On the use of Dynamic Topologies for Representing Spatially Moving Entities

Abstract

In many psychical systems, entities can only communicate when within the range of their sensors. When entities are static, the topology of the communication infrastructure can be known in advance and kept unchanged during simulation. However, representing moving entities becomes more complex since interconnections change with time as the relative distances between entities evolve. Given the dynamic nature of the problem, a common solution involves the use of publish/subscribe communication and the corresponding multicast message passing. In this paper we exploit the use of a peer-to-peer infrastructure (p2p) exhibiting a dynamic topology as an alternative representation for spatially moving entities. We use the Heterogeneous Flow Systems Specification (HFSS), a modular modeling formalism designed to represent hybrid systems with time-variant topologies. We present a simplified model of a defense system comprising airborne radars, drone detection and cancellation. We demonstrate the benefits of p2p communication over publish subscribe interaction.