|Authors||P. R. Grønsund|
|Title||Cognitive Radio From a Mobile Operator's Perspective: System Performance and Business Case Evaluations|
|Afilliation||Networks, Communication Systems|
|Publication Type||PhD Thesis|
|Year of Publication||2013|
|Publisher||University of Oslo|
Cognitive Radio (CR) is a promising technology to solve the spectrum scarcity problem and to increase spectrum efficiency. By using spectrum sensing and databases, one is able to obtain information about and utilize white space spectrum. Spectrum costs, which tend to be high for mobile operators, can then be reduced significantly. For mobile operators, CR technology brings threats and opportunities. Two evident threats are the risk of increased interference and competition when secondary users access white spaces in the operator's own spectrum. Considering the high level of activities on CR in research, standardization and regulation in addition to commercial activities, CR might eventually emerge in the telecommunications market. Therefore, instead of focusing on the threats only, mobile operators should focus on understanding how they can benefit from using CR. One opportunity is to use CR to get access to more spectrum to cope with the increasing wireless data demand and spectrum scarcity. Another opportunity is to use CR to access spectrum in, and to enter, new markets where there are no available spectrum licenses. The objective of this thesis is to study and understand how a mobile operator can benefit from using CR as a potential sustaining or disruptive innovation to opportunistically access white spaces. We show, by example case studies, that there are potentials for a mobile operator to use CR to access white spaces and achieve well performing technical and economic viable solutions. In particular, we study three important areas for CR with focus on the mobile operator's perspective. First, we characterize spectrum usage and analyze potential capacity for CR access in primary OFDMA networks. We show that there is a potential for CR systems to utilize white spaces. Furthermore, we propose that cooperation with the primary operator is important to maximize spectrum utilization. Second, we study the concept of a sensor network aided CR system. We propose three business cases and evaluate the economic viability. The most promising business case for an operator is that of a joint venture that gets the rights to use the ``unused'' spectrum resources of spectrum owners. We find that high reuse of existing base station sites by the CR system is a business critical parameter. Furthermore, it is found challenging to achieve high reuse of existing base station sites when evaluating technical performance using a simulation model. Hence, this points in the direction of shorter range and less expensive access points such as femtocells. However, we show that full reuse of base station sites can be achieved by relaxing interference requirements for the CR. Then, we propose a promising business case that uses cognitive femtocells aided by a sensor network to offload the LTE network. Third, we use simulations to evaluate performance of the first CR standard IEEE 802.22. We find that the activity of wireless microphones as the primary users should be quite high to reduce throughput and delay. Interference to the wireless microphone is found to be low in general and to occur only for short periods when using novel sensing strategies. Furthermore, we show that the guaranteed bit rate QoS service for VoIP can be prioritized. Though, the spectrum sensing strategy is important to satisfy strict QoS requirements for throughput and delay. Finally, we explore spectrum selection functions, which are used as basis for channel selection. We show that selection functions that utilize long-term spectrum usage statistics based on historic, accumulated sensing results can enhance over-all performance.