Performance Optimization Theory And Technology Research On Small Cell Networks

The cellular system is evolving to the fifth generation (5G). Compared to the fourth generation (4G), 5G takes into consideration not only the gain of peak data rate,spectrum efficiency, mobility and latency, but also the improvement of new key performance metrics such as the user experienced data rate, area traffic capacity,network energy efficiency and connection density. To implement the business use of 5G in 2020 which is the target proposed by International Telecommunication Union,the evolution of 5G should begin with the discussion of spectrum, requirement and key technologies, rethink about the evolution experience of current cellular systems and key technologies, explore new evolution ideas and select the potential evolution paths.According to the forecast by Cisco, the global wireless traffic of 2020 will be 8 times as much as that of 2015. To deal with the soar of traffic requirement, the densification of network is inevitable. Small cell is one of the main promising solutions to the dense network deployment, because small cell is cheap, energy efficient, plug and play. So the performance optimization theory and technology research on small cell networks are powerful engines for the evolution of 5G. This thesis aims at improving the system throughput, system energy efficiency and system fairness by proposing solutions to the interference management issues, energy efficiency optimization problem and load balancing in small cell networks. The proposed solutions of interference management,system energy efficiency and load balancing are excellent methods to relieve the challenges of current cellular network and are of significant importance to the evolution of 5G. The contributions of this thesis include the following three aspects.First, two cooperative power control schemes for small cell networks are proposed to improve the system throughput. The first scheme is named as resource block exclusion based power control (RBEBPC) and is proposed by sharing the interference correlated information. RBEBPC consists of two steps that are iteratively conducted:(1) based on current power allocation results, partial system resource blocks are excluded by playing the formulated cooperative coalition formation games; (2) the transmission power of each small cell is determined by solving a modified throughput maximization problem after the resource block exclusion. As the generated interference is constrained in the second step, part of the small cells transmit without full power.Thereby, the overall system interference keeps non-increasing after the adoption of RBEBPC. The second scheme is called Interference Allocation based Power Control(IAPC) and uses cooperation among small cells. In IAPC, two steps are repeated to manage interference: (1) a central entity (e.g., a macro cell) plays a canonical coalition game to constrain the interference allowed by each small cell, and (2) based on this interference allocation, each small cell solves a throughput maximization problem to control sub-channel transmit powers. Compared with Iterative Water Filling (IWF),IAPC improves system throughput by limiting interference cooperatively. This improvement is also a result of allocating less power to the sub-channels.Second, to optimize the energy efficiency of small cell networks with constrained backhaul links, power control based schemes and a small cell switch on/off scheme are proposed. The research on power control based schemes include two scenarios. The first scenario is a simpler one where the power consumption of backhaul links has no fixed charge component and each small cell serves only one subscriber. Then a more complex scenario where the power consumption of backhaul link has a fixed charge and each small cell serves more than one subscribers are studied. Two steps are repeated in the power control based schemes: (1) based on the backhaul link data rates, the small cells optimize the transmission power to reduce the interference; the system throughput is improved with lower transmission power, because the waste of transmission power in the form of interference is relieved; (2) by redistributing the backhaul link data rate,the backhaul link data rates match the front link data rate. The system energy efficiency in the power control based schemes is guaranteed to be convergent in a non-decreasing manner. In the research on small cell switch on/off scheme, the minimum number of small cells that could satisfy the subscribers’ data rate requirements are first switched on. Then the promising small cell and subscriber couple that could improve the system energy efficiency will be switched on. The system energy efficiency keeps increasing as long as a new small cell and subscriber couple is found. The small cell switch on/off scheme also indicates that open the minimum small cells to guarantee the data rate requirements of subscribers cannot always achieve the optimal system energy efficiency when the effect of backhaul links is considered.Third, to guarantee the fairness of small cell networks, a load definition is proposed and used to construct the load balancing problem. The load of small cell is defined as the maximum ratio of attached subscriber’s required data rate to the corresponding attached subscriber’s achieved data rate. The proposed load definition jointly considers the differences of small cells’ service ability and the differences of subscribers’ requirements. Based on the definition, the load balancing problem of small cell networks is formulated as a lexicographically vector minimization problem. The properties of the constructed problem are analyzed and a lower complex scheme is proposed to approach the load balancing bound.