Uplink Channel Estimation And Pilot Design In One-bit Massive MIMO Systems

With the rapid development of mobile communications,the fifth generation of mobile communication systems requires the ability to provide a higher data transfer rate.Due t o the increasing tension of radio resources,massive multiple input multiple output(MIMO)technology can make full use of space resources,which can greatly improve the energy efficiency and spectral efficiency,and has become one of the key technologies of 5G.At the same time,there are massive data processing in massive MIMO system,the total power consumption of the system is huge,and the hardware cost is expensive.An effective technique for solving the above problems is to use a low-precision analog-to-digital converter(ADC)in the base station receiver to reduce the cost and power consumption of massive MIMO system deployments.However,in order to make the system perform reliable and high-speed communication on the basis of low precision quantizati on,it is necessary to obtain more accurate channel state information(CSI).Since all users in the cell work at the same time and frequency,in order to avoid intra-cell interference,users in a single cell need to send mutually orthogonal pilot sequences.Based on the above requirements for CSI and pilot,the thesis focuses on uplink channel estimation and pilot design in one-bit massive MIMO systems.Firstly,it is introduced that single-user SIMO system model,the massive multi-user MIMO system model and the corresponding channel model.Furthermore,it is analyzed that uplink pilot transmission and data transmission in TDD communication mode.Secondly,it is introduced that the uplink channel estimation of single-user SIMO system.Firstly,the influence of different types of CSI on system channel capacity is expounded.Then,the pilot-based channel estimation algorithm is introduced,and then it is analyzed the data detection under linear reception.Finally,we derive the achievable rate.Then,based on uplink channel estimation of one-bit single-user SIMO system,the least squares(LS)channel estimation and minimum mean square error(MMSE)channel estimation in one-bit large-scale multi-user MIMO system are studied.However,since the above two algorithms are based on the amount of mutual information between the channel input and the quantized output,the influence of the quantization noise on the channel estimation performance and the reachability can not be intuitively visualized.Then,the linear least mean square(LMMSE)channel estimation is deduced by linear decomposition based on Bussgang's decomposition theory to represent the relationship between one-bit quantized input and output.Finally,the complexity,mean square error(MSE)of the three algorithms and the reachable rate of the system are analyzed.The simulation results show that the LS estimation performance is the worst,but the computational complexity is lower,and it is also more suitable for massive MIMO systems.Finally,it is proposed that an new orthogonal pilot generation scheme based on cyclic shifted Zadoff-Chu(ZC)sequence.Firstly,the characteristics of the commonly used reference signal sequence are analyzed,and the generation of orthogonal pilots in the existing literature is studied in detail,in which all the users in the control cell send their own pilots in different time slots frequency).Then,a new orthogonal pilot generation scheme is proposed in this paper,which is generated by ZC sequence in the time domain cyclic shift to generate different user's pilot(ZC pilot).And then the pilot length with the maximum reachable rate is optimized.The simulation results show that compared with the pilot generation scheme in the existing literature,the proposed scheme can redu ce the MSE of the channel estimation by about 12%.As the pilot length is optimized,the achievable rate achieves an increase of about 10 %.