance for acoustic echo canceller. In this paper, double-talk and echo path fluctuation are also studied as the first stage for application to acoustic echo canceller. The two subjects can be resolved by introducing two switches and delays into the evaluation process of the correlation coefficient.

This paper proposes a new adaptive algorithm for acoustic echo cancellers with four times the convergence speed for a speech input, at almost the same computational load, of the normalized LMS (NLMS). This algorithm reflects both the statistics of the variation of a room impulse response and the whitening of the received input signal. This algorithm, called the ESP (exponentially weighted step-size projection) algorithm, uses a different step size for each coefficient of an adaptive transversal filter. These step sizes are time-invariant and weighted proportional to the expected variation of a room impulse response. As a result, the algorithm adjusts coefficients with large errors in large steps, and coefficients with small errors in small steps. The algorithm is based on the fact that the expected variation of a room impulse response becomes progressively smaller along the series by the same exponential ratio as the impulse response energy decay. This algorithm also reflects the whitening of the received input signal, i.e., it removes the correlation between consecutive received input vectors. This process is effective for speech, which has a highly non-white spectrum. A geometric interpretation of the proposed algorithm is derived and the convergence condition is proved. A fast profection algorithm is introduced to reduce the computational complexity and modified for a practical multiple DSP structure so that it requires almost the same computational load, 2L multiply-add operations, as the conventional NLMS. The algorithm is implemented in an acoustic echo canceller constructed with multiple DSP chips, and its fast convergence is demonstrated.