A load balanced two-tier DHT with improved lookup performance of non-popular data items
GPC'11 Proceedings of the 6th international conference on Advances in grid and pervasive computing
A framework for interest-based community evolution and sharing of latent knowledge
International Journal of Grid and Utility Computing
Expressive Query Support for Multidimensional Data in Distributed Hash Tables
UCC '12 Proceedings of the 2012 IEEE/ACM Fifth International Conference on Utility and Cloud Computing
Hi-index | 0.00 |
In this paper, we propose a two-tiered Distributed Hash Table (DHT) based overlay for an implementation of publish/subscribe. In this overlay, upper tier consists of stable nodes with more resources whereas lower tier contains less stable nodes with limited resources. In our design, all nodes are required to join at lower tier and nodes with better uptime and resources may dynamically migrate to upper tier. Further, taking the advantage of prevalent skewness in subscription patterns, popular subscriptions are cached at upper tier nodes, which leads to better lookup efficiency. Overlay construction and maintenance costs in our scheme are also less compared to flat implementations as we are required to update finger tables and successor lists of upper tier nodes with less frequency. Experimental and analytical results indicate that our proposed infrastructure scales well with increasing number of nodes, suggesting its applicability for large scale distributed applications.