Abstract: Analyzing nodes behaviors in a PeertoPeer (P2P) network can effectively provide different services and encourage nodes to share resources, thereby improving the system performance. Using states to represent nodes behaviors, P2P node state model was proposed to describe the process of states transition. The model was combined with nodes behaviors assessment mechanism to specify conditions of node state. Experimental results show that the model, which has good scalability and availability, can achieve service differentiation and promote resources sharing by nodes.

Key words: node state, incentive mechanism

摘要： 对等网络中，研究节点的行为差异有利于实现区分服务，激励节点共享资源，提高系统的性能。用节点状态表示节点行为表现，提出P2P节点状态模型，描述状态的动态变化规律，并将模型与节点行为评价机制结合，确定节点状态的约束条件，提高模型的可用性。实验证明模型有效地支持区分服务，能促进节点的资源共享，具有良好的扩展性和实用性。

关键词: 对等网, 节点状态, 区分服务, 激励机制, Peer-to-Peer, differentiated service