Parallel algorithms for wireless LAN planning

Abstract

Nowadays wireless services are a part of the everyday life of billions of people around the world. To decrease not only their design and maintenance costs but also to enhance user comfort there is a high demand for specialized software that helps during theirs designing. Unfortunately commonly available and used software is demanding in terms of computing power in order for receiving the results in the reasonable time. The aim of the presented research is verification whether using GPU it is possible accelerate computation in reasonable way. The paper presents a study of possible parallelization of algorithms used for signal power simulation as well as algorithms that optimize placement of routers. The main goal of the paper is to show the impact of GPU usage on different stages of wireless network planning process. The first received results are very promising, the speedup received is even up to 46 for simulation algorithm and up to 15 for optimization algorithms.