The cross-entropy method for network reliability estimation

被引:74
作者
Hui, KP [1 ]
Bean, N
Kraetzl, M
Kroese, DP
机构
[1] Def Sci & Technol Org, IN Div, Edinburgh 5111, Australia
[2] Univ Adelaide, Dept Appl Math, Adelaide, SA 5005, Australia
[3] Def Sci & Technol Org, ISR Div, Edinburgh 5111, Australia
[4] Univ Queensland, Dept Math, Brisbane, Qld 4072, Australia
关键词
network reliability; cross-entropy; rare events; importance sampling; permutation Monte Carlo; merge process;
D O I
10.1007/s10479-005-5726-x
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
Consider a network of unreliable links, modelling for example a communication network. Estimating the reliability of the network-expressed as the probability that certain nodes in the network are connected-is a computationally difficult task. In this paper we study how the Cross-Entropy method can be used to obtain more efficient network reliability estimation procedures. Three techniques of estimation are considered: Crude Monte Carlo and the more sophisticated Permutation Monte Carlo and Merge Process. We show that the Cross-Entropy method yields a speed-up over all three techniques.
引用
收藏
页码:101 / 118
页数:18
相关论文
共 35 条
[11]  
de Boer PT, 2002, EUR T TELECOMMUN, V13, P303
[12]  
DEBOER PT, 2000, THESIS U TWENTE
[13]  
DEMELLO TH, 2002, UNPUB RARE EVENT PRO
[14]  
DUBIN U, 2002, THESIS TECHNION HAIF
[15]  
EASTON MC, 1980, IEEE T RELIAB, V29, P27, DOI 10.1109/TR.1980.5220696
[16]   ESTIMATION OF NETWORK RELIABILITY USING GRAPH EVOLUTION MODELS [J].
ELPERIN, T ;
GERTSBAKH, I ;
LOMONSOV, M .
IEEE TRANSACTIONS ON RELIABILITY, 1991, 40 (05) :572-581
[17]   ASSOCIATION OF RANDOM VARIABLES WITH APPLICATIONS [J].
ESARY, JD ;
PROSCHAN, F ;
WALKUP, DW .
ANNALS OF MATHEMATICAL STATISTICS, 1967, 38 (05) :1466-&
[18]   A MONTE-CARLO SAMPLING PLAN FOR ESTIMATING NETWORK RELIABILITY [J].
FISHMAN, GS .
OPERATIONS RESEARCH, 1986, 34 (04) :581-594
[19]  
Gertsbakh I., 2000, Reliability theory with applications to preventive maintenance
[20]  
HELVIK BE, 2001, 3 INT WORKSH MOB AG