Machuca, Carmen Mas; Eberspächer, Jörg (Technische Universität München, Deutschland)
Jäger, Monika; Gladisch, Andreas (T-Systems, Berlin, Deutschland)
Network operators are continuously performing techno-economic studies trying to minimize the cost of their networks. Nowadays, most of these studies investigate the Capital Expenditures (CapEx) of a certain technology, which are related to the network dimensioning and design. However, it is well known that not only CapEx but also the Operational Expenditures (OpEx) highly impact the overall cost of the network. This fact has triggered the interest of network operators in modeling the operational costs of their networks, aiming at minimizing the overall cost of their networks (i.e. the sum of CapEx and OpEx). Incumbent network operators are running different service specific platforms, based on different technologies, and the maturity of each technology is evolving quite fast. Currently, operators are trying to converge their platforms in order to reduce cost. The suitability of technology and network architecture underlying the operator’s production platform(s) depends on the service portfolio the operator intends to offer. From the service point of view, the operator has to decide on the services that are offered, which technology will support which services (there may be several technologies able to support the same type of service), and which platform is suitable for offering a service. One important aspect in the decision process is related to existing services in platforms that should not be upgraded any more but substituted by a new platform. The existing services can either remain in the “old” platform, requiring this platform in operation until the end of life of all services, or they are migrated to a “new” platform. For minimizing the cost of old platform operation and service migration, it needs to be investigated a) how much it would cost to migrate these services from an old platform to a new platform; and b) how much should be invested to migrate services within a given period of time. In this study we present how the operational steps of service migration can be modeled, and the relation between cost, required number of employees and the total number of services that are effectively migrated (i.e. that are not released during the migration time).