Interest-based content delivery in wireless mesh networks with hybrid antenna mode

Abstract

Driven by the great performance improvement using flexible combination of omnidirectional and directional antennas, wireless mesh networks (WMNs) with hybrid antenna mode have become a promising technology for providing users with better network service. Since multimedia contents account for most of the network traffic, content delivery is a big issue when designing network protocols. By combining the use of both directional and omnidirectional antennas, the performance of content delivery in WMNs can be improved. Currently, research works have shown that the relationship between nodes' interests in the network reveals the small-world pattern, so we can analyze nodes' interests and design optimal mechanism to deliver contents based on their interests. In this paper, we use the interest-sharing graph to capture the similarities between nodes' interests. Based on the interesting-sharing graph, we cluster nodes in the network using graph partitioning algorithm. The purpose of clustering is to make nodes in the same cluster share similar interests. After clustering nodes, we design a content delivery mechanism in WMNs which takes advantage of both omnidirectional and directional antennas. The simulation results show that our proposed content delivery mechanism can adapt to the WMNs and performs better than using omnidirectional or directional antennas only.