Bayes Weighted Availability for a Digital Radio Transmission System

Gamma priors, which are rich in shapes and conjugate to the exponential distributions, are assigned in this study to the parameters of exponentially distributed component up and down times. A closed-form availability estimator is used by applying Bayes theorem. Unlike classical estimation techniques, the Bayes approach applies when information on performance is limited or nonexistent. Parameters of the prior distributions for the Bayes approach represent similar systems in the telecommunication network for which historical data exist. The Bayes availability estimator is applied to a digital ratio transmission system. The digital radio example includes random processes for both equipment failure and signal propagation anomalies. In contrast to traditional availability definitions, a weighted availability estimator is employed to account for situations where the system is partially available. The k-out-of-n:G system is shown to be a special case of weighted availability.