Research On The Pilot Decontamination Algorithm For Massive MIMO Systems

The massive MIMO(Multiple-Input Multiple-Output,MIMO)technology is equipped with a large antenna array(usually dozens or hundreds of antennas)at the base station.By increasing the spatial freedom degree,the multiplex gain and diversity gain can be obtained for the communication system.It can not only greatly improve the capacity of the system,but also reduce the transmission power of the base station,then be one of the key technologies of 5G.However,the coherent time of the system channel is limited,and the pilot training sequences of all users can not be orthogonal to each other.Thus,the bottleneck problem of the large-scale MIMO system,pilot contamination,is caused.The research shows that the performance of channel estimation in large-scale MIMO system is mainly limited by pilot contamination.Therefore,pilot contamination elimination technology has become one of the core technologies in the field of 5G.At present,there are many research achievements in this field.The design of pilot allocation scheme is the main research focus.The design principle is to optimize the pilot allocation scheme and take account of the spectrum efficiency.In this paper,the problem of how to use shorter pilot sequences in the pilot contamination elimination scheme is studied in order to reduce the contamination and improve the transmission efficiency.The main work of this paper is as follows:1.In order to solve the problem of long pilot sequence and low spectral efficiency in the traditional pilot distribution scheme,this paper proposes a scheme based on division sectorization configuration pilot.The idea is to divide the sectorization in the community.The users in the same sectorization use orthogonal pilot sequence and the user pilot sequence between different sectorizations,so that the pilot contamination mainly exists between different sectorizations.When the users angle-of-arrival is not miscible,we can get rid of the pilot contamination by Bayes estimation and the difference of spatial information.The simulation resultsshow that the proposed pilot configuration scheme eliminates pilot contamination and effectively reduces pilot sequence length compared with traditional schemes,and achieves a significant increase in efficiency and speed with low pilot overhead.However,this scheme requires two known channel statistics of all users.2.In order to eliminate pilot contamination in multi sectorization large scale MIMO system with low pilot cost and avoid dependence on all two order statistical information,a pilot contamination elimination algorithm based on uplink and downlink training is proposed.Through the design of small pilot multiplexing,inter sectorization orthogonal pilot,the number and length of pilot area of training sequence is proportional to,and respectively design of channel two stages of training and training of the uplink downlink channel estimation process,and the whole process remain unchanged,the complete elimination of pilot contamination.The algorithm not only does not require two order channel statistics,but also trains pilot sequence length to reduce to the number of sectorizations,which greatly improves the transmission performance of large-scale MIMO system.