Efficient Resource Allocation In Distributed Coordinated Networks

As a key approach of inter-cell interference mitigation and system energy saving, the Distributed Coordinated Network (DCN) is considered as the one of the main architectures of4G system. Based on the different scenarios in the DCN, this dissertation focus on the efficient resource allocation problems, aiming at improvement of both traditional capacity-oriented requirements, such as transmission rate, fairness utility and weighted sum rates, and energy-oriented requirements, such as energy efficiency. The main results are listed as follows:1) Power allocation improving algorithm efficiency with single cell coordinated diversity.The transmission rate maximizing problem is addressed in the downlink OFDM DCN. Based on the optimization theories, an optimal power allocation algorithm is proposed, which reduces the exponomial complexity of existing work to a polynomial complexiy with respect to the subchannel number. Furthermore, by power iteration between multiple APs, another algorithm with a nearly linear complexity is proposed, achieving a performance very close to the optimum, which makes the scheme more practical.2) Fairness utility improvement in multicell DCN with coordinated diversity.According to the definitions of coordinated cells sets in CoMP standardization, a two-phase resource allocation strategy is proposed based on the coordinated proportional fair (PF) scheduling and the binary power allocation. By combining both the coordinated scheduling and the joint transmission, the proposed scheme achieves improvement in both cell-edge and cell-average performance, comparing with the single cell PF.3) Multiple resource joint optimization in multicell system with coordinated multi-antenna beamforming.For the multicell DCN with zero-forcing beamforming, the jointly optimizing problem among user grouping, subchannel and power allocation. By the dual decomposition, an iterative algorithm is proposed with a linear complexity with respect to the subchannel number. Then, another algorithm is proposed with a linear complexity with respect to both the subchannel number and the cell number, by introducing the greedy users selection into the dual problem. The simulation results demonstrate that, thanks to the joint optimization, the prosed schemes achieve a significant performance improvement, at the cost of a low complexity.4) Energy efficiency analysis with single cell coordinated diversity.The energy efficiencies are analyzed in the DCN with single cell coordinated diversity. Based on the generalized convexity theory and parametric programming, the nonlinear fractional problems are solved optimizing the energy efficiencies for both the coordinated MRC diversity and coherent transmission diversity. The optimal energy efficiencies are derived, along with the corresponding power allocation algorithms. And it is proved that, when the power constraint on each AP is sufficient large, the orignal problems in the DCN could be reduced to an equivalent problem in a single AP transmission system.Based on the above4aspects, this dissertation addressed the highly efficient resource allocation problem in DCNs, and proposed low-complexity algorithms to instantenously improve the performance of both cell-edge and cell-average, and to jointly optimize multi-dimension resources. Furthermore, the energy efficiency of DCN is investigated theoretically, deriving the optimal bound. Right now, the DCN has been promised to be the one of the main architectures in4G standardization. And a part of above works have been submitted to the corresponding standard organizations, while some of them also been used to apply for the technical patents.