A switched nonlinear system is considered, and the interval between two consecutive switchings is assumed to be greater than a value called "the dwell time." When switching among nonlinear systems, using a constant dwell time generally fails to lead to stability. In this letter, a state dependent dwell time function with convergence guarantees is presented for discrete-time stable nonlinear systems.

The market for IP convergence services is expanding rapidly due to the rising number of Internet users. To respond to market trends, service systems must provide services quickly. This paper discusses that application server called the service agent which provides IP convergence services. The service agent meets the requirements for four application servers, centralized intelligence, supporting various interfaces: service creativity and scalability. The architecture is based on that of AIN systems, but whole system is written in Java especially to achieve service creativity and scalability. As a result of trial manufacture, feasibility of the service agent and scalability was achieved. Enough performance was also confirmed to obtain for commercial services.

In this paper we propose a parallel composition based adaptive notch filter for eliminating sinusoidal signals whose frequencies are unknown. The proposed filter which is implemented using second order all-pass filter and a band-pass filter can achieve high convergence speed by using the output of an additional band-pass filter to update the coefficients of the notch filter. The high convergence speed of the proposed notch filter is obtained by reducing an effect that an updating term of coefficient for adaptation of a notch filter significantly increases when the notch frequency approaches the sinusoidal frequency. In this paper, we analyze such effect obtained by the additional band-pass filter. We also present an analysis of a convergence performance of cascaded system of the proposed notch filter for eliminating multiple sinusoids. Simulation results have shown the effectiveness of the proposed adaptive notch filter.

Because most link-state routing protocols, such as OSPF and IS-IS, calculate routes using the Dijkstra algorithm, which poses scalability problems, implementors often introduce an artificial delay to reduce the number of route calculations. Although this delay directly affects IP packet forwarding, it can be acceptable when the network topology does not change often. However, when the topology of a network changes frequently, this delay can lead to a complete loss of IP reachability for the affected network prefixes during the unstable period. In this paper, we propose the Cached Shortest-path Tree (CST) approach, which speeds up intra-domain routing convergence without extra execution of the Dijkstra algorithm, even if the routing for a network is quite unstable. The basic idea of CST is to cache shortest-path trees (SPTs) of network topologies that appear frequently, and use these SPTs to instantly generate a routing table when the topology after a change matches one in the caches. CST depends on a characteristic that we found from an investigation of routing instability conducted on the WIDE Internet in Japan. That is, under unstable routing conditions, both frequently changing Link State Advertisements (LSAs) and their instances tend to be limited. At the end of this paper, we show CST's effectiveness by a trace-driven simulation.

The adaptive cross-spectral (ACS) technique recently introduced by Okuno et al. provides an attractive solution to acoustic echo cancellation (AEC) as it does not require double-talk (DT) detection. In this paper, we first introduce a generalized ACS (GACS) technique where a step-size parameter is used to control the magnitude of the incremental correction applied to the coefficient vector of the adaptive filter. Based on the study of the effects of the step-size on the GACS convergence behaviour, a new variable step-size ACS (VSS-ACS) algorithm is proposed, where the value of the step-size is commanded dynamically by a special finite state machine. Furthermore, the proposed algorithm has a new adaptation scheme to improve the initial convergence rate when the network connection is created. Experimental results show that the new VSS-ACS algorithm outperforms the original ACS in terms of a higher acoustic echo attenuation during DT periods and faster convergence rate.

A new beamforming method based on simplex downhill optimaization process has been presented for the reverse link CDMA systems. The proposed system performs code-filtering at each antenna for each user. The new beamforming method gives lower computations and faster convergence properties than existing algorithms. The simulation results show that the proposed algorithm has a better BER performance in the case of the time-varing channel.

The purpose of this paper is to discuss length estimation based on digitized curves. Information on a curve in the Euclidean plane is lost after digitization. Higher resolution supports a convergence of a digital image towards the original curve with respect to Hausdorff metric. No matter how high resolution is assumed, it is impossible to know the length of an original curve exactly. In image analysis we estimate the length of a curve in the Euclidean plane based on an approximation. An approximate polygon converges to the original curve with an increase of resolution. Several approximation methods have been proposed so far. This paper proposes a new approximation method which generates polygonal curves closer (in the sense of Hausdorff metric) in general to its original curves than any of the previously known methods and discusses its relevance for length estimation by proving a Convergence Theorem.

