Interference Suppression Research Of Heterogeneous Cellular Networks Based On SVR Beamforming

In recent years,so many kinds of wireless terminals such as smart mobile phones,smart wearable device and tablet computers etc.have increasingly entered people's lives,and the demand for wireless network communication services has shown explosive growth.In order to ensure that users get better signal quality and higher system capacity,Heterogeneous Cellular Networks(Het Net)are widely used and deployed.Although the deployment of a Het Net can improve the overall performance and capacity of the network,the overlap of the coverage of the MBS and the SBS makes the interference received by the user more complicated.Therefore,the research on noise interference suppression in the Het Net is particularly important.This thesis first summarizes the research status at home and abroad,and proposes an interference suppression scheme based on beamforming at the user end.The two-layer Het Net was modeled and analyzed,and the model topology and channel fading model were determined respectively.The SBSs and user positions were generated based on random geometry,and the interference in the heterogeneous cellular network was further analyzed.Secondly,research on beamforming algorithms based on machine learning.The proposed beamforming algorithm bases on the Support Vector Regression(SVR)method.We would replace the structural risk item of SVR with the array output energy,while replacing the empirical risk item of SVR with the array output error.The proposed algorithm does not need to obtain the angle of arrival of the desired signal in advance,thereby avoiding the impact of the direction of arrival mismatch on the performance of the beamformer.At the same time,the proposed beamforming algorithm based on SVR effectively inherits the small sample advantage of SVR.Finally,the existing Het Net interference suppression is mostly solved at the base station,so for the problem of complex interference in Het Net,an interference suppression scheme based on SVR beamforming at the user end is proposed and studied.In this solution,the base station uses an omnidirectional antenna,and the user end uses a directional antenna and deploys an SVR-based beamforming algorithm.In order to verify the feasibility of the method,a two-layer Het Net was simulated in different scenarios,and the locations of base stations and users were described through Poisson Point Processes(PPP)in random geometry.The locations of base stations and users were described according to the principle of random geometry correlation,while deducing the user-side access probability and spectrum efficiency and other indicators.Subsequently,simulations were performed on the deployment strength of different SBSs,the number of user-side antennas,and the number of training snapshots.The results show that the Het Net based on SVR beamforming can effectively suppress interference,improve user throughput and improve spectrum efficiency.