Dynamic Spectrum Management Based On Interference Mitigation In Macro/Femto Heterogeneous Network

With the arrival of the mobile Internet era, wireless data traffic is increasing rapidly. In the future, mobile telecommunications will be faced with the problem of insufficient spectrum resources. It is urgent to develop novel technologies to make more efficient use of spectrum bands, in order to alleviate the tense situation of spectrum usage. Current static spectrum allocation rules always leave much spectrum not fully utilized. Dynamic Spectrum Management (DSM) scheme allows multiple radio systems to reuse and share the unutilized spectrum holes. DSM scheme relies on Cognitive Radio (CR) technology which can autonomously learn environment, adaptively adjust parameters, and opportunistically use the spectrum holes in time or space dimension. Thus DSM scheme based on CR greatly improve the utilization of radio spectrum resource and eliminate the bottleneck of spectrum scarcity.Another technology which can effectively improve spectrum efficiency is Macro/Femto heterogeneous network. Femtocell improves the indoor signal coverage and system capacity by reducing the distance between the transmitter and receiver. Exploiting cognitive technology into femtocell can joint their advantages and better realize the dynamic spectrum management. Moreover, the combination of cognitive radio and femtocell can mutually compensate each other. Cognitive radio can assist femtocell for detecting the surrounding environment in heterogeneous network. Femtocell is the practical application scenario of cognitive radio under the present spectrum allocation policy. It also can help cognitive radio to solve access control, hardware support, network transformation expenses, resource allocation and other important issues. Concerned what have been discussed above, the subject of this thesis is the dynamic spectrum management based on interference mitigation in Macro/Femto heterogeneous network. It focuses on the dynamic resource allocation to achieve system utility optimization, users’ different QoS requirements and interference mitigation. The research work and innovation points of this thesis are included as follows.To research DSM strategy in Macro/Femto heterogeneous network scenario, the thesis provides technical feasibility study on cognitive-femtocell. Firstly, the opportunities and challenges of both CR and femtocell are summarized. Secondly, the perception of cognitive-femtocell and related features are elaborated. Also the technical feasibility and key problems are discussed. Finally, the research emphases and research methods are put forward.To solve the interference detection problem in dynamic spectrum management, the thesis establishes "Cognitive Interference Domain" Model which is suitable for heterogeneous network scenario. Based on the interference model, then a power control method is proposed. Firstly, cognitive-femtocell should autonomously detect the channels’ condition in order to avoid cross-tier interference to the licensed macro users. The proposed "cognitive interference domain" model can be used to help judge the channel condition and provide decision basis for dynamic spectrum management. Secondly, the author proposes a power control scheme based on interference suppression with the help of interference detection model. Finally, system level simulation results and analysis show that the proposed scheme reduces the cross-tier interference, improves the overall network performance and spectrum utilization efficiency. Furthermore, the proposed scheme has lower computational complexity.Considering the intra/cross tier co-channel interference and users’ different QoS requirements, a spectrum allocation scheme based on the QoS Graph Coloring (QGC) algorithm is proposed. The aim is to mitigate the two kinds of interferences, satisfy users’ different QoS requirements and maximize the spectrum efficiency. Furthermore, it achieves a tradeoff between fairness and spectrum efficiency by introducing a regulatory factor when setting the correspondence between colors and spectrum bands. The simulation results demonstrate that the proposed scheme outperformed the conventional scheme and parallel scheme by providing better spectrum efficiency and average throughput.The last part of this thesis reviews and summarizes the research work. Further research directions are discussed.