On the Potential of Optimized Data Exchange in Distributed Embedded Simulation

Abstract

Simulation is an important and widely used method for the analysis of the behavior of large systems, many applications exist. Special branches of research are the simulation of very large models using distributed simulation, and embedded simulation, i.e. the coupling of virtual models with physical hardware. In our work, we approach the combination of both challenges, thus we use distributed simulation on the simulation side of an embedded simulation.Our work concentrates on Discrete Event Simulation (DES) and presents an approach that optimizes the data exchange in such scenario of distributed embedded DES based on a simulation of communication network as our case study. The evaluation shows that the optimization approach reduces the traffic in the distributed simulation system and increases the accuracy of the simulation. We conclude that in the special scenario of distributed embedded simulation, our approach improves the performance to a certain extent, however many open topics for further investigation exist that are pointed out in the conclusions.

Simulation methods and tools for collaborative embedded systems: with focus on the automotive smart ecosystems

“…Another aspect of CES development is the modeling of communication. For this, in the study presented in [19], the Universal Simulation Engine [20] is used to exchange messages within a simulation.…”

“…Especially with the integration of hardware, the simulators are often supplied with the same sensor data. In the approach presented in [19], the communication is optimized and reduced. Although it focused on distributed embedded systems, the approach can be extended to simulation of CESs.…”