Model checking implicit-invocation systems

While implicit invocation (publish-subscribe) systems have good engineering properties, they are difficult to reason about and to test. Model checking such systems is an attractive alternative. However, it is not clear what kinds of state models are best suited for this. We propose a structural approach, which factors the model checking problem into two parts: behavior specific to a particular implicit invocation system, and reusable run-time infrastructure that handles event-based communication and delivery policies. The reusable portion is itself structured so that alternative run-time mechanisms may be experimented with.