Every channel with time structure has a capacity sequence

Abstract

On August 25th 2008 we lectured at the workshop in Budapest which honored I. Csiszár in the year of his 70th birthday, on the result that for every AVC under maximal error probability pessimistic capacity and optimistic capacity are equal. This strongly motivated us to think again about performance criteria and we came back to what we called already a long time ago [1] a (weak) capacity function (now sequence!). But this time we were bold enough to conjecture the theorem below. Its proof was done in hours. In the light of this striking observation we omit now the word “weak” which came from the connection with the weak converse and make the following