We present a novel bimanual body-directed travel technique, PenguFly (PF), and compare it with two standard travel-by-pointing techniques by conducting a between-subject experiment in a CAVE. In PF, the positions of the user's head and hands are projected onto the ground, and travel direction and speed are computed based on direction and magnitude of the vector from the midpoint of the projected hand positions to the projected head position. The two baseline conditions both use a single hand to control the direction, with speed controlled discretely by button pushes with the same hand in one case, and continuously by the distance between the hands in the other case. Users were asked to travel through a simple virtual world and collect virtual coins within a set time. We found no significant differences between travel conditions for reported presence or usability, but a significant increase in nausea with PF. Total travel distance was significantly higher for the baseline condition with discrete speed selection, whereas travel accuracy in terms of coin-to-distance ratio was higher with PF.

The knowledge of the propagation behavior of radio waves is a fundamental prerequisite for planning and optimizing mobile radio networks. Propagation effects are usually simulated numerically, since real-world measurement campaigns are time-consuming and expensive. Automatic planning algorithms can explore a vast amount of network configurations to find good deployment schemes. However, complex urban scenarios demand for a great emphasis on site-specific details in the propagation environment which are often not covered by automatic approaches. Therefore, we have combined the simulation of radio waves with an interactive exploration and modification of the propagation environment in a virtual reality prototype application. By coupling real-time simulation and manipulation tasks we can provide an uninterrupted user-centered workflow.

Conventional beam tracing can be used for solving global illumination problems. It is an efficient algorithm and performs very well when implemented on the GPU. This allows us to apply the algorithm in a novel way to the problem of radio wave propagation. The simulation of radio waves is conceptually analogous to the problem of light transport. We use a custom, parallel rasterization pipeline for creation and evaluation of the beams. We implement a subset of a standard 3D rasterization pipeline entirely on the GPU, supporting 2D and 3D frame buffers for output. Our algorithm can provide a detailed description of complex radio channel characteristics like propagation losses and the spread of arriving signals over time (delay spread). Those are essential for the planning of communication systems required by mobile network operators. For validation, we compare our simulation results with measurements from a real-world network. Furthermore, we account for characteristics of different propagation environments and estimate the influence of unknown components like traffic or vegetation by adapting model parameters to measurements.

In this paper we present a GPU implementation to accurately select 3D objects based on their silhouettes by a pointing device with six degrees of freedom (6DOF) in a virtual environment (VE). We adapt a 2D picking metaphor to 3D selection in VE's by changing the projection and view matrices according to the position and orientation of a 6DOF pointing device and rendering a conic selection volume to an off-screen pixel buffer. This method works for triangulated as well as volume rendered objects, no explicit geometric representation is required.

Fast radio wave propagation predictions are of tremendous interest, e.g., for planning and optimization of cellular radio networks. We propose the use of ordinary graphics cards and specialized algorithms to achieve extremely fast predictions. Our implementation of the empirical COST-Walfisch-Ikegami model allows the computation of several hundred predictions in one second in a 7 km urban area. Further, we present a ray-optical approach exploiting the programming model of graphics cards. This algorithm combines fast computation times with high accuracy.