Abstract: According to power-law distribution characteristic of scale-free complex networks, an algorithm based on stratified sampling（SSBA） is proposed, which can compute the reliability measures of a large scale-free complex network by the analysis of a sample networks drawn from it．The bootstrap based confidence intervals for these reliability measures are also given．Extensive experiments show that SSBA is efficient in estimating the reliability measures of large scale-free complex networks．

Key words: scale-free complex networks, reliability of complex networks, reliability measures, stratified sampling method, confidence intervals

摘要： 针对大型无标度复杂网络的幂律分布特性，提出了一种基于分层抽样技术的算法SSBA，通过分析样本网络推导出大型无标度复杂网络的可靠性度量参数，并给出这些参数的Bootstrap置信区间。大量的实验表明，SSBA算法能有效估算出大型无标度复杂网络的可靠性度量参数。

关键词: 无标度复杂网络, 复杂网络可靠性, 可靠性度量, 分层抽样方法, 置信区间