Mixed-ADC Massive MIMO Uplink in Frequency-Selective Channels

Abstract

The aim of this paper is to investigate the recently developed mixed-ADC architecture for frequency-selective channels. Multi-carrier techniques such as orthogonal frequency division multiplexing (OFDM) are employed to handle inter-symbol interference (ISI). A frequency-domain equalizer is designed for mitigating the inter-carrier interference (ICI) introduced by the nonlinearity of one-bit quantization. For static single-input-multiple-output (SIMO) channels, a closed-form expression of the generalized mutual information (GMI) is derived, and based on which the linear frequency-domain equalizer is optimized. The analysis is then extended to ergodic time-varying SIMO channels with estimated channel state information (CSI), where numerically tight lower and upper bounds of the GMI are derived. The analytical framework is naturally applicable to the multi-user scenario, for both static and time-varying channels. Extensive numerical studies reveal that the mixed-ADC architecture with a small proportion of high-resolution ADCs does achieve a dominant portion of the achievable rate of ideal conventional architecture, and that it remarkably improves the performance as compared with one-bit massive MIMO.Comment: 14 pages, 10 figures, to appear in IEEE Transactions on Communication

Performance Analysis of Multi-Cell Millimeter-Wave Massive MIMO Networks With Low-Precision ADCs

“…LEMMA 1 Lemma 1. The analog beamforming gain within the cell l, c llk , defined in (8) is bounded as follows…”

“…Furthermore, the spectral efficiencies of a mixed-ADC system under energy constraint has been studied in [7]. The mixed-ADC architecture in frequency-selective channels has been investigated in [8]. It has been demonstrated in [9] that low-precision, e.g., 2-3 bits, ADCs only cause limited sum rate loss under some mild assumptions for an amplifyand-forward relay uplink network.…”

Reliable OFDM Receiver With Ultra-Low Resolution ADC

“…As indicated by the capacity derivation for additive white Gaussian noise (AWGN) channels in [4], only 10%-20% of capacity is lost under the quantization precision of as few as 2−3 bits. The research in [7] revealed that in multiple-input and multipleoutput (MIMO) systems with mixed-precision ADCs, equipping only a small proportion of high-resolution ADCs can closely approximate the achievable rate of conventional architecture with all full-resolution ADCs. These results inspired subsequent studies into signal recovery techniques for dealing with strong nonlinear distortion caused by coarse quantization.…”

Efficient Equalization Method for Cyclic Prefix-Free Coarsely Quantized Massive MIMO Systems

“…The achievable rates and equalization with a mix of low and high-resolution ADC architecture with perfect Channel State Information (CSI) is discussed in [6]. The computational complexity of the proposed methods [5], [6] are in general high. In [7], low-complexity linear channel estimation and receive combiners for multiuser symbol detection were proposed and a lower bound on capacity was also derived using linear schemes.…”