An Overlay network construction scheme based on node location in physical Network

ABSTRACT Big data mining provides valuable and important information, not only in academic but also in industrial, business, science and government applications. To manage such data, data mining provides parallel data mining approach. The proposed method is the new approach for the neighbor node selection. Neighbor node selection scheme plays an important role in construction of overlays network. Our aim is to remove the nodes in overlay network which are reducing the performance in terms of service availability of the network. In order to improve the performance we remove such nodes which reduces the efficiency of the system. Therefore we modify the node selection approach along with the node removal method. We also recover the cluster with new super nodes to keep the robustness of the overlay network.