Histogram Based Live Streaming in Peer to Peer Dynamic Balancing & Clustering

We made a preliminary clustering analysis of an investiga measure research within Planet Lab. The application was an Internet architecture. Our clustering is inspired by the Regularity. Such approach was already demonstrated in appeared to be a powerful tool. Live streaming result suggests that the nodes of a large enough graph can be partitioned in few clusters in such a way that link distribution between most of the pairs look like random. Our main goal is to study what this type of clustering can tell us about p2p systems using our investigational system as source of data. We searched clustering’s of clustering type by using max like hood as leadership. Our graph is directed and weighted. The link direction designates a client server relation and the value is the proportion of all chunks obtained from such a link during the whole experimentation. We think that the first results are motivating. Most of the cluster pairs have very great patterns of link distribution, organizing peers effectively. The values of weights between clusters and their patterns. We end up with cluster pair