全球智库信息集成服务系统

Orchestration rules are crucial to the overall performance of Mosaic architectures

• Mosaic architectures can perform better than more-monolithic architectures across a variety of target demands but only with high-performing, robust orchestration rules.
• The best choice of rules is highly dependent on both platform and target factors. This dependence is difficult to predict in advance of an operation.

As the U.S. Department of Defense transitions from a focus on irregular warfare to great-power competition, several new approaches to fighting conflicts are under consideration to reduce costs and increase effectiveness and robustness. The Defense Advanced Research Projects Agency is investigating a warfighting construct known as Mosaic warfare, after the idea of creating a complex image out of many small, simple pieces. This approach relies on fractionation of capabilities from large multicapability platforms onto multiple smaller ones; the ability to employ heterogenous mixes of capabilities throughout a battlespace; and, finally, the ability to rapidly compose a set of needed capabilities in a time and place to accomplish a mission.

This report presents the benefits of fractionation, heterogeneity, and rapidly composable forces, by means of modeling and simulation in the context of servicing targets through a reduced-order, agent-based model. The performance of monolithic platforms (which contain all available capabilities) is compared with fractionated platforms (which contain a subset of the available capabilities) in various target environments. The report summarizes (1) that fractionation is beneficial in that it can increase the operational tempo in the face of simple targets (those that require a small number of capabilities) and (2) when those independent platforms can be effectively orchestrated to tackle more-complex targets. The authors examine several methods for orchestrating these capabilities and show that the problem of orchestration is a critical need for future investment to enable distributed autonomous system architectures, such as Mosaic warfare.

• Chapter One

  Introduction

• Chapter Two

  Agent-Based Modeling Approach

• Chapter Three

  Results

• Chapter Four

  Observations