Modelling Global and Local Name Spaces for Mobile Agents Using Object Nets

While it often makes sense to use pointer-based semantics (sometimes called reference semantics) to model components of a system that are in close proximity, this kind of modelling becomes increasingly infeasible for distant components. The notion of distance here is to be seen relative with respect to the system being modelled. We use local and global name spaces to model the necessary paradigm shift for systems of communicating mobile agents. As a concrete example of modelling that requires discrimination of notions of proximity or distance, we study a scenario of mobile devices with wireless communications capabilities. The communication can be done directly between two agents or via a system of wireless network routers. To model such a system, we introduce object Petri nets that integrate both semantic paradigms found in the literature, i.e. reference semantics and value semantics. Formal definitions are given and some basic properties are proved.