中文摘要：

转提出了一种基于发概率的动态数据转发策略FPAD （forwarding probability-based adaptive cata delivery algorithm）．FPAD适用于由不同类型传感器节点构成的可监测不同对象的异构延迟容忍移动传感器网络HDTMSN（heterogeneous delay tolerant mobile sensor network）在这种网络中，各类节点拥有不同的通信能力、运动速度与消息存储能力，并且获取的数据消息具有不同的大小和不同的延迟容忍度．针对异构网络的特点，FPAD一方面根据节点能量消耗和消息传输延迟计算出节点的传输概率和转发概率，并以此进行数据消息的传输；另一方面，提出根据消息当前的延迟容忍度作为消息丢弃依据的消息队列管理机制．仿真实验结果表明，与现有的几种数据传输算法相比，FPAD的数据传输成功率更高、传输延迟更小，而且网络寿命相对较长．

英文摘要：

This paper proposes a data gathering algorithm—FPAD （forwarding probability-based adaptive data delivery algorithm）. FPAD applies to HDTMSN （heterogeneous delay tolerant mobile sensor network）, which is composed by different nodes monitoring different objects. In such networks, nodes of different kinds have different communication capabilities, motor speed, and information storage capacity, and the data obtained have different sizes and delay tolerances. In terms of the characteristics of HDTMSN, FPAD, onthe one hand, calculates nodes transmitting and forwarding probability, and then forwards messages according to nodes energy consumption and message transmission delay. On the other hand, it proposes message queue management mechanism based on the current delay tolerance of the message. Simulation results show that compared with the existing algorithm of data transmission, FPAD has a higher delivery ratio and smaller delivery delay, and it has a relatively longer network lifetime.