AG Kommunikationstheorie

Thema:

Planning LTE Networks With Data Uncertainty Using Robust Optimization

Abstract:

Current methods for the planning of wireless networks require a static model of the problem. However, uncertainty of data arises frequently in wireless networks, e.g. fluctuating bandwidth requirements. Robust optimization is a new approach to deal with uncertainty in the framework of optimization models. The approach presented in this talk limits the number of uncertain entries in the parameter matrix by a robustness parameter and optimizes against the worst instances of these uncertain entries. Three linked problem formulations for throughput maximization and coverage maximization of LTE networks are presented in this talk. The related optimization models are formulated as mixed-integer linear programs and integer linear programs, respectively. For each linear program, a robust formulation is given. Additionally, first numerical results are presented to compare each linear program to its corresponding robust formulation.



zurück zur Terminübersicht