International C2 Journal: Issues
Vol 2, No 2
Guest Editor’s Introductory Remarks
This Special Issue focuses in on one of the most intellectually challenging aspects of Defence Modelling. Across NATO, and more widely, nations are considering the benefit and cost of enhancing their ability to deliver higher levels of Network Enabled Capability or Network Centric Operations. At the core of this capability objective is the hypothesis that increasing levels of information, and information sharing, will provide significant improvements in our ability to prosecute conflict within warfighting or stabilisation operations. The ‘transfer function’ which converts such information into operational outcomes is centred on the ability of human commanders to take the right decisions, and deploy their forces in effective ways, given the sensors, information sharing and shared awareness which they are provided with. Within limited defence budgets, it is the role of the operational analyst to give advice on how much of such Network Enabled Capability is enough, or the best way of proceeding, given a finite amount of resource.
In a previous paper published in this journal I sketched out the approach we have taken in the UK. Following on from this, it seemed appropriate to put out a call for papers to the community to stimulate further thought and work. The result is the special edition which we have now put together. I hope it will help in some small way to move forward the state of the art and the state of the practice in this important domain.
I am encouraged by the fact that many of the aspects of the ‘benefits chain’, from collecting information, to developing shared awareness and creating agility, are yielding to representation in mathematical form in our simulation models. The six papers in this Special Edition, described in summary below, exemplify this trend.
The first paper (Kewley and Embrechts; ‘A Multiagent System for Tactical Control of Automated Forces’) focuses on the ‘transfer function’ converting information into decision and course of action. It draws on ideas from Complexity Theory in order to develop algorithms for decision-making agents within a constructive simulation environment. Numbers of different types of agent develop plans and actions in a networked fashion in order to evolve plans and routes which are adaptive to a potentially rapidly changing enemy. This paper illustrates in particular the great utility of methods based on genetic algorithms.
The second paper (Nehme, Mekdeci et al; ‘The Impact of Heterogeneity on Operator Performance in Future Unmanned Vehicle Systems’) looks at the gathering of information in the future battlespace using a ‘swarm’ of unmanned robotic vehicles, and how they can be easily controlled by shifting the ‘balance of intelligence’ from the human controller to the machine. This leads to the development of queueing theory based algorithms which represent this shift in balance within a constructive simulation modelling environment, taking account of human effects such as situation awareness and the ability to switch easily between tasks.
In paper three (Hossain and Walmsley; ‘A Minimum Spanning Tree Approach to Identifying Collective Behaviour and Inferring Intent for Combat Models’), the focus is kept on the collection of information. The key question at issue here is how we make sense of this information in order to infer enemy intent, when enemy forces are in very irregular formations characteristic of the complex endeavours and asymmetry relevant to the 21st Century. A set of algorithms have been developed to do this, based on a minimum spanning tree approach which , from a full set of interconnections, prunes out certain links in order to produce a number of coherent groupings of purposeful subgroups, and their likely intents (in terms of their general heading). These algorithms have been embedded and tested in a constructive simulation environment.
The fourth paper (Hiniker; A View of the Combat CAS: Unifying Net-Enabled Teams’ ) looks at the Network Centric Operations/ Network Enabled Capability benefits chain (from sharing information, through sharing awareness to agility) through the lens of complexity. The paper then develops a number of ideas and approaches to both capturing the elements of this chain, and quantifying their benefit. A key idea is the development of shared ‘schema’ (as proposed by Murray Gell-Mann) based on a shared and common operational picture.
Paper five (Sprague and Dobias; ‘Modelling the Complexity of Combat in the Context of C2’) illustrates the benefit of using agent base distillations, which have a highly abstracted representation of the force elements, to explore wide ranges of scenario space in order to understand force level emergent properties. In particular, the paper discusses the benefit of complexity based metrics coupled with genetic algorithms, allowing agent adaptation, to improve the command decisions made within such modelling environments.
In the final paper six (Dean, Vincent et al; ‘Representing a Combat ID Analysis Tool within an Agent Based Constructive Simulation’) a case based reasoning approach is successfully used to develop a set of algorithms which capture the combat ID process (identification of friend or foe) at the individual agent level, including the influence of many human factors on the identification process. The paper then illustrates the benefit of using an agent based modelling environment to test out the various options for embedding such a set of algorithms within a constructive simulation.
References
J Moffat (2007) ‘Modelling Human Decision-Making in Simulation Models of Conflict’ The CCRP International C2 Journal vol 1 no 1, pp 31-60.