Leonard T. Bruton

Low-sensitivity digital ladder filters

A class of low-sensitivity digital ladder filters may be realized in the voltage-current domain by direct analogy with the continuous resistively terminated LC ladder filter. The problem of unrealizability that is implied by delay-free loops in the discrete signal flow graph is overcome by using transformations that correspond to LC elements that exhibit finite Cube factors. The resultant deterioration in the passband of the transfer function is determined from the Blowstein LC ladder sensitivity theory and is, thereby, shown to be low valued at high sampling frequencies. The sensitivity properties of this class of digital ladder filters are directly analogous to the LC prototype because each LC element is replaced by a digital multiplier. Consequently, if maximum power transfer is approximately maintained throughout the passband of the LC prototype, then the first-order sensitivity of the corresponding digital transfer function to multiply a coefficient quantization is necessarily low valued.

Three-dimensional cone filter banks

Three dimensional (3-D) narrow cone-shaped filter passbands are ideally required for the selective filtering of sampled broad-band 3-D plane waves on the basis of their directions of arrival. A method is proposed for approximating narrow 3-D cone-shaped passbands using a 3-D cone-filter-bank structure in which the subbands consist of band limited 3-D narrow-band infinite impulse response (IIR) beam filters having 3-D uniform bandwidths that are approximately proportional to their distance from the origin in the 3-D frequency space. The 3-D beam filters may be realized from a cascade of 3-D IIR frequency-planar filters. It is shown that the proposed 3-D cone filter bank achieves low distortion of broad-band passband 3-D plane waves, and significant attenuation of broad stopband plane waves.

Performance evaluation of digital audio watermarking algorithms

We propose an algorithm-independent framework for rigorously comparing digital watermarking algorithms with respect to bit rate, perceptual quality, computational complexity, and robustness to signal processing. The framework is used to evaluate five audio watermarking algorithms from the literature, revealing that frequency domain techniques perform well under the criteria.

Fractal block coding of digital video

A video coding method is proposed which is based upon fractal block coding. The method utilizes a novel three-dimensional partitioning of input frames for which a number of efficient block-matching search methods can be used, and permits spatio-temporal splitting of the input blocks to improve overall-encoding quality. After describing the basic fractal block coding algorithm, the details of the proposed three-dimensional algorithm are presented along with encoding and decoding results from two standard video test sequences, representative of video-conferencing data. These results indicate that average compression rates ranging from 40 to 77 can be obtained with subjective reconstruction quality of video-conferencing quality. The results also indicate that, in order to meet the compression rates required for very low bit rate coding, it is necessary to employ additional techniques such as entropy encoding of the fractal transformation coefficients. >

Nonideal performance of two-amplifier positive-impedance converters

Positive-impedance converting networks may be used to realize driving-point inductance and frequency-dependent negative resistance. The Q factors of these immittances are evaluated over a wide frequency range in terms of the relevant parameters. Three distinct types of Q- factor behavior are derived and it is shown that extended wide-band operation may be achieved such that Q factor is virtually independent of amplifier phase shift. The use of design-value asymptotes is explained as a means of determining the behavior of practical circuits.

A Speed-Optimized Systolic Array Processor Architecture for Spatio-Temporal 2-D IIR Broadband Beam Filters

For high-speed plane-wave filtering applications, real-time 2-D spatio-temporal linear-array broadband beam filters are required, operating at temporal frame rates in excess of hundreds of megahertz. The corresponding application specific VLSI circuits must have low critical-path latencies. A novel high-speed systolic array architecture for a first-order 2-D broadband frequency-planar spatio-temporal beam filter is proposed for this purpose and employs a field-programmable gate array (FPGA) circuit where the critical path latency is minimized by timing optimization of inter- and intra-parallel processor pipelines, together with 3-D look-ahead techniques. The method facilitates single-chip VLSI circuit implementations operating at real-time frame rates of several hundred megahertz.

Noise performance of low-sensitivity active filters

The noise limitations of some important low-sensitivity RC-active filter realizations are derived; the inherent rms noise (noise from essential network resistors) that exists at the output terminals of a bandpass active filter is determined for two-integrator loop and positive impedance converter filter realizations. Output noise spectral density and total rms output noise are determined for the two-integrator loop section. The noise contributed by the essential circuit resistors and the operational amplifiers is taken into account and it is shown that exact calculation of output noise may be obtained in terms of Q factor and impedance level. The results are used to obtain the theoretical noise limitations of these low-sensitivity active filters. It is found that the deterioration in signal-to-noise ratio (SNR), due to the internal filter noise, is independent of Q factor for high Q realizations.

Noise performance of RC-active quadratic filter sections

A method for calculating the noise voltage at the output of quadratic filter sections is developed. Multiple-feedback low-pass, bandpass, and high-pass quadratic filter sections realized using differential-input single-ended output operational amplifiers are analyzed. The amplifiers are assumed to have infinite input impedance, infinite gain, and zero output impedance. The noise sources associated with the amplifiers are assumed to be statistically independent, but can have both white and l/f noise components. A noise analysis of a fourthorder maximally flat low-pass filter realized by cascading two quadratic filter sections is included.

Transfer function synthesis using generalized doubly terminated two-pair networks

In Part I of this paper, a procedure for the synthesis of doubly terminated two-pair networks is developed. This procedure, coupled with certain properties of mirror and anti-mirror image polynomials, permits the design of digital filters with extremely low sensitivity of the steady-state magnitude of the transfer function to multiplier coefficient errors. In Part II, the ladder structure is examined as one possible internal structure for the two-pair specified in Part I. The synthesis procedure of Part I, combined with a special case of the ladder decomposition of Part II, leads to a previously proposed structure, the lossless digital integrator ladder, for which no direct design method was formerly available.

Synthesis of digital ladder filters from LC filters

Digital ladder filter networks may be realized by applying simple transformations to the flow-graph network representation of continuous domain resistively terminated LC two ports. It is shown that such methods have the disadvantage that there does not exist a transformation with the three requirements that the entire imaginary s -plane axis map to the z -plane unit circle, that the resultant discrete network is stable and finally that the resultant discrete network be computationally realizable due to Its freedom from delay free loops. A synthesis technique is proposed for low sensitivity digital ladders. The conventional transformation that is applied to the LC filter prototype does lead to a stable structure with the required mapping property; the delay free loops are eliminated by straightforward flow-graph manipulation. The coefficient sensitivity of the magnitude transfer function |H(e^{f{\omega}T}| is low valued throughout the passband.