中文摘要：

针对现有的P2P信任模型中对节点行为改变的动态适应能力不足以及欠缺激励机制的特点，提出了一种引入激励机制的动态信任模型（DJTrust）。该模型采用信任值动态衰减策略以提高其动态适应能力，同时通过推荐偏移量来计算推荐信誉值，增加了推荐信誉的可信度；最后引入虚拟信誉货币以鼓励节点提供服务和推荐信息，提高了节点模型参与的主动性。实验结果表明，DITrust较已有的模型能够抵御类型更广泛的恶意节点攻击，且具有更小的网络开销。

英文摘要：

Most existing trust models do not provide adequate support to cope with quick changes in peers＇ behavior and run short of incentive mechanism. Based on that, this paper presented a dynamic trust model with incentive mechanism （ DIT- rust）. Introduced dynamic attenuation mechanism of trust value to enhance the model＇ s dynamic adapting ability. At the same time, computed recommendation reputation through the deviation of recommendation to increase the reliability. Finally, imported virtual reputation currency in order to enhance the peers＇ positivity of participation. The simulation results show that DITrust can resist more types of deliberate attack than other models, and needs less overloads.