Soft-Output Finite Alphabet Equalization for mmWave Massive MIMO

Abstract

Next-generation wireless systems are expected to combine millimeter-wave (mmWave) and massive multiuser multipleinput multiple-output (MU-MIMO) technologies to deliver high data-rates. These technologies require the basestations (BSs) to process high-dimensional data at extreme rates, which results in high power dissipation and system costs. Finitealphabet equalization has been proposed recently to reduce the power consumption and silicon area of uplink spatial equalization circuitry at the BS by coarsely quantizing the equalization matrix. In this work, we improve upon finite-alphabet equalization by performing unbiased estimation and soft-output computation for coded systems. By simulating a massive MU-MIMO system that uses orthogonal frequency-division multiplexing and per-user convolutional coding, we show that soft-output finite-alphabet equalization delivers competitive error-rate performance using only 1 to 3 bits per entry of the equalization matrix, even for challenging mmWave channels.

An Overview of Signal Processing Techniques for Terahertz Communications

“…For a broadband SC THz system, MMSE precoding and detection is explored in [269] by assuming sparse channel matrices. Finite alphabit equalization [270], which reduces the complexity, power consumption, and circuit area (by coarsely quantizing the equalization matrix), can also prove to be useful for THz scenarios. Indoor THz communications Tbps rates under finite alphabets are demonstrated in [183], where a frequencydivision scheme of multiple sub-bands is utilized to relax the requirements on ADCs and DACs.…”

A Projected Gradient Descent Algorithm for Designing Low-Resolution Finite-Alphabet Equalizers in All-Digital Massive MU-MIMO Communication Systems

“…Remark 1: Spatial equalization matrices conventionally use ℓ-bit quantization for each coefficient in both the real and imaginary components [16], [17], [18]. For instance, the 1-bit finite-alphabet X 1 consists of four symbols: {+1 + j, +1 − j, −1 + j, −1 − j}, with one bit assigned to each component.…”

“…Nonetheless, the spatial equalization matrix in the baseband processing still requires 10-12 bits for quantization [12], [13], [14], [15], resulting in a significant increase in power consumption and silicon chip area. To mitigate these concerns, researchers have proposed the use of finite-alphabet equalizers, which crudely quantize the spatial equalization matrix coefficients [16], [17], [18]. This approach mitigates power consumption and silicon area concerns while taking advantage of the benefits of all-digital BS architectures.…”

Resolution-Adaptive All-Digital Spatial Equalization for mmWave Massive MU-MIMO

“…The vector µ contains post-equalization scaling factors that are represented with more bits (e.g., 10-bit). In order to obtain unbiased estimates of the transmitted symbols, the entries of the scaling vector µ are set to [22]…”