MTBO - a metric for assessing production failure rates in telecommunications networks

The Mean Time Between Outages (MTBO) metric provides the frequency of failures in telecommunications network elements (hardware and software) attributed to equipment supplier. By contrast, the traditional Mean Time Between Failures (MTBF) metric only addresses total failures that lead to element replacement. MTBO allows us to capture short duration outages (of the order of 1 second) which can interrupt such services as streaming video, telepresence, and online gaming. In contrast, the traditional Availability metric is not sensitive to short duration outages in the presence of long duration failures. Application of MTBO is demonstrated for a wide range of network elements in wire-line and wireless networks including IP routers, Ethernet networks connecting a large number of cell routers to a redundant concentrator, mobility equipment with redundancy and power amplifiers in the Universal Mobile Telecommunications System (UMTS) base station (nodeB). Requirements to MTBO measurement tools are outlined. In case of hardware failures, the router MTBO is evaluated using absorbing Markov models.