Integrity according to Bayes

Applying Bayesian analysis techniques, the paper shows that the positioning performance can be described as a weighed sum of error distributions, in which the weights depend both on the measurement residuals, the statistical properties of the measurements and the a priori probabilities of failure. This description is not only elegant in the insight it provides on the `distribution´ of accuracy and integrity over the different measurements, but also offers operational benefits over existing methods. Unlike with currently available methods, an exact probability of hazardously misleading information P(HMI) is readily derived. Furthermore, the processing scheme is simpler than for traditional fault detection based architectures