Architecting federations of systems: A framework for capturing synergy

Abstract

This paper defines Federations of Systems, introduces a tailored architecting framework, and demonstrates it in a satellite design case study. While certainly not being new concepts in the literature, federations are a class of Systems of Systems (SoS) that have not been extensively explored due to the scarcity of applied examples and still few practitioners facing their specific design challenges. This is expected to change with the advent of the Internet of Things, Machine to Machine applications, collaborative consumption, and Federated Satellite Systems. It has become henceforth necessary to develop approaches in design and architecting capable of addressing such interconnected sets of systems. Federations are SoS characterized by the absence of global goals and stakeholders, and cooperation enabled only on an opportunistic basis. Systems engage in federations when the perceived benefit exceeds the costs of the necessary interfaces. The proposed framework quantifies the upper limits to the net benefits of cooperation between systems through the concept of synergy. Using synergy as a heuristic to navigate federated design spaces, the framework predicts the appearance and evolution of federations. The demonstration of the framework in a six‐satellite federation shows that advantages of federating emerge with the collaboration between just two systems. Interface costs, architectural cost preferences, and network topology drive the achievable benefits. Specifically, when considering communication architectures under 80 MUSD for satellite systems, federating is advantageous when the interface lifecycle costs are below 15 MUSD. This demonstration case exemplifies how federating can assist systems in achieving design points that dominate the solutions of their local tradespace.

The strategy dynamics of collective systems: Underlying hindrances beyond two-actor coordination

“…Collaborative decision-making plays a pivotal role in facilitating information exchange and coordinating the activities of individual actors to guide the system toward optimized performance. However, despite the potential benefits of coordinated collective action, its seamless execution faces structural challenges related to the combinatorial complexity of aligning individual objectives and incentives [ 2 ]. These hindrances go beyond localized actor-to-actor coordination challenges and encompass intricate dynamics that result from the interplay of many subsystems, direct and indirect stakeholders, and decision-making agents.…”

Towards Federated Satellite Systems and Internet of Satellites: The Federation Deployment Control Protocol

“…In this context, a satellite could communicate with a ground station using the downlink contact of another satellite, increasing thus the overall downlink opportunities. The Federated Satellite System (FSS) concept [11,12] aims at addressing this situation by promoting collaborations between satellites to share unused resources, such as memory storage or downlink opportunities. These collaborations, called federations, are established sporadically and opportunistically depending on the availability of the resources to be shared.…”

In pursuit of autonomous distributed satellite systems

“…FSS can be seen as an alternative to fractionated spacecrafttechnically more feasible because it could be implemented with readily available technology. Satellite federations mostly rely upon inter-satellite networking and coordination mechanisms to create virtual, opportunistic infrastructures with emergent functionalities (Lluch, 2017). A parallel concept is that of cellularised systems (Kerzhner et al, 2013), where satellite units-or "satlets"-are virtually aggregated to spawn new, and more complex, systems.…”

“…A generic understanding of distributed satellite architectures has been introduced revealing some of their operational traits. Seeking to achieve the emergent capabilities analysed in (Corbin, 2015) and (Lluch, 2017), many of the DSS approaches present structural functions that could allow them to implement complex communication networks-also for EO applications-, maintain coordinated flight formation, or distribute computational load among spacecraft. Most of these structural functions rely upon a number of enabling technologies (Guo et al, 2009).…”

“…In SE, the expected utility of a system is essentially used to capture stakeholder preferences or value of a design (Ross et al, 2010). The fundamental goal of utility is to transform such preferences into a numerical figure, while preserving ordinality and transitive properties (Lluch, 2017). Therefore, for any given quantifiable attribute in the system, a utility function U (•) is provided which, for convenience and simplicity, encodes stakeholder satisfaction in the range [0, 1].…”