Abstract:
Cooperation in wireless content distribution networks has been proposed as an effective approach to improve system performance in terms of bit rate enhancement and energy consumption reduction. To enable such cooperation, wireless nodes need to be grouped into clusters; in each cluster, a node is selected as a cluster head and acts as a relay between the access point and other nodes. In this thesis, we address the problem of clustering in wireless content distribution networks with intelligent cooperation among nodes. We propose and analyze three clustering methods based on practical system parameters such as estimated signal strength levels among the nodes and the number of neighbors in the vicinity of each node. Simulation results are presented for various network scenarios to demonstrate the characteristics of the proposed approaches and compare their performance gains. In addition, we present and analyze via simulations a practical approach to detect a class of spoofing attacks on the proposed cluster formation methods in wireless content distribution networks.
