一种基于预测的自适应机会网络消息转发策略

Prediction-based strategy of message adaptive forwarding for opportunistic networks

针对机会移动网络数据转发机制进行了深入研究。机会移动传感网络以"存储—携带—转发"的方式实现节点间通信,需要减少网络中消息总的副本数量,提高消息传递成功率。基于简单实用的原则,提出结合消息副本率和与sink节点相遇概率的自适应消息转发策略(adaptive forwarding strategy,AFS)。当两个传感器相遇时,首先更新各自的消息副本率和与sink节点相遇概率参数;随后以这两个参数为依据,确定消息是否转发及转发方向。该方法可以有效提高消息传递成功率,降低网络中总体消息副本率,避免消息拥堵。实验结果表明,自适应消息转发策略是有效的,能够获得良好的网络性能。

In opportunistic mobile sensor networks, the communications can be accomplished by the way of ＂storing-taking-forwarding＂. To achieve the goal, it is necessary to reduce the total amount of message copies and promote the success ratio of message delivering. Based on the principle of simplicity and practicability,the paper put forward a strategy of adaptive forwarding strategy（AFS） for message delivery by the combination with message copy ratio and the probability of meeting sink node. Firstly, both sensor updated their message copy ratio and meeting probability respectively. Subsequently, whether need to forward the message and the forwarding direction could be decided according to these two parameters. This strategy could promote the success ratio of message delivery and reduce the total amount of message copies, efficiently. At the same time, it could also avoid conjesting of copied messages. The simulation results indicate that this strategy is effective and help to achieve a higher performance of the network.