Problems of Information Transmission最新文献

The graph model is an appreciable tool for data transmission network, where the feasibility of data transmission in site attack circumstances can be described by fractional critical graphs, and the vulnerability of networks can be measured by isolation toughness variant. This paper considers both the stability of the network and the feasibility of data transmission when the sites are destroyed, and determines the isolated toughness variant bound for fractional ((a,b,n))-critical graphs, where the parameter (n) represents the number of damaged sites at a certain moment. A counterexample proves the sharpness of the given isolated toughness variant bound. The main theoretical conclusion provides an equilibrium between performance and cost in network topology designing.

The paper presents linear predictors and causal filters for discrete-time signals featuring some different kinds of spectrum degeneracy. These predictors and filters are based on approximation of ideal noncausal transfer functions by causal transfer functions represented by polynomials of the Z-transform of the unit step signal.

We consider contact processes on locally compact separable metric spaces with birth and death rates that are heterogeneous in space. We formulate conditions on the rates that ensure the existence of invariant measures of contact processes. One of the crucial conditions is the so-called critical regime condition. To prove the existence of invariant measures, we use the approach proposed in our preceding paper. We discuss in detail the multi-species contact model with a compact space of marks (species) in which both birth and death rates depend on the marks.

We consider a geometric approach to the notion of metric entropy. We justify the possibility of this approach for the class of Borel invariant ergodic probability measures on sofic systems, which is the first result of such generality for non-Markovian systems.

We study sequences (left{A_nright}_{n=-infty}^{+infty}) of elements of an arbitrary field (mathbb{F}) that satisfy decompositions of the form

where (a_1,a_2,b_1,b_2colon mathbb{Z}tomathbb{F}). We prove some results concerning the existence and uniqueness of such sequences. The results are used to construct analogs of the Diffie–Hellman and ElGamal cryptographic algorithms. The discrete logarithm problem is considered in the group ((S,+)), where the set (S) consists of quadruples (S(n)=(A_{n-1},A_n, A_{n+1}, A_{n+2})), (ninmathbb{Z}), and (S(n)+S(m)=S(n+m)).

Abstract

We propose an approach for generalization of formulas previously obtained by the authors for the first and second queue length moments in a queueing system with a nonordinary Poissonian arrival flow, single server, and constant service time to the case of a variable service time. The service time is assumed to be a random variable with a finite set of values. This model is adequate for a vast class of batch transmission systems, since the batch transmission time in real-world systems can take only finitely many values.

Abstract

An attempt is made to give an accurate mathematical interpretation of the energy–time uncertainty relation.

Abstract

A covering code, or a covering, is a set of codewords such that the union of balls centered at these codewords covers the entire space. As a rule, the problem consists in finding the minimum cardinality of a covering code. For the classical Hamming metric, the size of the smallest covering code of a fixed radius (R) is known up to a constant factor. A similar result has recently been obtained for codes with (R) insertions and for codes with (R) deletions. In the present paper we study coverings of a space for the fixed length Levenshtein metric, i.e., for (R) insertions and (R) deletions. For (R=1) and (2) , we prove new lower and upper bounds on the minimum cardinality of a covering code, which differ by a constant factor only.

Abstract

We define the bidirectional distance profile (BDP) of a convolutional code as the minimum of the distance profiles of the code and its corresponding “reverse” code. We present tables of codes with the optimum BDP (OBDP), which minimize the average complexity of bidirectional sequential decoding algorithms. The computer search is accelerated by the facts that optimum distance profile (ODP) codes of larger memory must have ODP codes of smaller memory as their “prefixes”, and that OBDP codes can be obtained by “concatenating” ODP and reverse ODP codes of smaller memory. We compare the performance of OBDP codes and other codes by simulation.