提出了一种软件系统的非线性有源自回归(Nonlinear AutoRegressive models with eXogenous Inputs,NARX)网络模型的老化检测方法。解决了目前软件老化方法未考虑多变量间关联性及历史数据的延迟影响的问题。该方法首先通过对实验采集的HelixServer-VOD服务器性能数据进行主成分分析,确定网络的输入维数,根据AIC准则确定最佳模型阶数,最终选取合理的网络模型结构;使用已知的未老化状态样本对NARX网络进行训练,建立系统的辨识模型;然后运用序贯概率比检验(Sequential Probability Ratio Test,SPRT)对NARX辨识模型的残差进行假设检验,判断系统的老化状态。实验分析表明,基于NARX网络模型的故障检测方法能够有效地应用于软件老化的检测。
Secure and private authentication protocol is important in Radio Frequency Identification (RFID) technology.To date,researchers have proposed many RFID authentication protocols.However,these protocols have many flaws due to lack of theoretical support in designing these protocols.In this work,first we present the security and privacy requirements in RFID authentication protocols.Then we examine related works and point out problems in designing RFID authentication protocols.To solve these problems,we propose and briefly prove three theorems.We also give necessary examples for better understanding these theorems with concrete protocols.At last,we give our suggestions on designing secure and private authentication protocols.The security and privacy requirements,theorems,and suggestions will facilitate better understanding and designing of RFID authentication protocols in the future.