Self-Organized Networks

During recent years the number of WiFi networks has experienced rapid growth. The evergrowing number of wireless communications systems have made the optimal assignment of a limited radio frequency spectrum a problem of primary importance. With the common 802.11 wireless technology, few non-overlapping channels are available, and there is no standard mechanism for the access points to dynamically select the channel to be used in order to minimize interference with other access points. This has resulted in a situation where many WiFi networks use default or suboptimal channel assignments, leading to poor performance, and uneven spectrum usage. In this paper, we introduce three different approaches for the frequency assignment problem based on graph coloring that can significantly improve frequency distribution and reduce the number of collisions in the network