Delay-aware Resource Allocation Optimization In Cloud Redioaccess Networks

Cloud radio access networks(C-RANs)is an important component of the fifth generation wireless communication systems,which decrease the operating expenditure and capital expenditure for mobile operators and significantly improve the network performance.In the cloud radio access networks,traditional base stations are decoupled into base band units and distributed remote radio heads.The baseband unit pool aggregates all computational resource and provides large-scale processing and management functions for the transmitted signals.The remote radio heads with antennas work cooperatively and improve the spectrum efficiency of the system.However,since the fronthaul is constrained in C-RANs,it becomes a significant challenge and research hotspot of designing efficient radio resource allocation scheme to adapt to the dynamic load and satisfy the quality of service(QoS)in C-RANs.The traditional radio resource allocation optimization mostly focused on optimizing the energy efficiency and spectrum efficiency based on power control,subchannel allocation and cell association.Most of the existing works neglect the QoS requirement on the delay of different service for individual users.To ensure the fairness between users and improve the efficiency of resource utilization,it is significant to consider the delay characteristic of different service and the influence brought by the delay of packet service.Hence,to improve the quality of service of the network,the radio resource allocation optimization should take both the improvement of energy efficiency and spectrum efficiency and delay requirement for individual users into consideration.This paper investigates the cross-layer radio resource allocation optimization in cloud radio access networks,which considers the delay requirement of non-real time packet services based on the framework of the traditional radio resource optimization.The queues for packet services are constructed and the optimization problems under certain constraints are formulated.Corresponding delay-aware cross-layer resource allocation optimization schemes are proposed,which can achieve a tradeoff between network performance and the delay of service queues.Hence,the proposed schemes can satisfy the requirement for QoS and guarantee the performance of the network and improve the resource utilization efficiency at the same time.The main contributions of this paper are:First,the cross-layer resource allocation problem based on the beamforming vector design in C-RANs is investigated in this paper.An energy-efficient optimization objective function with instantaneous and average power constraint,inter-tier interference constraints and system queue stability constraint is presented in this paper for cloud radio access networks.The optimization problem is a non-convex problem and can be reformulated as a stochastic optimization problem by introducing the Lyapunov optimization approach.Under the Lyapunov optimization framework,the problem can be regarded as an EE maximization problem with power constraints and inter-tier interference constraint,which can be solved by generalized weighted minimum square error receiver and an iterative approach and based on which an optimal network-wide cooperative beamformer design algorithm is proposed.The proposed delay-aware resource allocation solution achieves a tradeoff between average EE performance and average delay.With the increase of the control parameter V,the EE performance improves at the cost of sacrificing the delay performance.Then,the delay-aware dynamic multi-dimensional resource optimization problem in heterogeneous C-RANs is researched in this paper.A throughput utility optimization problem in heterogeneous C-RANs is formulated.To make the average utility of user throughput arbitrarily close to the optimum and maintain system queue stability under instantaneous and average power constraints,a dynamic multi-dimensional resource allocation scheme which jointly considers cell association,power allocation and subchannel allocation is proposed in this paper.The proposal maximizes the utility of throughput under the queue stability constraint and achieves a tradeoff between average utility performance and queuing delay.To meet the great challenge of designing effective delay-aware resource allocation optimization solution,this paper formulated two optimization problems which based on the optimization objective function of maximizing EE and throughput utility,respectively.The corresponding delay-aware resource allocation solutions are proposed,both of which optimizes the performance of the network under the queue stability constraint by dynamically allocating the resource and finally achieve a tradeoff between system performance and queuing delay.In the end of this paper,the summarization of the paper is made and a future research direction is pointed out.