A Hybrid Modeling and Simulation Formalism for Static and Dynamic Analysis of DoDAF Architecture

Moini, A., Madni, A.M., and Ciardo, G.

Abstract

The objective of architecture analysis is to determine whether a proposed system-of-systems (SoS) is capable of meeting performance and effectiveness requirements defined by measures of performance (MOPs) and measures of effectiveness (MOEs), respectively. Within the defense community, diverse stakeholders from a variety of organizations (e.g., operational, acquisition, training test and evaluation) have been using architectural models to acquire insights into the dynamic behavior of business processes and their supporting systems, identify logical errors or inconsistencies in system specifications, an quantify performance and operational utility of systems. Many of these systems are physics-based (e.g., weapon platforms, ISR assets), operate in continuous space-time continuum, and interact with discrete (time and state) processes. It is in response to these characteristics that Fluid Stochastic Petri Nets (FSPNs) were first introduced as a scalable formalism to model hybrid systems (i.e., physics-based systems with continuous quantities controlled by discrete logic) in a manner also contain that the state space explosion. A key benefit of using FSPNs as a modeling formalism is its seamless treatment and unification of traditionally independent modeling concerns (e.g., logical, structural, behavioral, temporal and dynamic stochastic behavior) within the confines of a single, unified framework. This paper explores the use of FSPNs as a scalable, cost-effective modeling and simulation formalism for performing logical, behavioral and performance analysis of DoDAF architectures. Specifically, it discusses some of the key issues associated with FSPN simulation and explores the feasibility of using FSPNs for analyzing static and dynamic aspects of DoDAF architectures.

From: Moini, A., Madni, A.M., and Ciardo, G., Conference on System Engineering Research (CSER), 2008.