摘要: This paper presents methods for overbounding themissed-detection probability for monitors with nominallychi-square noise distributions. Such distributions occurcommonly for integrity monitors in safety-criticalapplications, such a... 展开 This paper presents methods for overbounding themissed-detection probability for monitors with nominallychi-square noise distributions. Such distributions occurcommonly for integrity monitors in safety-criticalapplications, such as in Space-Based and Ground-BasedAugmentation Systems for GNSS. Robust boundingmethods are needed when noise parameters, such as themeasurement covariance matrix, are not precisely known.Simple inflation of the covariance model for the inputdata does not, typically, ensure conservative bounding forall bias directions. To address this problem, conservativemodels of monitor-statistic noise are introduced, whichensure bounding for all covariance matrices in a specifiedrange, between lower and upper bounds. 收起