Data detection in 1-bit quantized massive MIMO
Thesis event information
Date and time of the thesis defence
Place of the thesis defence
L5 hall, Linnanmaa Campus
Topic of the dissertation
Data detection in 1-bit quantized massive MIMO
Doctoral candidate
Master of Science Amin Radbord
Faculty and unit
University of Oulu Graduate School, Faculty of Information Technology and Electrical Engineering, Centre for Wireless Communications
Subject of study
Communications Engineering
Opponent
Professor Gonzalo Seco-Granados, Universitat Autònoma de Barcelona, Spain
Custos
Associate Professor Italo Atzeni, University of Oulu
Data detection in 1-bit quantized massive MIMO
Future wireless systems operating in high-frequency bands will rely heavily on large antenna arrays to support increasing data-rate and connectivity demands. Massive multiple-input multiple-output (MIMO) is a key enabler for such systems, but the large number of required data converters increases power consumption and hardware complexity. Low-resolution data converters, especially 1-bit analog-to-digital converters (ADCs) and digital-to-analog converters (DACs), provide an attractive solution, although the resulting severe quantization distortion changes the signal statistics and complicates data detection.
This thesis addresses data detection in 1-bit quantized massive MIMO systems. The first part develops an analytical framework for uplink massive MIMO systems with 1-bit ADCs. The expected values of the soft-estimated symbols after linear combining are characterized under the effects of 1-bit quantization in both channel estimation and uplink data transmission. Based on this analysis, a linear minimum mean dispersion (LMMD) receiver and a joint data detection (JD) method are proposed to improve data detection by accounting for the expected values of the soft-estimated symbols and their interdependence among the interfering user equipments (UEs). Numerical results examining the symbol error rate show that the proposed LMMD receiver significantly outperforms conventional receivers. Moreover, the proposed JD and its low-complexity variant provide a significant boost compared with the UE-specific data detection.
The second part investigates data detection in a point-to-point massive MIMO system with 1-bit DACs at the transmitter and either full-resolution or 1-bit ADCs at the receiver. The linearly precoded signal is dithered prior to quantization, and soft-estimation-based data detection methods with symbol-independent dither removal are developed. In addition, a symbol-dependent linearization of the transmitted signal at the output of the 1-bit DACs is introduced and used to derive maximum likelihood (ML)-based data detection methods for the receivers with either full-resolution or 1-bit ADCs. Moreover, low-complexity variants of these ML-based methods are also proposed. The proposed ML-based methods and their low-complexity variants achieve significant gains over a baseline based on binary ML detection via a homotopy algorithm.
This thesis addresses data detection in 1-bit quantized massive MIMO systems. The first part develops an analytical framework for uplink massive MIMO systems with 1-bit ADCs. The expected values of the soft-estimated symbols after linear combining are characterized under the effects of 1-bit quantization in both channel estimation and uplink data transmission. Based on this analysis, a linear minimum mean dispersion (LMMD) receiver and a joint data detection (JD) method are proposed to improve data detection by accounting for the expected values of the soft-estimated symbols and their interdependence among the interfering user equipments (UEs). Numerical results examining the symbol error rate show that the proposed LMMD receiver significantly outperforms conventional receivers. Moreover, the proposed JD and its low-complexity variant provide a significant boost compared with the UE-specific data detection.
The second part investigates data detection in a point-to-point massive MIMO system with 1-bit DACs at the transmitter and either full-resolution or 1-bit ADCs at the receiver. The linearly precoded signal is dithered prior to quantization, and soft-estimation-based data detection methods with symbol-independent dither removal are developed. In addition, a symbol-dependent linearization of the transmitted signal at the output of the 1-bit DACs is introduced and used to derive maximum likelihood (ML)-based data detection methods for the receivers with either full-resolution or 1-bit ADCs. Moreover, low-complexity variants of these ML-based methods are also proposed. The proposed ML-based methods and their low-complexity variants achieve significant gains over a baseline based on binary ML detection via a homotopy algorithm.
Created 9.11.2026 | Updated 9.11.2026