Research On Resource Allocation In Backhaul-Aware Heterogeneous Cellular Networks

In recent years,with the vigorous development of the Fifth Generation(5G)mobile networks,wireless data traffic has grown exponentially.Driven by the new generation of devices(smartphones,laptops,Internet-connected devices,etc.)and bandwidth-demanding applications(UHD video,AR/VR games,etc.),capacity demand is growing much faster than spectrum efficiency,especially in hot spots.Dense heterogeneous cellular networks can increase network coverage,expand network capacity and provide users with higher data transmission rates by increasing the deployment of small base stations,to cope with the explosive growth of data traffic in the 5G era.The backhaul link between the macro station and core network can provide a stable data transmission guarantee in most cases.However,with the intensive deployment of a large number of small zones and the demand for higher data transmission rates by users,the backhaul link is overwhelmed and under great pressure.The available backhaul capacity becomes a valuable resource and one of the important factors limiting the network performance,restricting the upper limit of network throughput,and affecting the quality of service(QoS)of users.Due to the diversity and high dynamic of user traffic,the data traffic in the backhaul link changes in real-time.In a network where the available backhaul capacity changes dynamically,if the macro station can sense the available backhaul capacity in the current TTI,it can adjust the user association,resource allocation,and power allocation strategies to improve the network performance.Therefore,how to obtain the available capacity of the macrocell backhaul link is studied in this paper.When the number of users in a small cell increases or heavy traffic service is used,the wireless backhaul link of data transmission between a small base station and the macro station will also limit the backhaul capacity of a small cell.The dual constraints of the available backhaul capacity of the macro cell and the wireless backhaul capacity of a small cell restrict the improvement of the whole system throughput.How to optimize the performance of the heterogeneous cellular network and guarantee the QoS of users through appropriate user association strategy under the dual backhaul constraint is also a further study in this paper.We will show the main research contents of this paper below.(1)This paper studies the available bandwidth detection model and proposes a real-time detection method of available bandwidth in 5G scenarios,which can obtain the present available backhaul capacity in TTI.Firstly,the available bandwidth detection model based on probability is studied,and a rate adjustment algorithm is designed to improve the detection accuracy of link utilization.Secondly,the Kalman filter algorithm is used to estimate the parameters of the linear model.Then,to compare the performance of the proposed algorithm under various background traffic,three background traffic models are modeled:the CBR model,Pareto model,and Burst model,which respectively represent different types of network traffic in real networks.Finally,in the 5G scenario,the algorithm simulation is completed and compared with the classical available bandwidth detection methods Pathload and Forecaster,showing the superior performance of the proposed algorithm.(2)This paper studies the resource allocation scheme of a two-layer heterogeneous cellular network under the dual limitation of available backhaul capacity of the macro cell and wireless backhaul capacity of a small cell.Based on the dual constraints of the available backhaul capacity of the macro cell and the wireless backhaul capacity of the small cell,and considering the QoS requirements of users and the co-tier interference,this paper proposes an optimization problem to maximize the network throughput of the system.For the convenience of solving the problem,the original problem is decoupled from the user association subproblem and the joint resource allocation and power allocation subproblem.In the user association subproblem,the small cell backhaul constraint is taken into account.The optimal solution is obtained by the Lagrangian dual decomposition method.Finally,the effectiveness of our algorithm is proved by numerical results,and the network performance under the dual limitations of available backhaul capacity and small cell wireless backhaul capacity is demonstrated.