Accelerating of line of sight analysis algorithms with parallel programming

Yılmaz, Gökhan
Line of sight (LOS) analysis is a set of methods and algorithms to determine the visible points in a terrain with reference to a specific observer point. This analysis is used in simulations, Geographic Information System (GIS) applications and games. For this reason, it is important to have a capability to get results quickly and facilitate analysis in such a way that the interaction with the changing reference points is possible. Van Kreveld, R2 and R3 are the most frequently used algorithms in line of sight analysis. The purpose of this research is to develop parallel adaptations of these particular algorithms by making use of the capabilities of a modern Graphics Processing Unit (GPU) and to evaluate these adaptations in terms of performance and memory usage. By analyzing which algorithm is more suitable to be implemented on the GPU, the algorithm that will provide the most appropriate and quick solution to the probing problem can be determined. In this research, Van Kreveld's algorithm, which is basically a sequential algorithm, was developed partly in parallel, and the speed-up was 1.5x compared to the sequential version of the algorithm. Speed-up rates increase up to 10.5x for R2 and 160x for R3 algorithms, respectively. The results can be used to combine CPU / GPU approaches in order to perform hybrid or full parallelization of Van Kreveld’s Algorithm on the GPU. The results presented in the thesis will serve as a guide for the selection of the appropriate algorithm by evaluating the strengths and weaknesses of different algorithms.