PROBLEMS OF INFORMATION TRANSMISSION
A translation of Problemy Peredachi Informatsii


Volume 27, Number 1, January–March, 1991
Back to contents page

CONTENTS                   Powered by MathJax

 

Output-Uniform Universal Coding of Discrete Memoryless Sources
Yu. M. Shtarkov
pp. 1–10

Abstract—The redundancy of output-uniform universal coding of discrete memoryless sources is defined as a function of the mean delay and an asymptotically optimal method is proposed for generating the set of input messages. The problem of weighted universal coding is formulated and solved, with redundancy a function of the mean delay of the encoded source.

 

Bounds on $E$-Capacity of a Channel with a Random Parameter
M. E. Haroutunian
pp. 10–18

Abstract—We obtain a random-coding lower bound for the $E$-capacity of a channel with a random parameter under complete information about the encoder states [S.I. Gelfand and M.S. Pinsker, Probl. Contr. Inf. Theory, 9, no. 1, 19–31 (1980)] and also improve the upper bound of [E.A. Haroutunian and M.E. Haroutunian, Probl. Contr. Inf. Theory, 17, No. 2, 99–105 (1988)].

 

Weight Distributions of Subfield Subcodes of Algebraic-Geometric Codes
S. G. Vladut and A. N. Skorobogatov
pp. 19–29

Abstract—We consider subcodes of algebraic-geometrical codes over a prime subfield that generalize BCH codes and classical Goppa codes. The deviation of the weight distribution of such a subfield code from the weight distribution of a random code is estimated by two different techniques. If the code length is large compared with the genus of the curve and the degree of the divisor defining the code, then the distribution is close to the random-code distribution.

 

Signal-Code Constructions without Zero Spectral Component
V. V. Zyablov and S. L. Portnoi
pp. 30–39

Abstract—A method is proposed for generating signal-code constructions whose spectrum does not contain the zero frequency. The constructions are based on generalized concatenated codes with signals in the inner stage and error-correcting codes in the outer stage. Examples of constructions are given. Viterbi maximum-likelihood decoding of the constructions is considered. Examples of decoder simulation in a Gaussian channel are described. Numerical results for the frequency spectrum are presented.

 

Sequential Decoding Algorithm Producing the Maximum-Likelihood Estimate for Low-Density Gallager Codes
V. B. Balakirsky
pp. 40–48

Abstract—A sequential decoding algorithm is proposed for low-density block codes introduced by Gallager [Low-Density Parity-Check Codes, MIT Press, Cambridge (1963)]. If the low-density code is used for transmission of information in a binary symmetric channel, the proposed algorithm produces a maximum-likelihood estimate of the transmitted codewords. The characteristics of the algorithm are reported for a block code of length 20 and minimum distance 6. An asymptotic upper bound is obtained on the average number of decoder computations over the code ensemble.

 

Detection of a Signal with Unknown Moments of Appearance and Disappearance
V. G. Repin
pp. 49–58

Abstract—We consider the description of a signal with unknown moments of appearance and disappearance when the a priori probability distributions for the appearance moment and the signal length are not available. The adaptive Bayesian approach is used to derive simple recursive decision-making algorithms for finite and semifinite observation intervals, based on computation of one-dimensional Markovian sufficient statistics. The detection characteristics are determined. An algorithm that detects the appearance and disappearance of a signal is described.

 

Hypothesis Testing with Incompletely Known Distribution Function of the Observations and Unknown A Priori Probabilities of the Hypotheses
A. A. Slepokurov
pp. 58–66

Abstract—An optimal decision rule is derived for testing two hypotheses under incompletely known distributions of the observed signals and unknown a priori probabilities of the hypotheses. Optimality is understood in the sense of minimizing the maximum excess of average risk over Bayesian risk. Some examples are considered.

 

Demodulation of Discrete-Frequency Signals in a Multipath Channel
D. R. Ankudinov and Yu. S. Shinakov
pp. 67–73

Abstract—We present the results of analysis of second-order demodulation algorithms for discrete-frequency signals. Demodulation methods are compared for the case where the channel transmission coefficients are known and unknown.

 

Optimal Rates of Convergence of Estimates in the Stochastic Problem of Computerized Tomography
A. P. Korostelev and A. B. Tsybakov
pp. 73–81

Abstract—We consider the stochastic tomography problem, i.e., reconstruction of an unknown image from observations of its integrals over hyperplanes (lines in the two-dimensional case) in the presence of random noise. The minimax lower bound on image reconstruction accuracy is established in classes of smooth functions. An image estimation method is proposed which achieves this bound by the order of rate of convergence.

 

Data Compression in Remote Sounding
V. P. Yakovlev
pp. 82–85

Abstract—We compare the compression of digital data from high- and low-resolution satellite instruments. We show that entropy coding of separate measurements reduces the description by approximately one half; a similar reduction is additionally achieved by allowing for statistical dependence between neighboring measurements.