An LMS adaptive digital filter using distributed arithmetic (DA-ADF) has been proposed. Cowan and others proposed the DA adaptive algorithm with offset binary coding for the simple derivation of an algorithm and the use of an odd-symmetry property of adaptive function space (AFS). However, we indicated that a convergence speed of this DA adaptive algorithm degraded extremely by our computer simulations. To overcome these problems, we have proposed the DA adaptive algorithm generalized with two's complement representation and effective architectures. Our DA-ADF has performances of a high speed, small output latency, a good convergence speed, small-scale hardware and lower power dissipation for higher order, simultaneously. In this paper, we analyze a convergence condition of DA adaptive algorithm that has never been considered theoretically. From this analysis, we indicate that the convergence speed is depended on a distribution of eigenvalues of an auto-correlation matrix of an extended input signal vector . Furthermore, we obtain the eigenvalues theoretically. As a result, we clearly show that our DA-ADF has an advantage of the conventional DA-ADF in the convergence speed.

The electronically steerable passive array radiator (ESPAR) antenna is one kind of the parasitic elements based single-port output antennas with several variable reactances. It performs analog aerial beamforming and none of the signals on its passive elements can be observed. This fact and one that is more important--the nonlinear dependence of the output of the antenna from adjustable reactances--makes the problem substantially new and not resolvable by means of conventional adaptive array beamforming techniques. A novel approach based on stochastic approximation theory is proposed for the adaptive beamforming of the ESPAR antenna as a nonlinear spatial filter by variable parameters, thus forming both beam and nulls. Two learning rate schedule were examined about output SINR, stability, convergence, misadjustment, noise effect, bias term, etc., and the optimal one was proposed. Further development was traced. Our theoretic study, simulation results and performance analysis show that the ESPAR antenna can be controlled effectively, has strong potential for use in mobile terminals and seems to be very perspective.

The electronically steerable passive array radiator (ESPAR) antenna performs analog aerial beamforming that has only a single-port output and none of the signals on its passive elements can be observed. This fact and one that is more important--the highly nonlinear dependence of the output of the antenna from adjustable reactances--makes the problem substantially new and not resolvable by means of conventional adaptive array beamforming techniques. A novel approach based on stochastic approximation theory is proposed for the adaptive beamforming of the ESPAR antenna as a nonlinear spatial filter by variable parameters, thus forming both beam and nulls. Our theoretic study, simulation results and performance analysis show that the ESPAR antenna can be controlled effectively, has strong potential for use in mobile terminals and seems to be very perspective.

In order to maintain the diversity of structures in the population and prevent premature convergence, I have developed a new genetic algorithm called DCGA. In the experiments on many standard benchmark problems, DCGA showed good performances, whereas with harder problems, in some cases, the phenomena were observed that the search was stagnated at a local optimum despite that the diversity of the population is maintained. In this paper, I propose methods for escaping such phenomena and improving the performance by reinitializing the population, that is, a method called each-structure-based reinitializing method with a deterministic structure diverging procedure as a method for producing new structures and an adaptive improvement probability bound as a search termination criterion. The results of experiments demonstrate that DCGA becomes robust in harder problems by employing these proposed methods.

In this paper, a discrete-time convergence theorem for continuous-state Hopfield networks with self-interaction neurons is proposed. This theorem differs from the previous work by Wang in that the original updating rule is maintained while the network is still guaranteed to monotonically decrease to a stable state. The relationship between the parameters in a typical class of energy functions is also investigated, and consequently a "guided trial-and-error" technique is proposed to determine the parameter values. The third problem discussed in this paper is the post-processing of outputs, which turns out to be rather important even though it never attracts enough attention. The effectiveness of all the theorems and post-processing methods proposed in this paper is demonstrated by a large number of computer simulations on the assignment problem and the N-queen problem of different sizes.

The sub-recursive least squares (sub-RLS) algorithm estimates the coefficients of adaptive filter under the least squares (LS) criterion, however, does not require the calculation of inverse matrix. The sub-RLS algorithm, based on the different principle from the RLS algorithm, still provides a convergence property similar to that of the RLS algorithm. This paper first rewrites the convergence condition of the sub-RLS algorithm, and then proves that the convergence property of the sub-RLS algorithm successively approximates that of the RLS algorithm on the convergence condition.

