However, faster execution rate is necessary in applications where real-time interactivity is mandatory. Enabling interactions with multiple hardware devices while maintaining expected realism for both visualization and simulation require complex algorithms and considerable CPU time. Multimodal interactions require interfaces such as mouse, trackballs, haptic and tracking devices.
These may include realistic visual rendering engine to display photo-realistic images and physics engine to simulate real world phenomena including rigid body motion, deformation of solids, fluid flow, and multi-physics phenomenon. MIS systems especially virtual reality applications require multiple components to work synergistically to achieve a high degree of immersion. Such systems have wide applications in medical training, flight simulation, video gaming, movies and other industries.
Multimodal interactive simulations (MIS) are highly demanding in terms of computational resources. Test results are presented on multiple hardware platforms with diverse computing and graphics capabilities to demonstrate the effectiveness of our approach. texture size, canvas resolution) and simulation parameters (e.g. The optimum solution is used for rendering (e.g. This data is utilized in conjunction with user specified design requirements in the optimization phase to ensure best possible computational resource allocation. In the identification phase, the computing and rendering capabilities of the client device are evaluated using an exploratory proxy code. Our approach includes three distinct phases: identification, optimization and update. Hence, we present a mixed integer programming model to optimize the performance of graphical rendering and simulation performance while satisfying application specific constraints. However, optimization of simulation performance on individual hardware platforms is not practical. For such applications where visual quality and the performance of simulations directly influence user experience, overloading of hardware resources may result in unsatisfactory reduction in the quality of the simulation and user satisfaction. This paper presents a technique for optimizing the performance of web based multimodal interactive simulations.