Research On Digital Domain Full-Duplex Self-Interference Cancellation Method

With the widespread deployment of 5G,spectrum resources become more scarce and more efficient spectrum utilization technology is needed.Full duplex communication can simultaneously transmit and receive signals at the same time and in the same frequency band.Compared with time division duplex and frequency division duplex,it can double the spectrum utilization.The obstacle of full duplex communication is the introduction of serious self-interference(SI)in the receiver.If the SI can not be cencelled,the performance of full duplex communication will be seriously affected.According to the current research,the full duplex SI cancellation method is completed in three steps: antenna domain cancellation,analog domain cancellation and digital domain cancellation.In this paper,the method of SI cancellation without additional analog domain cancellation is studied.The relevant theories are analyzed in detail and verified by simulation.The specific contents are as follows:Firstly,this paper proposes a hierarchical quantization digital cancellation method using two analog-to-digital convertors(ADC)with different precision,which are defined as ADC＿Pre and ADC＿S.By selecting the quantization precision of ADC＿Pre,the quantization signal contains only the principal value of the SI,while the desired signal(DS)and the residual SI are remained in the quantization error.Then the residual SI is cancelled in the digital domain,and the SI is completely cancelled while the DS is saved.In addition,the interference factors affecting the hierarchical quantization method are analyzed.Secondly,according to the working principle of the hierarchical quantization method,the constraint conditions of quantization precision of two ADC are analysed in this paper.And the quantization precision range of ADC＿Pre and ADC＿S is proposed.In addition,it is proved that there is at least one set of positive integers which make the two precision ranges intersect.Thirdly,this paper proposes an improved method to solve the problem of the algorithm performance deterioration when the channel is in a non-ideal state.By analyzing the symbol time offset(STO)of the receiver caused by the time-varying channel,the influence on the sampling time of non-uniform sampling is obtained.Then,two different estimation methods of STO are used to correct the sampling time,so as to solve the problem of SI cancellation under the condition of STO caused by the time-varying channel.Finally,the proposed hierarchical quantization SI cancellation method and the improved non-uniform sampling SI cancellation method are verified experimentally.The experimental results indicate that the hierarchical quantization SI cancellation method can achieve at least 120 d B cancellation ability under different conditions.The improved non-uniform sampling SI cancellation method has at least 28 d B improvement of the cancellation ability.