A Permission based Multilevel Parallel Solution for Distributed Mutual Exclusion

Due to the growing application of peer-to-peer computing, the distributed applications are continuously spreading over an extensive number of nodes. To cope with this large number of participants, various cluster based hierarchical solutions have been proposed. Cluster or group based solutions are scalable for a large number of participants. All of these solutions exploit the idea of coordinators, leaders or proxies of the clusters. If any such node fails, the election of a new one is required. Thus, fault tolerance of these algorithms is low. Again, as the number of participating nodes increases every day, it is necessary to devise highly scalable distributed mutual exclusion algorithms. This research presents a permission based parallel solution of distributed mutual exclusion by modeling a multilevel clustered network, where clusters are formed at different hierarchies. This technique enhances the scalability by reducing the cluster size, as it requires consensus from only one cluster at each level. As the algorithm has no use of coordinators, it possesses high fault tolerance. The paper also addresses the problem of achieving optimal level of clustering in a network for distributed mutual exclusion.