Interference Exploitation-Based Hybrid Precoding with Low-Resolution DACs

Hybrid analog-digital precoding with low-resolution DACs, especially the one-bit case, has been shown to greatly reduce hardware costs and operational power consumption in massive MIMO systems, at the expense of increased radiated power. To address this challenge, we extend the constructive-interfer...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP) S. 604 - 608
Hauptverfasser: Zheng, Xin, Hegde, Ganapati, Pesavento, Marius
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.12.2019
Schlagworte:
Antennas
Computer architecture
Interference
MIMO communication
Optimization
Precoding
Radio frequency
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Hybrid analog-digital precoding with low-resolution DACs, especially the one-bit case, has been shown to greatly reduce hardware costs and operational power consumption in massive MIMO systems, at the expense of increased radiated power. To address this challenge, we extend the constructive-interference-based precoding to this low-cost architecture. We formulate an optimization problem for computing the constructive-interference-based hybrid analog-digital precoders with 1-bit resolution DACs. Our objective is to maximize the quality-of-service at the users subject to a given transmit power budget. Since the resulting problem is a mixed-integer nonlinear program, which is difficult to solve optimally, we decompose the problem into two subtasks, namely, analog precoding and digital precoding. For fixed analog precoders, we formulate a mixed-integer convex program to compute the 1-bit resolution digital precoders. Furthermore, we devise a suboptimal method to solve the formulated mixed-integer convex program problem with low complexity. Simulation results are provided to evaluate the proposed methods.
DOI:10.1109/CAMSAP45676.2019.9022657