Multi-region capacity planning model with contracts of varying duration under uncertainty : a satellite capacity acquisition case study

Abstract

This paper highlights the issues associated with and presents a modeling framework for long-term capacity planning problems constrained in a similar fashion to satellite capacity acquisition. Although ambiguities exist, the satellite capacity planning problem also contains many clearly defined constraints. In the immediate time horizon, universal satellite capacity is fixed, as satellite launches are rare, and each incremental satellite launch only marginally increases capacity. In the world of Commercial Satellite Communications (COMSATCOM), satellites have fixed coverage areas as they are typically in geo-synchronous orbit. Other military applications such as observation and satellite phone require, low-earth orbit and medium-earth orbit satellites. The contracts used to acquire satellite capacity from the commercial sector also clearly define constraints, such as start and end date, bandwidth, price per unit bandwidth and allocated satellite system, which also implies frequency and available capacity. The paper first reviews the issues involved with planning capacity acquisition, and then presents a modeling framework for developing strategies to meet criteria and minimize cost. A simulation environment is developed for testing contracts and satellites of different parameters against varying demand distributions. A dynamic program is used to solve the integer program associated with selecting satellite capacity levels to cover demand.