Resource Allocation Methods With Delay Constraint In For Radio Access Networks

The fog radio access network(F-RAN)can not only effectively reduce the capacity burden of the fronthaul link,but also significantly improve the network spectrum efficiency and energy efficiency.However,how to guarantee the service demands of real-time services while achieving high spectral efficiency and high energy efficiency performance of F-RAN is still a very challenging issue.On the other hand,resource allocation methods play a key role in further enhancing the network performance and ensuring the service demands of different types of services.Therefore,it is very important to study resource allocation methods for F-RAN to guarantee the service demands of delay-sensitive services.In this thesis,taking time-varying channel conditions and random traffic arrivals into account,delay constraint is modeled as network stability constraint.Then we focus our minds on the resource allocation problems under delay constraint in F-RAN,aiming at maximizing the throughput and energy efficiency respectively.In order to solve these optimization problems,we propose the corresponding resource allocation algorithms.The main work and contributions of this thesis are listed as follows.In order to maximize the overall throughput of F-RAN under network stability constraint,based on Lyapunov optimization theory and depth-first branch-and-bound method,a joint mode selection and resource allocation algorithm is proposed.Firstly,Lyapunov drift plus penalty expression is applied to transform the stochastic dynamic optimization problem into a deterministic static optimization problem,which can be decomposed into three sub-problems:transmission mode selection,uplink receiver beamforming design,and power control.All these three sub-problems can be iteratively solved.The depth-first branch-and-bound method proposed in this thesis chooses the node with the smallest lower bound to branch,and then takes the variable with the smallest partial derivative as the branch variable,which helps to achieve fast search and reduce computation complexity.The simulation results verify that the proposed algorithm can improve overall throughput of F-RAN and reduce service delay.Moreover,it also provides a simple mothod to balance the throughput performance and delay performance by adjusting the control parameter.Aiming at maximizing the energy efficiency of F-RAN,by solving the fractional stochastic optimization problem with network stability and queue length probability constraints,a multi-slot energy-efficient resource allocation algorithm is proposed.Leveraging on the Dinkelbach method and Markov inequality,the original optimization problem is equivalently transformed into a problem that can be solved by an iterative method.In each iteration,the optimum beamforming vector is obtained by using the weighted minimum mean square error approach.Besides,we design a low-complexity solution to the joint resource block allocation and power control problem,a mixed-integer nonlinear programming problem.System-level simulation results demonstrate that the proposed algorithm can maximize network energy efficiency while ensuring the requirements of different types of services.This thesis proposes effective resource allocation algorithms to guarantee the requirements of real-time services in F-RAN,making the F-RAN architecture suitable for the application in current and future mobile communication networks.The research on delay-constrained resource allocation problems in F-RAN not only provides theoretical analysis and performance reference for supporting delay-sensitive services in F-RAN,but also makes great contributions to the commercial deployment of F-RAN.