A method is presented for generating random numbers with uniform distribution using linear recurrence sequences with very large period lengths. This method requires an irreducible polynomial modulo 2 to define the sequence. A suitable method for generating an infinite number of such polynomials is presented. The polynomials generated in this way can have an arbitrarily large degree, and a large enough order to make them suitable for practical applications.

In this paper, we introduce the neighborhood of binary self-dual codes. Further, we show that for codelength divisible by 8 such a neighborhood consists of three self-dual codes, two of them are doubly-even and one is always singly-even. We investigate the relationship between neighboring codes. Finally, we prove that no better Type I code exists than the best possible Type II code of the same length.

No abstract

Parallel computing paradigms are of high interest because parallel computers are everywhere. In this paper connections of two parallel computing paradigms, namely intervalvalued computing and cellular automata are shown. The intervalvalued paradigm is a new computing paradigm that is universal and in this framework hard problems can be solved efficiently. One-dimensional cellular automata are also universal computing models that uses a high level of parallelism. We show that interval-valued computing can simulate cellular automata and also interval-valued computing can be simulated by cellular automata.