A new sufficient feasibility test for asynchronous real-time periodic task sets

The problem of feasibility analysis for asynchronous periodic task sets (i.e. where tasks can have an initial offset) is known to be co-NP-complete in the strong sense. A sufficient pseudo-polynomial test has been proposed by Baruah et al., which consists in analyzing the feasibility of the corresponding synchronous task set (i.e. all offsets are set equal to 0) for which a pseudo-polynomial test is known. If the test gives a positive result, then the original asynchronous task set is feasible; else, no definitive answer can be given. In many cases, this sufficient test is too pessimistic, i.e. it gives no response for many feasible task sets. In this paper, we present a new sufficient pseudo-polynomial test for asynchronous periodic task sets. Our test reduces the pessimism by considering the value of the offsets and obtaining a set of critical arrival patterns. We show, with a set of extensive simulations, that our test outperforms the previous sufficient test.