Performance Analysis And Optimal Design Of Mixed-ADC Massive MIMO Systems

With the continuous increase of mobile communication equipment,mobile data traffic is growing rapidly,which places higher requirements on mobile communication systems,and the fifth generation of mobile communication systems(5G)came into being.Massive multiple input multiple output(MIMO)technology can fully exploit the spatial resources and achieve substantial gain in spectral efficiency,energy efficiency,delay and transmit reliability.Therefore,massive MIMO technology has become one of the key technologies in 5G.However,massive MIMO systems need to deploy a large number of antennas,which will cause problems such as huge data processing,high hardware cost and huge power consumption.One of the promising means to solve the above problems is to reduce the resolution of the analog-to-digital converter(ADC)in receivers,which is able to reduce the hardware cost,power consumption and computational complexity.However,low-resolution ADC massive MIMO systems will encounter quantization noise that cannot be ignored and more antennas are required to ensure performance.As a compromise between the traditional full-resolution ADC massive MIMO systems and the low-resolution ADC massive MIMO systems,the mixed-ADC massive MIMO systems can exploit advantages of both the above two structures.In this dissertation,massive MIMO systems with mixed-ADC structures are studied.Performance analysis and optimization of mixed-ADC massive MIMO systems are carried out.On the one hand,the uplink achievable rate and energy efficiency of mixed-ADC centralized massive MIMO systems are analyzed.The additive quantization noise model(AQNM)is used to model the ADC quantization noise.Based on this model and assuming the perfect channel state information is known at the receivers,when the base station use maximum ratio combining(MRC),zero forcing(ZF)and minimum mean square error(MMSE)receivers,the asymptotic expressions for the achievable rate are derived respectively and accurate approximate expressions for the MMSE receivers are also provided,we obtain some relatively simple achievable rate expressions.By using the theory of majorization,it is found that for MRC and ZF receivers,the equal-resolution ADC structure can achieve the optimal performance,while the mixed-ADC structure is optimal for MMSE receivers.Based on this conclusion,an optimal quantization bit allocation algorithm is designed for MMSE receivers.Simulation results show that the obtained asymptotic expressions are very accurate.The mixed-ADC massive MIMO system can achieve better performance than the massive MIMO system with low-resolution ADCs,and the energy efficiency of the mixedADC massive MIMO system significantly outperforms that of the full-resolution massive MIMO system.On the other hand,the uplink achievable rate of mixed-ADC distributed massive MIMO systems is studied,where each remote antenna unit(RAU)is considered to use different resolution ADCs for quantization.Based on this system model,the asymptotic expressions of achievable rate and energy efficiency for MRC and ZF receivers are derived.In addition,the performance of MMSE receivers is evaluated by simulation.Simulation results show that the asymptotic expressions obtained are very accurate for MRC detection.The performance of the distributed massive MIMO system with mixed-ADC structure is not much different from the full-resolution ADC architecture.When the cell is relatively large,the achievable rate of the distributed massive MIMO system is better than the centralized massive MIMO system.For ZF receivers,the achievable rate of the full-resolution ADC structure is much better than mixed-ADC and lowresolution-ADC systems,but the energy efficiency of mixed-ADC and low-resolution ADC systems is far ahead.When MMSE receiver is used,the achievable rate of the mixed-ADC structure is better than that of low-resolution ADC structures under high signal to noise ratio(SNR)region.