Delay Adaptive Radio Resource Management Technology In Heterogeneous Network

With the rapid development of communication technology, the future wireless communication system performance requirements are constantly improved. To enhance the network capacity and the performance of spectrum and energy efficiency as well as meet the high-quality user experience, the technology of relays, picocells, femtocells has been proposed in heterogeneous wireless networks. Heterogeneous wireless networks provides users with flexible services, but also brings some problems, such as the difference in quality of service requirements," the diversity of users’behavior and the complexity of the interference types, etc. As a result, in heterogeneous wireless networks system, how to manage the network resources reasonably, to provide users with a variety of multimedia services and guarantee the quality of service as well as improve the spectrum utilization, has become a research hotspot. In general, the goal of traditional radio resource management scheme is to optimize the spectral efficiency and the energy efficiency while ignoring the service delay performance. Actually, besides the high spectrum efficiency and energy efficient which should be considered, the delay of packet data service should be highlighted in radio resource management in order to improve the network quality of service and fairness. This thesis focuses on the delay adaptive radio resource management in heterogeneous wireless networks. The main contribution is as follows:The first contribution is a radio resource management algorithm based on MDP. The service queue dynamics can be modeled as a markov decision process problem, in which the delay requirements of different users reflected in the queue weighting factors. By solving minimization problem of the total weighted delay, a distributed user scheduling and power control algorithm is proposed. Based on the dynamics of channel state information (CSI) and queue state information (QSI), we focus on finding an adaptive user scheduling and power control policy to achieve a desirable network delay level. In order to reduce the computational complexity and facilitate the distributive implementation, a distributive stochastic learning algorithm which only requires local CSI and local QSI to determine the optimal user scheduling and power control policy is derived. The simulation results show that this algorithm can achieve desired time delay performance as well as the low complexity.In the second point, a delay performance analysis scheme based on Lyapunov approach has been proposed. Firstly, the queue stability can be studied by using Lyapunov approach. Then, the relationship between the optimization goal and the delay performance can be derived based on the queuing theory. Take the relationship between energy efficiency and delay performance as an example, the tradeoff between energy efficiency and delay performance has been presented. By choosing appropriate tradeoff parameter, the target delay performance can be obtained, which has important significance for guiding the radio resource management scheme.