Abstract: The energy of nodes of wireless sensor networks is limited, using traditional link selection algorithm（empirical-algorithm） needs to send many data packets as testing samples, but it is not allowed in wireless sensor networks. The paper designs two link-selection algorithms based on Bayesian estimation and one link-selection algorithm based on hierarchical Bayesian estimation, marked as BLSP-B1, BLSP-B2 and BLSP-HE. Simulation result shows that BLSP-B1, BLSP-B2 and BLSP-HE have a 10%~20% higher success rate than empirical-algorithm in selecting the high quality link in the case of the small sample. Among them, BLSP-HE has better and most robust performance.

Key words: wireless sensor networks, Bayesian estimation, hierarchical Bayesian estimation, link selection

摘要： 无线传感器网络（WSN）节点能量有限，采用传统的链路选择的方法（经验法）进行链路选择，需要发送大量的数据包作为测试样本，这在WSN中是不合适的。设计了两种基于Bayes估计与一种基于多层Bayes估计的WSN链路选择算法，分别记为BLSP-B1、BLSP-B2、BLSP-HE。仿真实验发现，在小样本的条件下，BLSP-B1、BLSP-B2、BLSP-HE选择高质量的链路的概率比经验法要高出10%～20%，其中BLSP-HE算法最稳健，性能较好。

关键词: 无线传感器网络, Bayes估计, 多层Bayes估计, 链路选择