Fair and QoS-Oriented Spectrum Splitting in Macrocell-Femtocell Networks

• Main Authors: Erturk, M C, Aki, H, Güvenc, İsmail, Arslan, Hüseyin
• Format: Conference Proceeding
• Language: English
• Subjects: Artificial neural networks; Indexes; Interference; Macrocell networks; Measurement; Quality of service; Resource management
• ISSN: 1930-529X; 2576-764X
• DOI: 10.1109/GLOCOM.2010.5683397

In split spectrum macrocell-femtocell networks, certain portion of the available spectrum resources are dedicated to each tier in order to avoid interference problems among the tiers. In such a setting, effective partitioning of the spectrum resources carries critical importance for maximizing the total capacity and satisfying quality of service (QoS) requirements of the users within each tier. This paper proposes spectrum splitting methods for a macrocell-femtocell network considering factors other than the capacity maximization. First, a capacity maximizing spectrum splitting solution is derived based on certain simplifying assumptions. Then, a modified QoS-oriented fairness metric is introduced which captures important characteristics of tiered network structures that are not captured by the Jain's fairness index. Finally, a spectrum splitting strategy that simultaneously considers capacity maximization, fairness constraints, and QoS constraints is proposed. Example numerical results show that while only considering total capacity maximization may yield unfair spectrum splitting, a more balanced spectrum allocation can be achieved by also considering fairness and QoS constraints.