Analyzing telecommunication network availability performance using the downtime probability distribution

To develop a better understanding of the expected performance of a population of systems, and to establish a basis for developing downtime objectives that better define and control expected service performance, the authors develop a parallel system equipment availability model. The model is used to obtain the steady-state downtime distribution over a specified time period using common channel signaling (CCS) network access links as the reference architecture. It is shown how the downtime distribution can be qualified by modeling a CCS network access segment as a system with two independent, restorable units operating in a load-sharing, parallel mode. It is shown that a steady-state system downtime distribution over a specified time period exists and has a mean equal to the system average downtime objective. Using these results, it is also shown that systems with a small downtime objective achieve steady state rapidly. The proposed availability model shows that with the current average downtime objective of two minutes per year, many access segments (98.35%) will experience zero downtime during one year of steady-state operation, while a few (1%) may have downtime exceeding one hour or more