We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: Design Trade-offs for Decentralized Baseband Processing in Massive MU-MIMO Systems

Abstract: Massive multi-user (MU) multiple-input multiple-output (MIMO) provides high spectral efficiency by means of spatial multiplexing and fine-grained beamforming. However, conventional base-station (BS) architectures for systems with hundreds of antennas that rely on centralized baseband processing inevitably suffer from (i) excessive interconnect data rates between radio-frequency circuitry and processing fabrics, and (ii) prohibitive complexity at the centralized baseband processor. Recently, decentralized baseband processing (DBP) architectures and algorithms have been proposed, which mitigate the interconnect bandwidth and complexity bottlenecks. This paper systematically explores the design trade-offs between error-rate performance, computational complexity, and data transfer latency of DBP architectures under different system configurations and channel conditions. Considering architecture, algorithm, and numerical precision aspects, we provide practical guidelines to select the DBP architecture and algorithm that are able to realize the full benefits of massive MU-MIMO in the uplink and downlink.
Comments: 2019 IEEE Asilomar conference paper
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Cite as: arXiv:1912.04437 [cs.IT]
  (or arXiv:1912.04437v2 [cs.IT] for this version)

Submission history

From: Kaipeng Li [view email]
[v1] Tue, 10 Dec 2019 01:05:04 GMT (770kb,D)
[v2] Sun, 15 Dec 2019 04:37:55 GMT (801kb,D)

Link back to: arXiv, form interface, contact.