With the development of mobile communication technology, the number of mobile terminal users keeps increasing. Human users mostly carry mobile devices. Users can use mobile devices for data transmission and data sharing across spaces. However, mobile users are highly mobile and have a certain degree of selfishness, and their encounters are random. Traditional probabilistic routing methods are difficult to adapt in the application of social network mobile opportunity transmission since they ignore the nodes' sociality and cooperation. To reduce the impact of uncertain factors on data transmission, we propose a predict the probability method of encounter and forwarding cooperation with node (PNECP), the method predicting mobility probabilities based on the social relationship and forwarding collaboration relationship between mobile users. This method uses mixed-relations matrix decomposition to predict the probability of the user's mobility encounter, taking into account the user's encounter and the ability to collaboratively forward information. Designing the way of information collection and transmission, when the data is sparse, can also give prediction results. Simulation results show that our proposed method has a better performance on the delivery rate and average delay performance compared with other probabilistic transmission methods.

随着移动通信技术的发展，移动终端用户数量不断增加。人类用户大多携带移动设备。用户可以使用移动设备跨空间进行数据传输和数据共享。但是，移动用户的移动性很高，具有一定程度的自私，他们的遭遇是随机的。传统的概率路由方法由于忽略了节点的社交性和合作性，因此难以适应社交网络移动机会传输的应用。为了减少不确定因素对数据传输的影响，我们提出了一种预测与节点相遇和转发合作的概率方法（PNECP），该方法基于移动用户之间的社交关系和转发协作关系来预测移动概率。该方法使用混合关系矩阵分解来预测用户遇到移动性的可能性，同时考虑到用户的遇到以及协同转发信息的能力。当数据稀疏时，设计信息收集和传输方式也可以给出预测结果。仿真结果表明，与其他概率传输方法相比，本文提出的方法在传输速率和平均延迟性能上具有更好的性能。