Self-Organizing Content Distribution in a Data Indexed DHT Network

The paper discusses the access-skew problem in a data indexed DHT network. The difference in popularity of data indexes is likely to create access hotspots that overwhelm the host node. The capability of relieving heavily loaded nodes is essential to the scalability of an overlay network. In this paper, we propose an effective content distribution method to relieve hotspots in a self-organizing manner. An algorithm is given to reduce the query traveling distance and balance the load among replica nodes. Moreover, we give a small-world model based algorithm to reduce the load-balancing cost. Experiment results show that the mechanism significantly reduces the number of dropped queries and the average query delay while keeping load-balancing cost under control