Abstract
Analysis of military campaigns through the use of computational models and simulations, is one of the fundamental methods used within defence Operational Analysis at the Defence Science and Technology Laboratory. It helps to develop understanding behind the value of investment so that an agile defence capability can be built and maintained; in order to face the challenges of an ever-changing world. However, many of these models have been used and adapted continuously over decades, resulting in code-bases that have become unmaintainable in the face of constrained budgets. To address this problem, a number of software modelling framework, based upon the reuse of code and concepts have been developed. However, many of these ultimately did not achieve their full potential, because they merely iterated upon the same software patterns which had been used to produce models to-date. The most recent attempt however, known as the Generic Aggregator Model Valuator, was very different in this regard, due to its exploitation of the emerging Object-Functional design paradigm.As an emerging design paradigm, Object-Functional is still to be more formally understood. There is currently a lack of widely accepted design patterns for the paradigm and explicitly acknowledged examples of its use in projects. This thesis examines the Object-Functional paradigm in greater detail, by performing a qualitative evaluation of defence models built using the paradigm compared to extant models that use other approaches such as Object-Orientation. The evaluation aims to answer two key questions: what benefits does the exploitation of this paradigm bring to defence modelling? But also, what challenges? The Generic Aggregator Model Valuator frameworkâs implementation of the paradigm is also presented in detail, illustrating the patterns it uses, standing as an example that can contribute to the further refinement of this paradigm in the future.
Date of Award | May 2019 |
---|---|
Original language | English |
Awarding Institution |
|
Supervisor | Rich Boakes (Supervisor) |