AZB Rectangle Shrinkage Method and Heterogeneous Computing Accelerated Full Image Theory Method Ray Tracing Enabling Complex and Massive Outdoor 6G Propagation Modeling

Abstract: Until now, despite their high accuracy, the utilization of the conventional image theory method ray tracers was limited to simple simulation environments with small number of field observation points and low maximum ray bouncing order due to their poor computational efficiency. This study presents a novel full-3D AZB rectangle shrinkage method and heterogeneous computing accelerated image theory method ray tracing framework for complex and massive outdoor propagation modeling. The proposed framework is divided into three parts: 1. Visibility preprocessing part. 2. Visibility tree generation part: in this part, a novel AZB rectangle shrinkage method that accelerates and reduces generation speed and size of visibility tree is proposed. 3. Shadow testing and field calculation part: in this part, a heterogeneous computing algorithm that can make possible to handle a large amount of field observation points is proposed. It is demonstrated that the proposed framework is faster more than 651 times than the image theory method solver of WinProp. Also, it is confirmed that the proposed ray tracing framework can handle 1km x 1km wide and dense urban outdoor simulation with up to the maximum ray bouncing order of 6 and thousands of field observation points. The proposed ray tracing framework would be a cornerstone of future image theory method ray tracing techniques for complex and massive scenarios that was exclusive to the shooting and bouncing rays method ray tracers.