Abstract: This paper proposes an improved universal generating function(UGF)algorithm oriented sensing data fusion and uses it to evaluate the reliability of linear topology wireless sensor network. Firstly, a two-way continuous k out of n: F system model is established according to WSN topology and data transmission under PEGASIS protocol. Then, sensor node’s universal generating function and composition operator are redefined according to the characteristic of data transmission and data fusion in WSN. Finally, the two-way continuous k out of n: F model is divided into two unidirectional models, and the reliability of two-way continuous k out of n: F model can be calculated by composing the reliability of two unidirectional models. An example is presented to test the proposed improved algorithm. The results show that the improved algorithm can effectively solve the problem of wireless sensor network reliability evaluation.

Key words: universal generating function, system reliability, two-way continuous k out of n:F system, wireless sensor network

摘要： 提出了面向感知数据融合的通用发生函数(UGF)改进算法,并使用该算法对线性拓扑结构的无线传感网络(WSN)可靠性进行了评估。首先对PEGASIS协议下WSN的线性拓扑结构及数据传输过程进行抽象,建立了双向连续k/n:F系统模型。然后根据WSN感知数据传输及融合方式,在改进算法中重新定义了传感节点的UGF表达式和组合算子。最后对双向连续k/n:F模型进行单向化分解,根据得到的单向模型可靠性推导出双向模型的可靠性表达式。通过具体实例对提出的改进算法进行了验证,计算结果显示改进的算法可有效解决传感网络线性拓扑结构可靠性评估问题。

关键词: 通用发生函数, 系统可靠性, 双向连续k/n:F系统, 无线传感网络