Analysis of delays in networked flight simulators

Scope of study. Networking of manned-flight simulators over wide geographical areas have recently become feasible because of the great advances in the technologies of microprocessors, local area networks, and long haul networks. One of the key questions to be answered during the development of a simulator network is "How much delay can be tolerated in inter-simulator communication?" In order to analyze the networked flight simulator problem, it is necessary to have a mathematical model of the pilot. One of the most useful approaches for modeling humans performing manual control tasks is the Optimal Control Model (OCM). This study describes how the OCM concept can be used in the analysis of some networked flight simulator experiments.; Findings & conclusions. The first phase of this study investigated the effect of the length of the inter-simulator delay on pilot performance in a simple formation flight task. There were several interesting, and unexpected, results from the analysis of this task. The first was that the tracking error is not a monotonically increasing function of the time delay between the two simulators, as might be expected. The error perceived by the second pilot decreases initially as the inter-simulator delay is increased. After reaching a minimum, the error then increases. In addition, the actual error (by-passing the inter-simulator delay) has the opposite pattern. It was also found that the point at which the perceived error reaches a minimum will increase as the intra-simulator delay of the second pilot increases, but will decrease as the intra-simulator delay of the first pilot increases. The second phase of this study investigated the feasibility of applying differential games theory to modeling of an air-to-air combat task by considering the possibility of combining the optimal control model and differential games for a simple task. A missile-aircraft problem in which the aircraft (evader) is given noisy delayed information was analyzed. Again, we observed that the mean squared terminal miss (MSTS) as a function of inter-simulator delay is non-monotonic; it initially increases and will reach a maximum before decreasing. In addition, MSTS will increase as the intra-simulator delay of the evader increases. These trends are exactly the same trends observed for the actual error in formation flight task problems.