Research On Beam Resource Management And Control In 5G Cloud Radio Access Network

With the development and application promotion of 5G technology,the number of various intelligent terminal devices has shown explosive growth,making people put forward higher requirements for communication bandwidth and communication delay.Millimeter wave beamforming is one of the important technologies to solve the key performance indicators of next-generation mobile networks.C-RAN and its derivative network architecture can effectively improve the performance of millimeter wave beamforming.Artificial intelligence technology has also been widely used in the field of communication with its development in theory and practical applications.This thesis takes the 5G C-RAN as the research background,and takes the green communication and intelligent communication for 5G,B5G and even 6G communication as the research goal,and studies the technology of beam resource management and control in millimeter wave communication.The research work and contributions of the thesis are as follows:(1)This thesis first extends the centralized control beamforming system based on the optical real time delay module to the multi-cell multi-user communication scenario to adapt it to the C-RAN architecture of RRH dense deployment.The optical true time delay module deployed in the BBU pool provide the service beam mode required by users for multi-cell base station.At the same time,compared with the traditional scheme of deploying one optical true time delay module for every base station,this thesis proposes a beam resource allocation scheme which can save about half of the physical resources of deployment,that is,?n/2(?)optical delay modules are deployed for n base stations in a multi-cell system,which can effectively reduce the complexity and deployment cost of the physical resource deployment of the system.The non-optimal probability proposed shows that the probability of conflict between the service beam patterns required by users is very small.(2)On the basis of the multi-cell centralized control beamforming system based on the optical true time delay module,the thesis designs a convolutional neural network for the optical true time delay module which generates the service beam pattern in the BBU pool to makes the system adapt to the intelligent mobile communication network,and proposes a beam resource allocation scheme based on convolutional neural network.In this scheme,the convolutional neural network introduces synthetic noise when generating training data.The proposed traversal codebook coding design algorithm can obtain the output label of the training data with the optimization goal of maximizing the system spectrum efficiency.Therefore,in the prediction stage,a channel matrix with a certain degree of damage can be accepted as input,and the precoders F_RF and W_RF with good performance can be output,which has better anti-interference ability.The research results show that,in addition to having better performance than the traditional scheme,the proposed resource allocation scheme benefits from a large amount of training data of the convolutional neural network,which can output results faster and reduce communication delay.