Impact of Terrain Resolution on Analytic Modeling of Individual Soldier Target Detection

摘要

During the last ten years analytic individual combatant modeling and simulation users have continuously demanded higher and higher resolution terrain databases. Tremendous resources have been dedicated to constructing a limited number of these higher resolution terrain databases. The time and investment associated with developing and using these databases in both visual and Computer Generated Forces (CGF) applications has caused a number of managers to question the need for terrain databases of this resolution.In connection with a parametric analysis of the effects of individual combatant signature reduction, a comparison was conducted of the impacts on simulation execution and results for 2 different resolution terrain databases for the same geographical area. Both terrain databases were version 5 Compact Terrain Database (CTDB). The combat scenario involved an overcast starlight battle between 2 attacking infantry squads and 1 or 2 defending squads. The simulated combat took place over a 160 meter range and lasted 6 minutes. One terrain database had a spatial resolution of 10 meters while the other terrain database had a spatial resolution of 110 meters (DTED Level 1).The metrics that served as the basis for the comparison were numbers of line of sights (LOS) that existed, the number of targets detected and the time to complete a simulation iteration. Additional data about shots fired, targets hit, casualties inflicted were also collected but not analyzed for this comparison.The simulation that was used for the comparison was the Integrated Unit Simulation System (IUSS), a Windows95/NT based engineering level simulation designed to support individual combatant trade off analyses. 60 simulation iterations were executed for each attacking versus defending combination. A 266MHz Pentium Pro computer was used to execute the simulation.The comparison found that while the higher resolution terrain database requires a significantly longer period to complete the 60 replications, 5 hours versus 45 minutes, there was minimal difference in the percent of targets detected, .4% versus .5%. There was an increase in the average number of LOS that existed in the higher resolution terrain databases when compared to the lower resolution terrain databases, 3630 versus 2859.More detailed comparisons of this type are planned. These comparisons will look at different geographical locations and seek to isolate the factors that have the greatest impact on the simulation outcomes.