The adaptive system design by 16-bit fixed point processing enables to employ an inexpensive digital signal processor (DSP). The narrow dynamic range of such 16 bits, however, does not guarantee the same performance that is confirmed beforehand by computer simulations. A cause of degrading the performance originates in the operation halving the word length doubled by multiplication. This operation rounds off small signals staying in the lower half of the doubled word length to zero. This problem can be solved by limiting the multiplier to only its sign () like the signed regressor algorithm, named 'bi-quantized-x' algorithm in this paper, for the convenience mentioned below. This paper first derives the equation describing the convergence property provided by a type of signed regressor algorithms, the bi-quantized-x normalized least mean square (NLMS) algorithm, and then formulates its convergence condition and the step size maximizing the convergence rate. This paper second presents a technique to improve the convergence property. The bi-qiantized-x NLMS algorithm quantizes the reference signal to 1 according to the sign of the reference signal, whereas the technique moreover assigns zero to the reference signal whose amplitude is less than a predetermined level. This paper explains the principle that the 'tri-qunatized-x' NLMS algorithm employing the technique can improve the convergence property, and confirms the improvement effect by computer simulations.

This letter focuses on the design of a unified estimator for scheduled control in nonlinear systems with unknown parameter. An estimation law with a finite convergence time is formulated to compute the unknown scheduling parameter that drives a scheduled controller. This estimator can also be extended to the types of scheduled controllers addressed in the literature.

The normalized least mean square (NLMS) algorithm has the drawback that the convergence speed of adaptive filter coefficients decreases when the reference signal has high auto-correlation. A technique to improve the convergence speed is to apply the decorrelated reference signal to the calculation of the gradient defined in the NLMS algorithm. So far, only the effect of the improvement is experimentally examined. The convergence property of the adaptive algorithm to which the technique is applied is not analized yet enough. This paper first defines a cost function properly representing the criterion to estimate the coefficients of adaptive filter. The name given in this paper to the adaptive algorithm exploiting the decorrelated reference signal, 'normalized least mean EE' algorithm, exactly expresses the criterion. This adaptive algorithm estimates the coefficients so as to minimize the product of E and E' that are the differences between the responses of the unknown system and the adaptive filter to the original and the decorrelated reference signals, respectively. By using the cost function, this paper second specifies the convergence condition of the normalized least mean EE' algorithm and finally presents computer simulations, which are calculated using real speech signal, to demonstrate the validity of the convergence condition.

It is well known that based on the structure of a transversal filter, the RLS equaliser provides the fastest convergence in stationary environments. This paper addresses an adaptive transversal equaliser which has the potential to provide more faster convergence than the RLS equaliser. A comparison is made with respect to computational complexity required for each update of equaliser coefficients, and computer simulations are demonstrated to show the superiority of the proposed equaliser.

The signed regressor algorithm, a variation of the least mean square (LMS) algorithm, is characterized by the estimation way of using the clipped reference signals, namely, its sign (). This clipping, equivalent to quantizing the reference signal to 1, only increases the estimation error by about 2 dB. This paper proposes to increase the number of the quantization steps to three, namely, 1 and 0, and shows that the 'tri-quantized-x' normalized least mean square (NLMS) algorithm with three quantization steps improves the convergence property.

Numerous scientific and engineering fields extensively utilize optimization techniques for finding appropriate parameter values of models. Various optimization methods are available for practical use. The optimization algorithms are classified primarily due to the rates of convergence. Unfortunately, it is often the case in practice that the particular optimization method with specified convergence rates performs substantially differently on diverse optimization tasks. Theoretical classification of convergence rates then lacks its relevance in the context of the practical optimization. It is therefore desirable to formulate a novel classification framework relevant to the theoretical concept of convergence rates as well as to the practical optimization. This article introduces such classification framework. The proposed classification framework enables specification of optimization techniques and optimization tasks. It also underlies its inherent relationship to the convergence rates. Novel classification framework is applied to categorizing the tasks of optimizing polynomials and the problem of training multilayer perceptron neural networks.

First order line seach optimization techniques gained essential practical importance over second order optimization techniques due to their computational simplicity and low memory requirements. The computational excess of second order methods becomes unbearable for large optimization tasks. The only applicable optimization techniques in such cases are variations of first order approaches. This article presents one such variation of first order line search optimization technique. The presented algorithm has substantially simplified a line search subproblem into a single step calculation of the appropriate value of step length. This remarkably simplifies the implementation and computational complexity of the line search subproblem and yet does not harm the stability of the method. The algorithm is theoretically proven convergent, with superlinear convergence rates, and exactly classified within the formerly proposed classification framework for first order optimization. Performance of the proposed algorithm is practically evaluated on five data sets and compared to the relevant standard first order optimization technique. The results indicate superior performance of the presented algorithm over the standard first order method.