The design of real IIR digital filters using conformal transform variables has been extensively researched; this paper deals with the design of complex IIR digital filters having non-symmetric transfer functions, and directly designed using transformed variables. The transformation used is an extension of a transform used previously for continuous-time complex filters; the extension adds a bilinear function which enables the transform to be used for designing discrete-time transfer functions. Approximations having equiripple and monotonically decreasing passbands are supported. Methods for improving numerical accuracy and minimizing coefficient sensitivities are described. A high-order cascade filter example simulated using Monte Carlo analysis demonstrates almost ideal accuracy.

The toolbox described in conference paper “Complex Cascade Digital Filters” [1] will be demonstrated. Examples will be presented going through the steps of: a) specifying the desired filter responses and normalizing them so the passband is between ±1, b) finding the analog protype transfer function, c) converting the analog prototype transfer function to a digital transfer function based on the bilinear transform and undoing the previous normalization of the specifications, and d) finding a good digital cascade realization of the transfer function.

Complex filters are filters that process two signals often denoted the real and imaginary signal components. This publication deals with how to realize digital complex filters using cascade architectures. A number of design techniques to improve numerical inaccuracies are presented. The design algorithms have been implemented using Matlab; methods to overcome Matlab limitations when manipulating complex systems are described. A number of examples are presented. The sensitivities of the filter realizations are investigated using Monte Carlo analysis.

The Accreditation Board for Engineering and Technology, Inc. (ABET) and the Computing Sciences Accreditation Board (CSAB) signed a Memorandum of Agreement in November 1998 to integrate CSAB's accreditation services with ABET, with a transition time of approximately two years. During the interim period, the operations of the Computer Science Accreditation Commission (CSAC) are contracted by CSAB to ABET. A committee with CSAC, CSAB, and ABET representation is working to set up the new commission for accrediting programs in the computing sciences. This new commission will probably be called the Computing Accreditation Commission (CAC). Other activities are underway to try to assure that this integration goes as smoothly as possible.

A frequency-difference detector (FDD) based on an adaptive notch filter is proposed. Frequency detectors are useful in decreasing acquisition times in phase-locked-loop and Costas-loop-based clock and carrier-recovery systems. To avoid certain drawbacks associated with signal down-conversion to baseband, the input is initially processed in the bandpass domain. In applications where limits on acquisition time demand a wide closed-loop tracking bandwidth, the output of the proposed FDD has less inherent ripple and bias when compared with a quadricorrelator, a commonly used FDD. Two applications of the proposed FDD in the demodulation of BPSK data are presented.