Arithmetic codes resembling neural encoding

From MaRDI portal





In a residue code with the moduli \(m_ 1,m_ 2,...,m_ n\) the integer X is represented by the vectorial form \(\{r_ 1,r_ 2,...,r_ n\}\), such that \(r_ i\equiv X mod m_ i\), \(i=1,2,...,n\). Since the moduli are not pairwise relative primes, not every n-tuple of integers \(\{r_ i|\) \(0\leq r_ i<m_ i\), \(i=1,2,...,n\}\) represents a code word: only those satisfying \(r_ i\equiv r_ j mod g_{ij}\), where \(g_{ij}\) is the greatest common divisor of \(m_ i\) and \(m_ j\). Thus redundancy is introduced in a distributed way, inherent in the structure of the system. Known theorems on the error detecting, identifying, and correcting capacity of such codes are generalized. Application to brain modeling is touched upon.











This page was built for publication: Arithmetic codes resembling neural encoding

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1084076)