Search

Examples from the area of combat systems simulation are used to illustrate the advantages of using precursory analyses to examine simplified versions of systems or selected components, often in a reduced space. These analyses may yield valuable insights resulting in the identification of important areas and reduction in the number of trials. In some circumstances, the analysis leads to the adoption of the simplified model in place of the originally proposed more complex one. The presumption that the conclusions of these analyses have relevance to the real systems and extended simulations has been exonerated in practice. The techniques used are not specific to combat systems nor to the system size but encompass a wide range of simulation categories and scales. Examples quoted in the paper include a number of novel, previously unpublished solutions.

The use of information entropy as a quantitative measure of uncertainty for radar resource management and control objectives is explored and applied to issues of current interest. It is seen to be rigorous and objective, and therefore potentially superior to heuristic, rule-based approaches for problems which can be formulated in probabilistic terms. It is particularly appropriate for sensor systems in general which have as their objective the acquisition of information, but which are dominated by uncertainty and subject to time and resource constraints. Examples of the application of this control and management methodology are given to radar problems on widely differing time-scales: the scheduling of track updates in a beam-agile radar (ms), and the tasking of a constellation of SAR surveillance satellites for maritime search and tracking (hours).

Maximal exploitation of space-based surveillance resources is unique due to the nature of the platforms, their sensors and intermittent communications links to ground stations for target information processing and sensor tasking. An optimal formulation of the problem of integrating a range of disparate surveillance sensor systems based on Bayesian inference and entropy measures of eflectiveness for dealing with decision making under information uncertainty forms the basis for this paper. Here a general methodology is proposed for planning, tasking and control of surveillance assets by formulating surveillance as a stochastic optimal control problem, which has as its objective the minimisation of uncertainty associated with the information collected, subject to surveillance system time and resource constraints, and sensor imperfections. The approach presented provides a rational basis for surveillance asset optimisation that sets information as the objective, incorporates platform anamics and constraints, semor capabilities, stochastic eflects and control variables within a single formulation.