Research On Dynamic Channel Allocation Algorithm Of Adaptive Traffic In WLAN

Wireless LAN as the product of the combination of the computer network and wireless communication technology has been rapid developed. As the complement of cellular networks, it strongly supports the mobile and portable communication. With the widespread popularity of Wireless LANs, many enterprises and campuses are in the cover scope of wireless network.WLAN operates in the2.4GHz ISM band, and the number of channels is very limited. The widespread popularity of WLAN network means that the access points (APs) density increases. Because of the broadcasting characteristics of wireless communications, high density distribution of AP will have a serious interference and competition, so the throughput of the overall network will decrease. With the continued growth of deployed WLANs and network users, the interference will increase seriously. So it needs the right channel allocation to minimize interference.WLAN has two kinds of channel allocation algorithm. They are Variable-bandwidth and Fixed-bandwidth. This paper studies the two kind’s algorithm and proposes a corresponding improvement algorithm for them.The performance of the channel allocation algorithm based on Game-theory and Coloring-theory is analyzed in this paper. In order to overcome the shortage of them, the adaptive flow dynamic channel allocation algorithm based on Game-theory and Coloring-theory are studied. The principle and flow of improved algorithm are introduced in detail, then the performance compare are made. It finds that the proposed algorithm improves the system throughput and fairness in the dynamic network environment. The improved algorithm based on Game-theory has the better fairness of the system than the improved algorithm based on Coloring-theory in any scene. In the larger network, the improved algorithm based on Coloring-theory has better throughput than the improved algorithm based on Game-theory. But in smaller networks, the improved algorithm based on Game-theory has better throughput than improved algorithm based on Coloring-theory.