Research On Interference Cancellation For Non-orthogonal Multiple Access

Non-orthogonal multiple access(NOMA)technology,a new multiple access technology which can superimpose multiple users' signals and transmit them in the same time-frequency domain,is considered to be the most potential multiple access technology in 5G communication system because of the advantages of high spectral efficiency,low delay and high connection density.However,serious multiple access interference occurs at the receiver since the multiple users' signals in NOMA system are superimposed in non-orthogonal form in the same subband.Therefore,eliminating the signals of other interfering users accurately,quickly and efficiently at the receiver is the key of NOMA system.And the focus of the thesis is interference cancellation technology of NOMA system.Firstly,this thesis introduces the downlink and uplink NOMA system theory in detail on the basis of introducing the development status of NOMA technology.Interference cancellation technology,an important part of NOMA,is simply classified.In addition,multiple access signature technology for separating multiple users' signals and the constellation design technology for reducing bit error rate(BER)are also introduced in this thesis.Thereafter,for the downlink NOMA system,a NOMA scheme based on user sequence phase rotation is proposed in this thesis.Considering the correlation between users' signals at the transmitter and the limitation of signal processing capability at the receiver,the signal processing at the transmitter and the receiver is modeled and the optimization of BER performance in downlink NOMA system is realized by solving the model.It is found from the simulation results that compared with the NOMA scheme based on constellation rotation,the bit error ratio gain of the near user has improved by 30.3%,and the bit error ratio gain of the far user has improved by 39.4% on the proposed NOMA scheme in this thesis.Finally,for the uplink NOMA system,an adaptive weighted factors and multiple access signature aided parallel interference cancellation(PIC)algorithm is proposed in this thesis.Considering the uplink NOMA system scenario composed of a single base station and multiple users,the problem of optimizing weighted factors of the uplink NOMA system is modeled and the biased estimation of the PIC algorithm is adjusted adaptively at the receiver by solving the optimization problem.It is found from the simulation results that the sum data rate of the uplink NOMA system and the BER performance at receiver are improved effectively by the proposed PIC algorithm.