A generalized uplink cell-free massive multi-input multi-output (MIMO) system assuming access points (APs) are paired with mixed-resolution analog-to-digital converters (ADCs) and digital-to-analog converters (DACs) is investigated in this article. Specifically, the mixed-ADC/DAC architecture sacrifices part of performance gains and advocates a few low-precision ADCs/DACs to reduce the power consumption. Based on this antenna structure, we derive closed-form achievable rate expressions for maximal ratio combining (MRC) receiver filters by drawing support from the additive quantization noise model (AQNM) and analyze the impacts of the key design parameters, i.e., the number of total antenna arrays, the precision of coarse ADCs/DACs, the proportion of ideal ADCs/DACs in the mixed ADCs/DACs structure, and the power control strategy. In addition, the energy efficiency (EE) of the considered mixed-ADC/DAC architecture is also explored by recalling a generic power consumption model. Aside from this, to guarantee uniformly good service wherever users are geographically located, a max–min fairness power optimization algorithm is proposed, which can be determined by solving a geometric programming problem. Numerical simulations validate our theoretical analyses and examine the effectiveness of the proposed algorithm.

In this article, a generalized uplink cell-free massive multi-input multi-output (MIMO) system with nonideal radio frequency (RF) chains and low-resolution analog-to-digital converters (ADCs) is investigated. Specifically, the RF impairments and ADC imperfections not only hinder the accuracy of channel estimation but also deteriorate the strength of received signal-to-interference-plus-noise ratio. Achievable closed-form rate expressions are first derived by leveraging on the Gaussian RF impairment model and linear additive quantization noise model, after that, we explore the uplink achievable rate surrounding the key design parameters, i.e., the quality of RF chains, the precision of low-resolution ADCs, and the number of total antenna arrays. Considering channel estimation, we find employing poor RF chains and ADCs could fetch finite ceilings on estimation accuracy. Besides, the RF scaling law is also established, which suggests that the access points (APs) can tolerate lower receiver quality as the number of total antenna arrays increases, and the receiver quality can be cut down at most by ${1}/\sqrt{MN}$ , where $M$ and $N$ are the number of APs and the number of antennas of each AP, respectively. Moreover, we observe 5-bit ADCs are enough to substitute infinite-bit ADCs with scarce performance degradation. Finally, the compensation between RF impairments and ADC imperfections is presented. The resultant analysis provides insights for designing a more economical cell-free massive MIMO system while keeping the achievable rate at an acceptable level.

In this paper, we present the detailed rate analysis for cell-free massive multiple-input multiple-output (MIMO) systems over spatially correlated Rayleigh fading channels. Taking the realistic impairment effects of spatial channel correlation, pilot contamination, and channel estimation errors into account, the lower-bounds of the achievable rates for both the uplink and downlink are derived with the low-complexity linear processing such as matched filter and conjugate beamforming, which enable us to take cognizance of the impacts of transmitted power, and the number of access points (APs). Based on the derived rate results, the asymptotic performance analysis is then carried out. Besides, we propose the sophisticated max-min power allocation strategies taking the actual requirements into consideration to provide uniformly good service to all users. However, the objective functions of the two optimization problems are both non-concave. Fortunately, the former for uplink can be characterized as geometric programming (GP), whilst the latter for downlink merging the efficient tools of second-order-cone programming (SOCP). Lastly, the numerical results are shown to verify our analytical results and the effectiveness of the proposed max-min fairness algorithms.

In millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems, the hybrid beamforming architecture has been put forward to reduce the high hardware cost and power consumption, which are resulted from the tremendous requirements of dedicated radio frequency (RF) chains. In this paper, we propose several strategies to design analog and digital precoders for a point-to-point (P2P) hybrid MIMO system. Aiming at minimizing the Euclidean distance between the optimal digital precoder and hybrid precoder, we decouple this matrix factorization problem into a nonconvex quadratically constrained quadratic programming (QCQP) problem and an unit-modulus least-squares (ULS) problem, which can be solved by the presented three alternating optimization algorithms. Simulation and analysis results indicate that the proposed semidefinite relaxation based alternating optimization (SDR–AO) algorithm can approach near-optimal spectral efficiency performance compared with previous algorithms in the literature, but shows extremely high computational complexity. The alternating direction method of multipliers based alternating optimization (ADMM–AO) algorithm is preferred in the case that the number of transmit antennas is much larger than that of receive antennas or the amount of data streams is small. Moreover, when equal number of RF chains and data streams are employed, the analytical constant modulus factorization based alternating optimization (ACMF–AO) algorithm is a better choice. Finally, the proposed algorithms can also be well applied in finite resolution phase shifters (PSs) of the analog component and are extended to wideband mmWave systems.