Accelerating Wireless Network Simulation in 3D Terrain Using GPUS
Leonid Djinevski, Sonja Filiposka, Igor Mishkovski and Dimitar Trajanov
In this paper we present a parallel implementation of the Durkins radio propagation algorithm on Nvidia GPU devices. We have modified the extension of the NS-2 network simulator, by developing a transparent module which uses the Durkins parallel algorithm for simulation of wireless mobile networks over 3D terrains. When evaluating the performance of the parallel implementation of the Durkins algorithm compared to our sequential implementation, the results show that we managed to achieve x118 maximal speedup. For the overall NS-2 network simulation using our parallel module, we obtained an overall maximal speedup of x67.
Keywords: GPGPU, OpenCL, speedup, network simulators, Durkins radio propagation model.