Using Importance Sampling and Control Variate in Estimating Network Relibility
碩士 === 國立臺灣科技大學 === 資訊管理系 === 93 === Performace of network with random components is usually difficult to evaluate. Simulation of the stochastic system often becomes the only feasible method to assess the network and quality of the estimator becomes the major concern. Importance sampling is a varia...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2005
|
Online Access: | http://ndltd.ncl.edu.tw/handle/47429015496681500114 |
Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 93 === Performace of network with random components is usually difficult to evaluate. Simulation of the stochastic system often becomes the only feasible method to assess the network and quality of the estimator becomes the major concern.
Importance sampling is a variance reduction technique which reduces the variation of the estimator
without increasing the sampling effort. In this study, importance sampling technique is used to alter the sampling scheme of a highly reliable network system to make the rare event of system failure accur more frequently. Then, the state of a key component is used as a control variate to further improve
the importance sampling estimator.
Simulation experiments are performed to evaluate the performace of the importance estimator and the combined estimator.
|
---|