Advertisement

Integer programming applied to eigenvector computation in a class of Markov processes

  • Andre Oisel
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 229)

Abstract

The encoding of data in a number of recording and transmission devices can be modelized by a Markov process. Several performance statistics of the encoded signal (e.g. : frequency spectrum, run-length distribution, error propagation, etc) can be derived from a probability state vector, which is an Eigenvector for the encoder transition matrix.

We develop a very simple integer algorithm, applicable in this case. The integer nature of the result, not only facilitates subsequent calculations (e.g. : autocorrelation function), but also saves the code structure, which might help in analyzing many other properties.

This algorithm is part of a fully integrated program for frequency spectrum calculation, running on a microcomputer.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Significant references

  1. [1]
    ADLER, COPPERSMITH & HASSNER “Algorithms for sliding Block Codes” IEEE — Vol IT 29 — N 1 — Jan 1983 — pp 5–22Google Scholar
  2. [2]
    CARRIDLARD & Al. “Spectral Analysis of Variable Lenght Coded Digital Signals” IEEE — Vol IT 28 — N 3 — May 1982 — pp 473–481Google Scholar
  3. [3]
    CHIEN “Upper Bound on Efficiency of DC-Constrained Codes” BSTJ — Vol 49 — N 9 — Nov 1970 — pp 2267–2287Google Scholar
  4. [4]
    DAVIS “Monte-Carlo Analysis of Recording Codes” INTERMAG Conference — HAMBURG — April 1984Google Scholar
  5. [5]
    EGGENBERGER & HODGES “Sequential Encoding & Decoding of Variable Length Codes.” US Patent 4115768 — 19 Sep 78Google Scholar
  6. [6]
    FRANASZEK “A Sequence-State Coding for Digital Transmission” BSTJ — Vol 47 — N 1 — JAN 1968 — pp 143–157Google Scholar
  7. [7]
    FRANASZEK “Run-Length Limited Variable Length Coding with error propagation limitation.” US Patent 3689889 — 5 Sep 72Google Scholar
  8. [8]
    FROEBENIUS “Ueber Matrizen aus Nicht-Negativen Elementen” BERLIN — 1912Google Scholar
  9. [9]
    IMAI & SAITO “State-Function Decoding Methods for (d,k) Run-Length Limited Codes” 7th Symposium on Information Theory and its Applications of Japan. KINUGAWA — Nov 1984Google Scholar
  10. [10]
    JUSTESEN “Information Rate and Power Spectra of digital Codes” IEEE — Vol IT 28 — N 3 — May 1982 — pp 457–472Google Scholar
  11. [11]
    LEWIN “LOGICAL DESIGN OF SWITCHING CIRCUITS” ... In French: “SYSTEMES LOGIQUES” Editions SPES S.A. — LAUSANNE — 1972Google Scholar
  12. [12]
    LINHOLM “Power Spectra of Channel Codes for digital Magnetic Recording” IEEE — Vol MAG14 — N 5 — Sep 1978 — pp 321–323Google Scholar
  13. [13]
    TANG & BAHL “Block Codes for a Class of Constrained Noisless Channels” INFO & CONTROL — Vol 17 — Dec 1970 — pp 436–461Google Scholar
  14. [14]
    VARGA “MATRIX ITERATIVE ANALYSIS” — Prentice Hall — 1962Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1986

Authors and Affiliations

  • Andre Oisel
    • 1
  1. 1.CII-Honeywell BULLLes Clayes-S-BoisFrance

Personalised recommendations