Rolf Unbehauen

MOS Switched-Capacitor and Continuous-Time Integrated Circuits and Systems

Contents: Fundamentals of Sampled-Data Systems * MOS Devices for Linear Analog Integrated Circuits * Basic Properties and Systematic Analysis of Switched-Capacitor Networks * Basic Building Blocks of Linear SC Networks * Synthesis and Design of SC Filters * Design of Adaptive and Nonlinear Analog CMOS Circuits: Building Block Approach * CMOS Analog to Digital and Digital to Analog Conversion Systems * Subject Index.

Adaptive nonlinear digital filter with canonical piecewise-linear structure

A novel adaptive nonlinear filter with the least-mean-square (LMS) error criterion is presented. It is based on the so-called canonical piecewise-linear structure. As an alternative to approaches based on the Wiener-Volterra series which have so far been widely employed for adaptive nonlinear filtering, the proposed approach can exhibit adaptive performance, especially in strongly nonlinear cases, while saving computation and implementation cost. The performance of this adaptive nonlinear filter is illustrated by computer simulation results. >

Almost equiripple FIR half-band filters

Based on Chebyshev polynomials a novel analytical design procedure of almost equiripple FIR half-hand filters is developed. The closed form solution provides a direct computation of the frequency response. A formula for the impulse response coefficients is also derived. Several examples are included.

Degree, ripple, and transition width of elliptic filters

Simple analytic formulas inverting the degree education in both analog and digital equiripple filter approximations are presented. The inversion of the degree equation which is usually expressed as a ratio of theta functions, known in classical mathematics as a modular equation, is obtained in form of a finite product of Jacobian functions. From the numerical point of view it allows the recalculation of the parameters which control the optimization using an efficient arithmetic-geometric-mean procedure only. For the evaluation of k/sub 1/ from n and K (k, respectively) the zeros of the characteristic function are used and no additional computation is required. >

Switched-capacitor artificial neural networks for nonlinear optimization with constraints

Switched capacitor (SC) architectures for online solving of nonlinear optimization problems are proposed, and their properties are investigated. The proposed circuit structures are suitable for VLSI MOS implementations since they use switched-capacitor techniques. The structures exhibit a high degree of modularity, and a relatively small number of basic building blocks (computing cells) are required to implement many effective and powerful optimization algorithms. Basic mathematical operations, e.g. multiplication, addition, and nonlinear scaling transformation, are accomplished using advanced SC techniques. The validity and performance of the circuit structures are illustrated by intensive computer simulations using TUTSIM and NAP programs.<<ETX>>

Edge preserving filtering by combining nonlinear mean and median filters

The authors introduce a very general class of nonlinear filters, called mapping order-statistics filters (MOSFs). A subclass of the MOSFs, called L/sub p/-mean median filters (L/sub p/-MMF), is treated in detail. Theoretical analysis and computer simulations shown that the L/sub p/-MMF removes one kind of impulse noise as well as the nonlinear mean filter and attenuates Gaussian and uniformly distributed noise efficiently. The main advantages of the L/sub p/-MMF are that it can preserve edges as well as the median filter and can remove both the positive and negative impulse noise simultaneously. Further theoretical analysis shows that a finite-length discrete signal always converges to a fixed-point signal (root signal), if the same L/sub p/-MMF is applied iteratively. A necessary condition for a signal to be a root signal is found and proved. This necessary condition is also valid for the FIR-median hybrid filter.<<ETX>>

Note the window function with nearly minimum sidelobe energy

It is shown that a previously published class of window functions is complementary to the Chebyshev window function. The impulse response coefficients are expressed in closed form. Chebyshev window functions of the first and second kind are discussed. >

A systolic structure for complex digital filters

Systolic structures which possess a large degree of modularity, parallelism, and amenability to VLSI design for complex recursive digital filters are presented. The data can be fed into the systolic structures in a sequence and one value is allowed to all processing elements at a time. An error analysis of these systolic realizations is made, and expressions for the combined output error are given. It is observed that an additional error in the systolic realization compared with the direct-form realization is only the storage error and that all other types of errors are identically the same in both realization structures.<<ETX>>

Implementation of 2-D filters for real-time processing

An implementation technique for two-dimensional real-time linear filters with standard hardware elements is introduced. A cost analysis demonstrates that the RAM-based implementation of the elements z/sup -1//sub 1/ causes only a small part of the total cost. Therefore, as in the one-dimensional case, in two-dimensional linear signal processing the implementation cost is primarily dependent on the number of the multipliers and adders. >

Efficient evaluation of 1-D and 2-D polynomials at equispaced points

An extension to the two-dimensional case of a previously published fast algorithm for evaluating exponential functions and polynomials over periodically distributed points is generalized to evaluate two-dimensional polynomials efficiently over general periodic parallelograms. The problem of getting the initial values required for other fast algorithms is treated in a general manner. Fast algorithms are presented for computing initial values in both cases of one- and two-dimensional polynomials. It is observed that the two-dimensional recursion is not always advantageous compared with taking the one-dimensional recursion for each variable separately. >