Zero-clairvoyant scheduling with inter-period constraints

Abstract

This paper introduces two new mathematical modeling paradigms called Periodic Linear Programming and Periodic Quantified Linear Programming. The former is an extension of traditional Linear Programming, whereas the latter extends Quantified Linear Programming. We use these tools to capture the specifications of real-time embedded systems, which are characterized by uncertainty, complex timing constraints and periodicity. The strength of the modeling techniques lies in the ease with which these specifications can be represented and analyzed. © Springer-Verlag Berlin Heidelberg 2004.