|
内容記述 |
Multi-antenna transmission and reception technique, also known as MIMO,has evolved as one of the key enabling techniques to meet the ever-increasingdemand for high-speed wireless data access in current and emerging wirelesscellular networks. This dissertation addresses some critical issues on the re-source allocation for MIMO-enhanced cellular networks with centralized ordistributed antenna architecture.<br/>With the dramatic growth of mobile services provided by cellular net-works, the _rst problem that we should confront is to support and di_erentiatediverse services, particularly the quality of services (QoS) guarantee for real-time services. However, traditional algorithms for resource allocation fail toprovide a better solution that dynamically guarantees the QoS requirementswhile obtaining the throughput e_ciency, since they have neglected compet-ing and sharing characteristics between services from a systems perspective.In contrast, I consider this problem based on a cooperative game model, whichgives great insights into the nature of competing and cooperative relations.Consequently, I successfully formulated this problem on resource allocationas a cooperative game and obtained the notion of QoS guaranteed fairnessbased on the well known Nash bargaining solution. The algorithm based onQoS guaranteed fairness, can satisfy the QoS requirements of all services andachieve the Pareto optimal system throughput, which is validated throughsimulations and discussed at the end of Chapter 3. Moreover, this work alsoprovides a theoretical framework that paves the way to solving resource allo-cation problems in other similar scenarios. <br/> At the same time, the huge amount of tra_c have highly saturated thebandwidth available by current cellular networks, which pushes us to utilizebandwidth more e_ciently than ever so that universal frequency reuse is usu-ally considered by future cellular networks. However, this raised the secondproblem, severe inter-cell interference(ICI), which has become the bottleneckof further enhancement of spectral e_ciency. The spatial multiplexing trans-missions in MIMO-enhanced cellular networks, whose main advantage is adramatic improvement in spectral e_ciency, lose much of their e_ectivenessdue to this high levels of interference. Fortunately, the advances in MIMOtechnique such as cooperative transmission, especially that between base sta-tions (BS) within a cellular context, have emerged as one of the most promis-ing techniques to mitigate ICI and thus improves total system throughput.I proposed an algorithm in Chapter 4 for allocating wireless resources coop-eratively, which is aimed at mitigating ICI and e_ciently utilizing wirelessresources. Based on game theoretic analysis, the proposed algorithm achievesPareto optimal e_ciency and considers proportional fairness. Due to theprohibitive complexity of computation, I also developed a heuristic algorithmand compared it with a benchmark that is regarded as a Nash equilibriumoutcome in a non-cooperative scenario. The simulation and analysis resultsare also given at the end of Chapter 4. <br/> I also investigated the distributed antenna scenarios in both Chapter 5 and6 that have a topology of distributed antennas for the BS at each cell. Theintuitive advantages of this architecture are better signal coverage and lowerpower consumption. However, We expect to further exploit other advantagesiisince resource allocation with distributed antennas is more exible in cooper-ation and optimization than that in traditional architectures. In Chapter 5, Iproposed two energy-e_cient resource allocation algorithms, based on beam-forming transmission and selection transmission, respectively. the simulationresults shows that both algorithms have a higher energy e_ciency that con-ventional algorithms, and the selection transmission outperforms the beam-forming transmission algorithm in terms of energy e_ciency and complexity.The ICI problem in distributed antenna architecture is also investigated inChaper 6. I proposed a cooperative beamforming algorithm that mitigatesICI and achieves a higher system capacity. A comparison and analysis ofperformance between a scenario with co-located antennas and that with dis-tributed antennas are given, which clearly demonstrate the advantages of thearchitecture with distributed antennas. <br/> In summary, the cooperative resource allocation problems is investigatedin the MIMO-enhanced cellular networks. The game theoretic frameworkis proposed to provide QoS guarantee for diverse services, mitigate interfer-ence, and conserve transmission energy. All these algorithms can achieve thePareto optimal in terms of system throughput. my investigations with botharchitectures of distributed antenna and traditional co-located antenna is alsodiscussed in this dissertation. |
要旨・審査要旨 (286.5 kB)
