Service Reliability Analysis Using Competing Risk Models

The survival and failure analysis provide a lot of information regarding reliability and maintainability of systems. Arriving at suitable distribution and parameterization of data helps predicting the behavior of systems. Mobile communication has become one of the most valued and indispensable service sector due to its commercial, financial and entertainment applications, apart from usual verbal use. There have been dedicated efforts to improve its versatility, transportability and equipment quality. But lack of infrastructural facilities and the neglect of 'industrial best practices' has made this sector less dependable. The end to end availability of the network depends up on the reliability of each element in the series. Here I am analyzing six months live data of a network with about 2360 BTS serving 3.5 million customers using competing risk models. Both physical as well as the capacity related issues are focused together to get better idea of availability. The failure restoration time distribution patterns are arrived from the data.