Neighbor selection in Peer-to-Peer Computing using Multi-Layer Perceptron

In this paper, we propose the Multi-Layer Perceptron (MLP) technique for Neighbor Selection in Peer-to-Peer (P2P) Computing to reduce the communication overhead. The selection of Neighbor is one of the challenging areas in P2P Computing. Root Mean Square Error and Testing time are two Parameters considered for neighbor selection in P2P network. The objective of the proposed technique is to minimize the testing time with respect to Mean Square Error. The performance of the proposed technique is evaluated through simulation by considering a network size of 1000 nodes. Neural Network provides good accuracy, when the learning rate is greater than 0.5. Experiment has been conducted through simulation using transaction dataset and simulation result outperform for neighbor node selection in P2P network.