Bandwidth consumption efficiency using collective rejoin in hierarchical peer-to-peer

Having the advantage of scalability and efficiency in performing a succesful lookup query, a hierarchically structured P2P is a promising architecture for heterogeneous networks. However, in addition to potentially decrease the system performance, the presence of a superpeer failure event also forces normal peers under its influence to disconnect from the system. Optimally, the disconnected normal peers should perform rejoin, otherwise they will loose the granted access to the required service. Currently, the most common rejoin algorithm is based on the flat Chord maintenance algorithm, which is an individual rejoin. Addressing the high bandwidth consumption in individual rejoin, we propose a new approach, termed the collective rejoin algorithm. The analytical results, as well as simulation outputs, show that the rejoin process using the collective rejoin algorithm produces less traffic load than the individual rejoin algorithm.