Introduction

Abstract

Reasoning about the correctness of concurrent and distributed systems is an inherently difficult task. A main challenge comes from the fact that many processes run concurrently and interact with each other, which results in non-determinism and large execution and state spaces. Thus, it is easy for a system designer to miss a bug related to concurrency, e.g., an unforeseen race condition, deadlock, livelock, etc. Finding these kinds of bugs motivated research in model checking for many years [Baier and Katoen, 2008, Clarke et al., 1999, Emerson and Clarke, 1980, Grumberg and Veith, 2008, Queille and Sifakis, 1982]. The central problem for model checking (and of many other techniques) is state explosion. Research in model checking made several breakthroughs and developed methods to deal with this problem. Over the years, these methods have been implemented in industrial tools, which are used to verify hardware designs (e.g., microprocessors and cache coherence protocols), predominantly sequential software (e.g., device drivers), and network protocols [Grumberg and Veith, 2008]. The main line of research in model checking of concurrent systems considers systems with a fixed and a priori known number of processes.