Epidemic算法在某些场景中具有很高的传输成功率、很小的传输延迟,但算法的适应性较差,在另一些场景中算法性能会显著下降.理论和实验分析表明,挤出效应是导致Epidemic算法性能下降的主要原因.分析了具有免疫机制Epidemic算法的性能,指出了该机制的缺陷,提出了退避机制:当某一节点缓冲区饱和时,不再接收与之相遇节点的数据包.在ONE仿真平台上实现了具有退避机制的Ep-idemic算法,实验结果表明,在挤出效应显著的场景下,退避机制能有效地抑制挤出效应,改进后算法的传输成功率有大幅度的提高,路由开销也有一定程度的下降.%In some scenarios, Epidemic algorithm has high delivery ratio, small delivery delay, but poor adaptabilityd. Moreover, the performance of the algorithm will significantly degrade in other scenarios. On the basis of analysis of the factors affecting the algorithm performance, Crowding-Out effect is considered as the main reason leading to negative performance. In this paper, the performance of Epidemic algorithm with immune mechanism is analyzed and some defects of the immune mechanism are indicated. Therefore, an improved algorithm is formulated with a kind of Back-off mechanism, so that the node will no longer receive packets from meeting nodes when its buffer is close to saturation. The promising results on the ONE simulation platform show that the proposed algorithm can effectively suppress Crowding-Out effect and greatly improve the delivery ratio and reduce the routing overhead to some extend under various scenarios.
展开▼
