On the Characterization and Analysis of System of Systems Architectures

Abstract

A goal of an agile organization is the ability to adapt its structure to constantly changing operating environments so it can provide the multiple capabilities that enable mission accomplishment. A challenge for the system of systems (SOS) engineer is that while the SOS is being developed, the operating environment it was designed for changes. This situation causes significant uncertainty as to whether the SOS will meet the needs of the organization when finally deployed. To mitigate this uncertainty, SOS architectures should by assessed for their ability to deploy in more than one configuration. Past architecture assessments and performance characteristics are primarily system focused and do not address the dynamics of the interacting constituent systems of the SOS. This dissertation provides two measures, Adaptability and Agility, for assessing and comparing SOS architectures for their ability to adapt to the current operating environment and their ability to provide multiple capabilities concurrently. A SOS is defined as being composed of individual Elements that can be organized into Nodes. Each Element belongs to one and only one Node. Cohesion is a measure of the relatedness of the Elements within a Node. Coupling is a measure of the interdependence among the Nodes. Adaptability is defined as the ability of a SOS to respond to changes in the allocation of Elements to Nodes; it is computed using the concepts of Coupling and Cohesion. The Degree of Reuse measures the extent to which Elements support multiple capabilities. Agility measures the ability of the SOS to execute multiple processes concurrently and adapt to changing situations. Agility is a function of Adaptability and Degree of Reuse. The methodology provides the information required to assess the Adaptability and Agility of a proposed or actual SOS architecture. The process begins by identifying from the operational view of the architecture the capabilities that must be realized by the SOS alternatives. The SOS architecture describes how a particular subset of Elements organized into Nodes will realize the capabilities; it is the system view of the architecture. The SOS is transformed automatically into an executable model using Colored Petri Nets; invariant analysis and simulation are used to compute Coupling and Cohesion, while the Degree of Reuse is computed directly for each SOS. These three measures are then used to compute the Adaptability and Agility measures. Alternative architecture patterns are then compared in terms of their adaptability and agility. One advantage of the approach is that it can be applied early in the systems engineering process to help select preferred architecture alternatives. A case study is presented to illustrate the application of the assessment methodology and that different architecture types or patterns yield distinct values for the Adaptability and Agility measures that are consistent with the qualitative differences in the tested architectures.