Overcoming observability problems in distributed test architectures

Abstract

This paper investigates conditions that must be satisfied by an FSM for the existence of input sequences that can be applied in a distributed test architecture without encountering controllability and observability problems and without using external coordination messages. Such conditions have two potential values. First, they can be used to determine whether we require coordination messages and thus a network that connects the testers. Second, if we wish to avoid the use of coordination messages in testing then these conditions can be seen as testability conditions that can inform the design process. Results given in this paper differ from those in the following ways. First, the conditions are strictly weaker than those in since we are less restrictive in the ways we achieve our goals. Second, only considered observability problems; we consider both controllability and observability problems. In addition, only considered a particular type of observability problem and we generalize this. Finally, we investigate the situation in which we need only add input sequences to complement a given test/checking sequence ρ and prove that the conditions for this problem are equivalent to those for the original problem.