DHT网络规模估计算法的定量分析与设计  

Quantitative Analysis and Design of an Estimating Algorithm on DHT Network Size

作　　者：肖波[1] 聂晓文[2] 侯孟书[2] 

机构地区：[1]西南交通大学信息科学与技术学院,成都610031 [2]电子科技大学计算机学院,成都610054

出　　处：《电子科技大学学报》2011年第2期261-266,共6页Journal of University of Electronic Science and Technology of China

基　　金：国家自然科学基金(61073181)

摘　　要：为准确而高效地估计分布式哈希表(DHT)的网络规模,提出一种基于平均间距的估计算法。该算法利用节点之间的间距服从指数分布这一特性,通过测量平均间距长度估计网络规模。从最简单情况入手,逐渐增加估计的间距个数,定量分析了网络规模估计值的概率分布和置信区间。根据分析,如果采用多条间距估计网络规模,估计值将服从参数为(k,k)的伽马分布的尾分布。最后,仿真试验验证了算法的有效性与分析的正确性。To accurately and efficiently estimate the network size of distributed Hash table （DHT）, an algorithm based on the average interval is proposed. The algorithm makes use of the property that the intervals between nodes obey the exponential distribution to estimate the network size. From the simplest condition, the number of estimated intervals is gradually increased and the probabilistic distribution and the confidential interval of estimating values are quantitatively analyzed. According to the analysis, if we adopt multiple intervals to estimate the networks size, the estimated value of the network size will obey the tail Gamma distribution with the parameter （k, k）. Finally, simulations are carried out to verify the validity of the algorithms and the correctness of the analysis.

关 键 词：置信区间 分布式哈希表 概率分布 网络规模 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]