NASA Scientific and Technical Aerospace Reports
NASA Scientific and Technical Aerospace Reports
NASA Scientific and Technical Aerospace Reports
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
This paper had been prepared as a tutorial complement to its companion paper, the ARRAY MATRIX FOR SIGNAL<br />
PROCESSING [I], but could not be presented in the ICASSP-86 framework. Its objective was, <strong>and</strong> it is so now, to exemplify<br />
some of the adaptations <strong>and</strong> generalizations of matrix methods stemming from Tait’s ARRAY MATRIX theory [2]. That<br />
theory has since been extended <strong>and</strong> consolidated as a mathematics of polyxa [3], which can abbet, we believe, the smooth<br />
transition of vector-matrix methods into the array processing era. Through these papers we wish to encourage individual <strong>and</strong><br />
collective adoption of what is now called ‘extended’ matrix methods <strong>and</strong> notation, for communicating work on signal<br />
processing <strong>and</strong> in system theory. We exemplify this here on rederivation of the remarkable MARPLE algorithm.<br />
Author<br />
Matrices (Mathematics); Matrix Methods; Signal Processing; Systems Analysis<br />
20060002110 Fukui Univ., Japan<br />
On Generalized Feussner’s Principle in a Two-graph<br />
Matsumoto, Tadashi; Hirabayashi, Kimitaka; Nakano, Shigeyoshi; 1987 IEEE International Symposium on Circuits <strong>and</strong><br />
Systems, Volume 2; 1987, pp. 507-510; In English; See also 20060002103<br />
Contract(s)/Grant(s): MOE-58550229; MOE-60550245; Copyright; Avail.: Other Sources<br />
The general formulation of the generalized Feussner’s Principle in a two-graph is presented which permits the systematic<br />
<strong>and</strong> formal classification of all the common trees with respect to an arbitrary subset E(sub s) of edges. The necessary <strong>and</strong><br />
sufficient condition for the above class specified by E(sub s) to be an empty set or not is given which is useful for determination<br />
of the optimally classifiable edge-parameters for the finest classification of all the common trees of a two-graph.<br />
Author<br />
Classifications; Graphs (Charts); Trees (Mathematics)<br />
20060002159 California Univ., Berkeley, CA, USA<br />
Architecture Considerations for High Speed Recursive Filtering<br />
Kumar Parhi, Keshab; Wen-Lung, Chen; Messerschmitt, David G.; 1987 IEEE International Symposium on Circuits <strong>and</strong><br />
Systems, Volume 2; 1987, pp. 374-377; In English; See also 20060002103<br />
Contract(s)/Grant(s): N00039-86-R-0365; DCI-85-17739; Copyright; Avail.: Other Sources<br />
In this paper, we introduce pipeline interleaving in the incremental block-state structure for pipelined block<br />
implementation of high sampling rate recursive digital filters. We present a new approach to the state update implementation<br />
using a novel decomposition technique. This novel state update implementation, <strong>and</strong> the incremental output computation<br />
(based on the incremental block-state structure) lead to an efficient implementation of pipelined block recursive digital filters<br />
of (asymptotic) complexity linear in filter order (based on a quasi-diagonal state update matrix) <strong>and</strong> block size. The complexity<br />
of this implementation as measured by the number of multiplications is asymptotically same as that of non-recursive systems<br />
independent of the implementation methodology. However, for smaller block sizes, bit-level pipelined bit-serial word-parallel<br />
implementation may lead to reduced complexity realization as compared to bit-level pipelined bit-parallel word-parallel<br />
implementation. This comparison is assisted by the concept of an implementable delay operator introduced in this paper.<br />
Author<br />
Architecture (Computers); Digital Filters; Linear Filters; High Speed; IIR Filters<br />
20060002185 Illinois Univ., Urbana-Champaign, IL, USA<br />
Efficient Embedding of Planar Graphs in Linear Time<br />
Tamassia, Roberto; Tollis, Ioannis G.; 1987 IEEE International Symposium on Circuits <strong>and</strong> Systems, Volume 2; 1987,<br />
pp. 495-498; In English; See also 20060002103<br />
Contract(s)/Grant(s): RSCH-84-06-049-6; Copyright; Avail.: Other Sources<br />
In this paper we consider planar embeddings of n-node planar graphs in the rectilinear grid, where vertices are grid points<br />
<strong>and</strong> edges are nonintersecting grid paths. We present a new embedding algorithm that runs in linear time. The total number<br />
of bends in the embeddings constructed by our algorithm is very small. Furthermore, the embeddings occupy O(n(sup 2)) area,<br />
which is the best possible in the worst case. Our results are important in the design of VLSI chips. Other applications can be<br />
found in the areas of communication by light or microwave, transportation in space, <strong>and</strong> automatic graph drawing.<br />
Author<br />
Algorithms; Computational Grids; Very Large Scale Integration; Graph Theory; Edges; Embedding<br />
120