Models and algorithms for determining inter-unit network demand

Jeffrey P. Ridder; Samuel W. Brett; Craig M. Burris; Jimmie G. McEver; Jack E. O'Donnel; David T. Signori; Heather W. Schoenborn

doi:10.1117/12.918448

Translator Disclaimer

3 May 2012 Models and algorithms for determining inter-unit network demand

Jeffrey P. Ridder, Samuel W. Brett, Craig M. Burris, Jimmie G. McEver, Jack E. O'Donnel, David T. Signori, Heather W. Schoenborn

Author Affiliations +

Proceedings Volume 8405, Defense Transformation and Net-Centric Systems 2012; 840503 (2012) https://doi.org/10.1117/12.918448
Event: SPIE Defense, Security, and Sensing, 2012, Baltimore, Maryland, United States

Abstract

Analysts often use inter-unit demand for communication services as the basis for assessing network performance and the impact on mission effectiveness. Traditional methods base inter-unit demand on Information Exchange Requirements (IER's) most often derived from a variety of disparate sources that can result insignificant limitations. This paper describes models and algorithms that enable automated support for the challenging steps of tailoring the data from an established unit demand database in order to derive the inter-unit demand for specific scenarios. Such a capability is referred to as "demand parsing." The necessary operational constraints are modeled by applying an organizational distance metric, using weights associated with a small set of functionally driven usage patterns, to a node-link structure established at a level of resolution appropriate for the analytical context. An innovative agent based algorithm is applied to address the resulting multi-objective optimization problem by calculating solutions that satisfy both the operational constraints and those imposed by the unit demand. Using an agent based paradigm, the operational model and the algorithm were combined into a prototype tool that was applied within the parsing process to estimate the inter-unit demand for communications supporting units in a realistic air operation. The peak errors in meeting both types of constraints were found to be less than 20%. These levels are consistent with the errors in the unit database intended for first order assessments.

Citation Download Citation

Jeffrey P. Ridder, Samuel W. Brett, Craig M. Burris, Jimmie G. McEver, Jack E. O'Donnel, David T. Signori, and Heather W. Schoenborn "Models and algorithms for determining inter-unit network demand", Proc. SPIE 8405, Defense Transformation and Net-Centric Systems 2012, 840503 (3 May 2012); https://doi.org/10.1117/12.918448

ACCESS THE FULL ARTICLE