Toward Simulation-Assisted Architecture Design Space Exploration of Indoor Robotics Networks

Abstract

In this paper, we explore the combination of static architectural optimization with dynamic simulation. We describe an architecture through abstract models of components and a set of constraints that drive the design space exploration process to few promising solutions. These are evaluated through simulation to extract detailed performance parameters. The end objective of this work is to then feed back the information in the form of extra constraints to converge to a solution that satisfies the application requirements. We tested and validated the tool extensions, optimizing and analyzing two indoor robotics scenarios for critical performance parameters, i.e., overall throughput and end-to-end delay (E2E).