中文摘要：

针对移动P2P网络大规模分布式和高度动态性的特点，提出一种基于动态反馈机制的信任模型。该模型引入了距离因子和推荐因子两个参数来控制推荐信任链的规模，提高推荐信任度的准确性。模型将反馈机制和惩罚机制集成到节点间的近期信任度和长期信任度的更新计算之中，提高了模型随时间和历史变化的动态适应性。模拟实验表明，该模型具有很好的动态适应能力，能够有效隔离恶意节点。

英文摘要：

Since the mobile P2P network is characterized by large-scale distribution and high dynamics, a trust model based on dynamic feedback mechanism was proposed. Two new parameters, distance factor and recommend factor, were introduced to control the size of chain of recommended trust, so as to improve the veracity of recommend trust. In addition, the feedback mechanism and punishment mechanism were integrated into the calculation of the recent trust and long-term trust, so the trust model has better adaptability to the dynamics of trust. Simulation experiments show that the model has very good dynamic adaptability and is able to effectively isolate the malicious node.