Research on Key Technologies of Resource Management and Control in Super Base Station Architecture
Yuanyuan Wang · The Sydney eScholarship Repository (The University of Sydney) · 2016
Key technologies of resource management and control are studied in this thesis for centralized cellular access networks. Firstly, the resource virtualization management framework and key mechanisms are studied, including priority based double resource scheduling mechanism, computing resource allocation mechanism, and the realization model of resource management. Further, a seamless service migration mechanism based on data context is proposed. Secondly, an energy-efficient base station on/off control and user association algorithm is proposed. The system optimization problem is decomposed into two sub-problems. The first sub-problem is maximizing the system throughput under a certain base station on/off pattern. Second one is to maximize the energy efficiency of the system by base station on/off control. We propose a marginal effect based greedy user association algorithm (MGUA) and a simulated annealing based base station switching algorithm (SABO) separately. Simulation results show that: based on the proposed MGUA and SABO algorithms, compared to the traditional user access and base station switch mechanism, the system obtains 35% energy efficiency gain. Thirdly, a cooperative interference mitigation algorithm is proposed to solve the interference management problem in densely deployed centralized small cell networks. We model the power control problem of small cell networks as a cooperative game model, and prove the existence of pure strategy Nash Equilibrium. We propose a centralized random adaptive power control (RAPC) algorithm and a stochastic learning based discrete power control (SLPC) algorithm. Simulation results show that both of the proposed RAPC and SLPC algorithms have good convergence and optimization performance. In the case of densely network deployment scenario, using the proposed algorithm, the number of small base station transmissions satisfying their quality requirements, two times that of existing control algorithm in the literature.