Grey Model-Enhanced Risk Assessment and Prediction for P2P Nodes

Establishing high performance cooperation in the peer-to-peer (P2P) network architecture is a fundamental and challenging research area currently, due to peer anonymity, peer independence, high dynamics of peer behaviors and network conditions, and the absence of an effective security mechanism. Most assessment methods based on trust or reputation have some remarkable innate drawbacks, for some presumed restrictions to the data samples, and such methods are inherently disabled for identifying the malicious recommendations, which weakens the final results´ credibility and persuasiveness. To answer the issues of peer risk assessment in p2p networks, we propose a novel risk assessment-and-prediction method(RAPM) based on grey theory, to analyze peer´s credibility. Grey modeling, grey decision making, and Taste Concourse Method are presented to assess and predict peer´s risks. In our scheme, four major peer´s attributes are selected for demonstrating GM-based risk assessment. The initial results show that this scheme is an efficient method for risk assessment and prediction of incomplete information nodes in P2P networks, and there are fewer requirements for the quantity of sample data than traditional methods for risk assessment.