Comments on “Predictability of Failure Event Occurrences in Decentralized Discrete-Event Systems and Polynomial-Time Verification”

摘要

We show that the notion of copredictability studied in the considered paper is equivalent to the already existing notion of uniformly bounded coprognosability introduced in a 2010 article of Kumar and Takai. In fact, a weaker, more general notion of coprognosability, which does not need a uniform bound for prognosing an impending failure, was also introduced by Kumar and Takai in 2010. It was shown that for the case of regular languages, the two notions (the one with a uniform bound and the other without it) coincide. As a result, the algorithm for testing coprognosability for regular languages presented by Kumar and Takai in their 2010 paper also tests the copredictability concept in the considered paper, which presented a test of its own. Finally, the fact that copredictability is stronger than codiagnosability in the absence of unobservable cycles was also shown in the 2010 article of Kumar and Takai, and it is another result that is reproduced in the considered paper.