论文部分内容阅读
近年提出的云无线接入网,通过集中化改善了小区为中心无线接入网在成本和灵活性方面的瓶颈.但是,云无线接入网需要通过前传网大范围汇聚带宽极高的前传采样,造成了过高的建设与运营成本.本文针对云无线接入网过度依赖前传通信资源的问题,依据控制与数据分离的思想设计了一种软件定义超蜂窝网络新型架构,并提出了在该架构下通过通信与计算协同的资源部署大幅降低前传汇聚带宽的具体方案.首先基于排队论给出了虚拟基站池中计算资源统计复用增益与池规模的定量关系,并依据统计复用增益边际效应迅速递减的性质得到了部署中等规模基站池更经济的指导方针.然后基于图聚类框架针对基带处理功能分割部署问题给出了一种遗传算法,该算法给出的分割方案可以根据设计偏好在前传带宽成本与计算代价间进行灵活折衷.
In recent years, the proposed cloud radio access network improves the cell-centric radio access network’s bottleneck in terms of cost and flexibility by concentrating the network.However, the cloud radio access network needs to aggregate a large bandwidth pre-sampling through the pre-transmission network , Resulting in excessive construction and operation costs.In this paper, aiming at the over-reliance on pre-communication resources of cloud wireless access network, a new architecture of software-defined ultra-cellular network is designed according to the idea of control and data separation, Architecture to reduce the bandwidth of premultiplication.Firstly, based on the queuing theory, the quantitative relationship between the statistical multiplexing gain of virtual resource pool and the size of pool is given, and according to the statistical multiplexing gain margin The rapid diminishing effect of the effect is more economical for the deployment of medium-sized base station pooling.And then, a genetic algorithm is proposed based on graph clustering framework for the segmentation and deployment of baseband processing functions. The proposed segmentation scheme can be based on design preferences Flexibility between pre-delivery bandwidth costs and compute costs.