Resource Availability Optimization for a Point-to-Point Connection on a Telecommunication Network

In this paper, a Telecommunication Provider serving clients with different priorities is studied. Some of the services are telephony, internet service, pay TV, etc. The provider uses a network topology where each node is connected to exactly two other nodes. The interconnection of the nodes is made by optical fiber with specific bandwidth interfaces for point to point connection. When the connection between two nodes is lost or there is an increased demand in a network segment, bandwidth and availability issues occur for the point to point connections of the network as data rerouting takes place. Initially all classes have access to all system resources but an increased demand can lead to resource decrementation or even to resource exhaustion. A Markov model-based resource availability optimization problem is formulated in order to increase availability of high priority classes by reserving system's resources and simultaneously serving as many lower priority class requests as possible. There is a trade-off between guaranteeing availability for high priority classes and the cost of denying requests of lower priority. Empirical data are used and through a numerical example, the solution of the optimization problem reflects the best resource reservation policy.