Performance And Achievable Rate Analysis Of Communication Systems With Low-Resolution ADCs

The coming mobile communication systems will occupy higher-frequency bands and larger bandwidths to support higher communication rates.Therefore,the system with high-resolution analog-to-digital converters(ADCs)suffers from high power consumption and hardware cost.Using the low-resolution ADCs to sample and quantize the received signal is a direct and effective way to improve the energy efficiency.Unfortunately,the low-resolution ADCs incur severe nonlinear quantization distortion,thus the analyses for the systems with infinite-precision or high-resolution ADCs are no longer applicable.This thesis considers analyzing the performance and achievable rate of communication systems with low-resolution ADCs to provide guidance for practical system design.Firstly,this thesis analyzes the performance of orthogonal frequency division multiplexing(OFDM)receiver with low-resolution ADC over doubly selective channels for joint channel-and-data estimation(detection)in the Bayesian framework.An OFDM receiver model based on the basis expansion model is established,and the prior probabilities of the doubly selective channels are modeled according to their characteristics.The fixed point equations that can be solved through simple iterations are presented to describe the performance of the joint estimator by using the replica method from statistical physics.The influence of ADC parameters on system performance is analyzed through numerical simulation.The simulation shows that when the QPSK modulation is used,the performance of receiver with 4-bit quantization is close to that with infinite-precision quantization.Secondly,this thesis analyzes the ergodic sum rate of uplink multiuser multiple input multiple output(MIMO)systems with low-resolution ADCs under the Rayleigh fading channels.A general expression of the ergodic sum rate is derived by using the replica method and the central limit theorem,and the specific expressions of the ergodic sum rate are also derived when the prior probabilities of data are Gaussian distribution and M-QAM modulation.The accuracy and widespread application of analysis results are verified,and the impact of ADC parameters on the sum rate is also analyzed through simulation.The simulation resluts also show that the system with a finite set of data candidates can use low-precision ADCs to reduce power consumption and cost without achievable rate loss(or with slight achievable rate loss).Finally,this thesis analyzes the ergodic sum rate of uplink millimeter wave multiuser MIMO systems with low-resolution ADCs under the finite-dimensional channels.To capture both the sparsity and array structure,the finite-dimensional channel model is introduced.Based on the Bussgang linear quantization model,the expression of the ergodic sum rate is derived when data follows a Gaussian distribution by using the replica method.The algorithm for optimizing input covariance matrix is proposed to maximize the ergodic sum rate.For point-to-point systems,the expression for the ergodic channel capacity is given when the normalized array steering vector is the column of the unitary matrix.The accuracy and widespread application of analysis results and the fast convergence of the proposed algorithm are verified through simulation.The impact of ADC parameters on the ergodic sum rate is also analyzed through simulation.The simulation resluts also show that the achievable rate loss caused by the low-resolution ADCs can be reduced or eliminated by optimizing the input covariance matrix.