• The Journal of the Acoustical Society of Korea
• /
• v.18 no.2
• /
• pp.61-65
• /
• 1999

This paper derives a sufficient condition on the stability of recursive third-order Volterra filters based on their filter coefficients. A Volterra filter is very effective in modeling nonlinear systems with memory. However, it is well-known that the nonrecursive Volterra filter requires a large number of filter coefficients to describe a nonlinear system. For this reason, recursive Volterra filters are usually considered because the recursive implementation requires a smaller number of coefficients compared to the nonrecursive one. Unfortunately, the main problem of the recursive Volterra filters is their inherent instablility. In this paper. we present a simple condition for the output of a recursive discrete-time third-order Volterra filter to be bounded whenever the input signal to the recursive Volterra filter is bounded by a finite constant.

• Journal of Industrial Technology
• /
• v.14
• /
• pp.3-17
• /
• 1994

This paper develops a design technique for approximating nonseparable frequency characteristics by sums and products of separable transfer functions. Therefore nonseparable frequency characteristics includes the four-quadrant symmetry filters. The desired filter with half plan symmetry is obtained by shifting a low pass characteristic in the frequency domain, and by combining these shifted characteristics. Also the paper develops the technique for designing recursive and nonrecursive two dimensional digital filters by the application of a complex transformation to one dimensional low pass filter.

• Journal of the Institute of Convergence Signal Processing
• /
• v.5 no.2
• /
• pp.143-150
• /
• 2004

In this paper, a recursive digital filter realization for an ionospheric channel model is proposed. This realization is in the form of a cascade of identical second-order all-pass filters, and is determined by only three parameters; two coefficients of an all-pass section, and the number of sections. The values of these parameters are optimized by a nonlinear optimization algorithm called the "downhill simplex method", so that the resulting time delay function closely approximates that of the ionospheric channel model. Comparing with the nonrecursive digital filter realization, it can be shown that the proposed recursive-digital-filter-realization is advantageous in points of view for the numbers of filter coefficients and the realization.

• Speech Sciences
• /
• v.9 no.4
• /
• pp.275-287
• /
• 2002

In man-machine communication by speech under the noisy environment, the quality of speech may be degraded severely for the machine to recognize correctly. Especially when the corrupting noise occupies the same band as the speech, the conventional fixed filters cannot filter out the noise effectively. In recent, to resolve such a problem adaptive noise canceller (ANC) is frequently used, which is based upon adaptive filters. The Adaptive recursive filters perform better than adaptive nonrecursive filters due to the added poles, but the stability may be severely threatened. In this paper an ANC system employing the adaptive recursive filter is proposed to enhance the speech corrupted by noise. And the stability of the adaptive recursive filter is guaranteed by employing the adaptive compensator.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.4C
• /
• pp.377-385
• /
• 2010

In this paper, a general and explicit technique is presented for determining the filter coefficients of maximally flat (MAXFLAT) FIR filter with the magnitude response exactly passing through a prescribed cutoff point. This technique is based on a general formula (i.e. impulse response) with an arbitrary cutoff point and permits direct computation of the coefficients of this filter type with a specified cutoff point. The technique provides an explicit method for choosing the order of flatness of the filter with the specified cutoff point. Also, in the paper, it is shown to give a computationally efficient and accurate solution to the design of the filters with the desired cutoff point.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.23 no.6
• /
• pp.950-956
• /
• 1986

In this paper, a direct addressing mode of a microprocessor is introduced to save memory capacity, and also a dedicated digital filter is constructed to speed up the filter processing and to enable an easy selection of the impulse response types. A theoretical analysis has been conducted on the errors caused by the finite word klength, rounding-off and multiplication procedures. The digital filter designed by the proposed method is made into a module which can function as a 7th-order recursive or a 14-order nonrecursive type with a simples witch operation. The proposed filter is implemented on a printed-circuit board. The frequency characteristics of this filter can be controlled by the multiplication values stored in ROMs. A low-pass, a high-pass and a band-pass filter have been designed and their frequency characteristics are verified by actual measurements. For a order higher filer, two filter modules have been cascaded into an integrated filter of 23rd-order non-recursive low-pass type and a 12th-order recursive multiband type. Their frequency characteirstics have been found to agree with the theory.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.23 no.3
• /
• pp.409-414
• /
• 1986

To eliminate the requirement of multiplications in nonrecursive filter realization, the use of a delta modulation(DM) has been studied by several researchers. However, the structure of DM filters inevitably contains an intergrator that is cascaded with the arithmetic unit to operate convolution summation. In this paper we porpose a method to determine the coefficients that may be used for implementation of a DM filter without an integrator. Also, we obtain the condition by which one can exclude the response errors due to the elimination of the integrator. By computer simulations it is shown that the performance of the proposed filter is very good, providing that the condition is satisfied.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.105-112
• /
• 2007

In general, the earlier methods for the design of MAXFLAT FIR filters have existent problems due to the approximation algorithms used to approach MAXFLAT(maximally flat) response in the passband and the stopband.. The proposed approach advanced by using of MAXSHCUT(maximally sharp cutoff) condition in this paper clearly overcomes these problems. In this approach, we use a key parameter represented with filter-order and cutoff-frequency parameters for obtaining the lowpass filters with the MAXFLAT and MAXSHCUT characteristics in the frequency domain. Consequently, this design technique leads to new MAXFLAT and MAXSHCUT FIR digital filter, which can achieve sharp-cutoff responses with the stopband attenuation exceeding 100 dB almost everywhere.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.7C
• /
• pp.451-458
• /
• 2011

Maximally flat (MAXFLAT) FIR filter design provides the advantage of giving a closed-form solution, but there still remains a problem of exactly and efficiently choosing the order of flatness for the accurate design of filters. This paper provides, through the modeling and analysis of a frequency-domain error function in the closed-form solution, how to determine the order of flatness. A proposed method, based on the frequency-domain error function, is accomplished by computing a minimum distance between its peak frequency and specified cutoff frequency. It is also shown that the proposed scheme is computationally efficient and accurate than the empirical formula given by Herrmann.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.758-761
• /
• 2007

Digital signal processing technique is applied in wide fields such as speech processing, image processing and spectrum analysis. Therefore, in order to do frequency selective operation digital filter is used in stead of analog filter and sharp filter characteristics can be implemented. Since finite impulse response (FIR) digital filter as nonrecursive type represents linear phase response characteristics and is always stable and is used in fields regarding wave information importantly such as data transmission. And due to frequency characteristics, in order to remove the Gibbs phenomenon generating around a discontinuous point, filter is designed through window function method. Therefore, in this paper to improve performance of FIR digital filter, a modified window function was applied. And the proposed method was compared with conventional methods using peak side-lobe and transition properties in simulations.