Load- and interference-aware dynamic channel assignment in IEEE 802.11 based wireless mesh networks

Abstract

Channel assignment aims to reduce interference by carefully assigning a channel to each radio. The dynamic channel assignment based on internal usage can improve the throughput of multi-channel, multi-radio wireless mesh networks by selecting the least used channel, thereby resulting in a low level of interference. However, simply assigning the least-used channel among mesh nodes is not sufficient to reduce the interference in a network. This is because today’s numerous urban areas around the world are covered by dense 802.11 Wi-Fi networks which may operate on the same channel used by mesh nodes, generating significant external interference. As a result, even though channel assignment schemes choose the least-used channel for a radio, if that channel is being heavily utilized by external nodes, it may not be the best channel. Therefore, relying exclusively on the dynamic channel assignment considering only internal usage can significantly degrade network throughput due to the external interference from co-located wireless network. In this paper, we proposed a distributed dynamic channel assignment algorithm which takes both the internal usage and external interference in a wireless mesh network into account. However, accounting for the external interference in channel assignments is a challenge since all external nodes do not allow internal mesh nodes to acquire their traffic information and do not cooperate with internal mesh nodes. Since a mesh node does not have information about which channel the external nodes are on, each mesh node has to monitor all the available channels to obtain the traffic information of the external nodes. However, monitoring all the available channels by channel switching can cause the flow disruption of current data traffic and performance degradation. To avoid these problems, the proposed algorithm does not perform channel switching to monitor external interference but the cooperation between the mesh nodes within a neighborhood...