Study On Fronthaul Compression In Cloud Radio Access Network

With the rapid development of mobile communication technology,wireless service and application scenarios are towards the diversification and intelligence evolution.High bandwidth,low latency and massive connections are the urgent need of the future network.The traditional access network architecture has been unable to meet the development of modern communication network.The emergence of a new type of access network architecture is necessary and inevitable.The C-RAN(Centralized,Cooperative,Cloud Radio Access Network,centralized,cooperative,cloud and wireless access network)came into being under this background.The nature of it is that the base station processing resources of all or part are of the focus,which can be beneficial to unified management and dynamic allocation of the baseband resource pool,thereby reducing the cost of construction and equipment,reducing the overall energy consumption,improving the coordination ability through centralized deployment.When using virtualization technology,it can promote the sharing of resources,as well as making the scheduling more flexible,as a result of which significantly improve the spectrum utilization.The new architecture has been the subject of widespread attention and study.C-RAN architecture is mainly composed of RRU(Remote Radio Unit)and a large number of BBU(Building Baseband Unit)which are parts of the virtual baseband pool,each performing its duties.The capacity of the fronthaul between the RRU and the BBU pools is limited.In order to improve the utilization of the fronthaul,an effective compression method is needed.Therefore,based on this reality,this paper studies and analyzes the problem of data compression and transmission of fronthaul link.On the one hand,it is necessary to study the practical compression scheme for fronthaul data.On the other hand,based on the network information theory,we can optimize the compression noise so as to maximize users' sum rate.First of all,this paper expounds the basic situation of C-RAN architecture,including the base station structure,bearing mode,application and challenge of C-RAN.Then,we introduce some theoretical knowledge of network information theory.We make the theoretical content of this part as a breakthrough point,and apply it to the analysis of fronthaul compression model under C-RAN architecture.This part lays a theoretical foundation for the later optimization process.Secondly,this paper analyzes the LTE compression schemes.There are mainly two kinds,one of which is IQ compression including frequency compression,resampling and quantization.EVM performance analysis and comparison of the 20 MLTE signal in different quantization are given.The other one is the discussion of compressed sensing based on LTE signal.Two algorithms are used to reconstruct and analyze the success probability of sparse signals.Then,this paper focuses on the analysis and research of the limited capacity of the uplink fronthaul,the framework for C-RAN uplink fronthaul compression model is introduced,respectively for point-to-point compression and distributed compression.When the receiver uses successive interference cancellation decoding scheme based on two users under the multiple access channel,the achievable rate region is analyzed.The study found that users transmit power and compression noise has a great relationship with the rate.Therefore the paper comes to the problem of the maximization of users' sumrate.Given the joint optimization of transmit power and compression noise power,user rate performance is improved.The paper proposed an optimization method which is called variable replacement algorithm,obtaining the same rate performance as the classic MM algorithm.Finally,this paper discusses the problem of downlink fronthaul transmission with limited capacity.The downlink fronthaul compression models can be divided into multivariate compression and independent compression.The joint optimization problem of compressed noise and precoding matrix can be solved by MM algorithm.According to the different precoding methods,different forms of optimization are carried out,and the rate performance of different compression forms under linear precoding and dirty paper coding is simulated and compared.