Vidosav Stojanovic

Digital filter implementation based on the RNS with diminished-1 encoded channel

A technique, based on the residue number system (RNS) with diminished-1 encoded channel, has being used for implementing a finite impulse response (FIR) digital filter. The proposed RNS architecture of the filter consists of three main blocks: forward and reverse converter and arithmetic processor for each channel. Architecture for residue to binary (reverse) convertor with diminished-1 encoded channel has been proposed. Besides, for all RNS channels, the systolic design is used for the efficient realization of FIR filter. A numerical example illustrates the principles of diminished-1 residue arithmetic, signal processing, and decoding for FIR filters.

The use of Residue Number System in the design of the optimal all-pole IIR digital filters

The paper discusses the Residue Number System (RNS) implementation of all-pole recursive digital filter. The RNS offers the advantage of using arithmetic based on integer operations and a simple hardware realization involving arithmetic operations implemented in parallel. In residue number system of IIR filter needs a scaling operation which is necessary to keep data within the limited dynamic range. The effects of residue scaling to performances of recursive digital filters are described here.

Transitional Butter–Elliptic filter suitable for complementary decomposition

Direct synthesis of recursive digital filters, whose characteristic function belongs to the new IIR filters class, is described in this article. Attenuation characteristic in the passband lies between the Butterworths and the Elliptics attenuation characteristics. All zeros of transfer function are placed on the unit circle so that the condition for complementary decomposition is fulfilled. Complementary decomposition into parallel connection of two allpass filters is performed. Then, lowpass filter is obtained summing these two filters and its complementary highpass filter – by their subtraction. The main advantage of this kind of filters is its smaller sensitivity to digital word length change compared to standard filters.

Group delay sensitivity of recursive digital filter transfer functions with multiple poles

Studies the group delay sensitivity of recursive digital filters whose group delay frequency response approximates a constant group delay in the Chebyshev sense. A technique is developed which predicts a group delay deviation of recursive digital filters due to changes in digital filter coefficients. It is illustrated that the technique can serve to determine and compare the group delay sensitivity of the transfer function with simple or multiple poles. It is shown that the multiple pole transfer function is less sensitive than a simple pole transfer function.<<ETX>>

Nearly monotonic passband low-pass filter design by using sum-of-squared Legendre polynomials

SUMMARY A new type of filter approximation, which utilizes orthonormal Legendre polynomials, referred to as sum-of-squared Legendre polynomials, is presented in this paper. Power transmission coefficient and the group delay of the proposed filter are compared with those of the Butterworth, Legendre–Papoulis, and Halpern filters. In order to illustrate the design of the proposed filter function, sum-of-squared Legendre polynomials coefficients and normalized element values of the low-pass LC (inductor-capacitor) ladder network prototype are given, up to the 10th degree. For continuous-time domain filtering, doubly terminated LC ladder network topologies have very low sensitivities to changes of component values. In order to determine the effect of variation of all reactive components on the filter attenuation characteristic, the new sensitivity function has been proposed. Copyright

The implementation of two channel IIR quadrature mirror filter bank based on residue arithmetic

Design of two-channel quadrature mirror filter (QMF) bank constructed by real infinite impulse response (IIR) digital all-pass filter is considered. Residue number system offers parallel processing and digital hardware implementation for binary operation of addition, subtraction and multiplication. There is no carry propagation between these operations. Hence, devices operating in this principle inherit properties of high speed and low power consumption. However, this properties makes overflow detection difficult during any operation. This paper discusses the application of RNS in designing the infinite impulse response (IIR) filter.

Transitional Butterworth-Thiran recursive digital filters

The transitional Butterworth-Thiran recursive digital filters (TBTh) whose characteristics change gradually from the characteristic Butterworth filter to the characteristic Thiran filter which have maximally flat group delay characteristic is present in this paper. The poles of the transition filter are obtained by interpolation between the poles of the Butterworth filter and poles of the Thiran filter.

Design of recursive digital filters without analog conversion

Design of recursive digital filters without analog conversion is described. The methods described in this paper enable the coefficients of polynomials containing poles of the transfer function to be obtained either in a closed, form or using iterative methods. These filters cannot be obtained using transforms from a continuous domain.