摘要： 标记概率计算作为概率包标记技术的关键内容，对算法的收敛性、最弱链、节点负担等方面具有重要影响。为此，分析现有算法的优缺点，结合无线传感器网络(WSN)的分簇结构，提出一种基于包标记的层次式混合概率包标记算法。扩大上下游节点的相对距离差，从而拉大节点标记概率之间的差距，增加上游节点标记的到达概率，在降低节点负担和算法复杂度的同时，提高算法收敛性。分析结果表明，该算法在收敛性、最弱链方面优于基本包标记法，在节点计算与存储负担方面优于自适应包标记法，可实现WSN资源约束条件下的整体优化。
关键词:
无线传感器网络,
溯源定位,
概率包标记,
收敛性,
最弱链,
节点负担
Abstract: As the sticking point of probabilistic packet marking techniques, marking probability has important influence over the convergence, weakest link, and node burden. Based on the analysis of the existing algorithm’s merit and demrit, this paper proposes a Layered Mixed Probabilistic Packet Marking(LMPPM) algorithm combined with the cluster structure of Wireless Sensor Network(WSN). The distance of nodes’ marking probability is enlarged and the reaching probability of nodes of advanced position is increased by amplifying the relative distance of nodes. Analysis result shows that LMPPM algorithm is better than Basic Probabilistic Packet Marking(BPPM) algorithm in convergence and weakest link respects, and better than Adapt Probabilistic Packet Marking(APPM) algorithm in node burden respect, which can realize whole optimization in WSN resource constraint conditions.
Key words:
Wireless Sensor Network(WSN),
traceback localization,
probabilistic packet marking,
convergence,
the weakest link,
node burden
中图分类号: