RANDOMNESS PROPERTIES OF NON LINEAR CONVOLUTIONAL CODING

PETRUZZI, D. M.; CASTIÑEIRA MOREIRA, J.; LEVIN, D. G.

Rio Cuarto, Córdoba

Congreso; XI Reunión del Procesamiento de la Información y Control (RPIC); 2005

UNRC, Córdoba

In this paper we present convolutional coding schemes based on a non-linear finite state sequence machine designed over a Galois Field GF(q), whose output shows optimal randomnessproperties. Schemes with the maximum output sequence length (all zero input response) arepresented for different Galois Fields. Two coefficients are used to quantitatively measure theproximity of the output behavior of these schemes with respect to the ideal white noise behavior. The proposed schemes outperform those presented by Frey in his paper, and achieve the optimal Quasi-Chaotic properties available for a given chaotic digital scheme. The proposed schemes have also a very little loss in BER performance so that they are a good alternative to the design of systems for which encryption and error correction are important joint goals. They can be  mplementedwith time variant coefficients in order to highly improve cryptographic properties of the ransmission.