Non-idealities Analysis And Compensation In Terahertz Communication Devices

With the rise of various emerging high-speed wireless transmission services such as cloud service,8K livestream and 3-D hologram and the increasing of the number of Internet users,it puts forward higher requirements for the bandwidth,transmission rate,delay and the number of access equipment of mobile communication system.The fifth generation mobile communication system can not fully meet people's application needs,terahertz communication is regarded as one of the potential key alternative technologies of the sixth generation mobile communication.Terahertz communication has gradually become the research focus of industry and academia because of its rich spectrum resources and transmission rate reaches several gigabits or even terabit.However,due to the increase of frequency band,device technology level,system structure and cost,the terahertz communication system performance inevitably suffers from the radio-frequency(RF)non-idealities.This thesis investigates the common RF non-idealities,and analyzes their impact on the communication system performance.Based on the single carrier frequency domain equalization(SC-FDE)system,the compensation scheme for the two RF non-idealities are proposed.The research work of this paper is arranged as follows:The first part introduces the communication link and transmission parameters based on the IEEE 802.11 ad standard.Phase noise,I/Q imbalance and power amplified nonlinearity are briefly summarized.Moreover,the phase noise model based on the 802.11 ad standard and terahertz phase noise model are investigated.The simulation demonstrates that there is correlation between terahertz phase noise sampling points.The second part focues on the frequency-independent I/Q imbalance.Firstly,it analyzes the system model impacted by the I/Q imbalance at the transmitter and receiver,and manifests that the impact of I/Q imbalance on the communication system is reflected in the mirror interference from the signal itself.Besides,an I/Q imbalance parameter estimation method based on minimizing channel change energy is studied.In order to improve performance in NLo S channel because of too many approximate processing,an I/Q imbalance compensation scheme used to classify signal impacted hy by I/Q imbalance is proposed.The system uses the two channels of I/Q data after frequency domain equalization at the receiver as the feature training neural network to complete the correct classification and judgment.The third part analyzes the SC-FDE system affected by phase noise,and shows that the impact of phase noise on the SC-FDE system is reflected in common phase error(CPE)and intersymbol interference(ISI).Then,the traditional CPE non-iterative compensation scheme and the iterative least mean square(LMS)compensation scheme based on decision-directed are studied.Secondly,to solve the drawbacks of the traditional phase noise compensation scheme,an adaptive memory recursive least squares(RLS)scheme based on decision-directed is proposed,and the effectiveness and advantages of phase noise compensation algorithm is proved.Especially in the non line of sight(NLo S)channel,its convergence speed and bit error rate perform better than the traditional scheme.