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

Download:

Current browse context:

cs.IT

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

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

Computer Science > Information Theory

Title: Joint Source-Channel Coding for Semantics-Aware Grant-Free Radio Access in IoT Fog Networks

Abstract: A fog-radio access network (F-RAN) architecture is studied for an Internet-of-Things (IoT) system in which wireless sensors monitor a number of multi-valued events and transmit in the uplink using grant-free random access to multiple edge nodes (ENs). Each EN is connected to a central processor (CP) via a finite-capacity fronthaul link. In contrast to conventional information-agnostic protocols based on separate source-channel (SSC) coding, where each device uses a separate codebook, this paper considers an information-centric approach based on joint source-channel (JSC) coding via a non-orthogonal generalization of type-based multiple access (TBMA). By leveraging the semantics of the observed signals, all sensors measuring the same event share the same codebook (with non-orthogonal codewords), and all such sensors making the same local estimate of the event transmit the same codeword. The F-RAN architecture directly detects the events values without first performing individual decoding for each device. Cloud and edge detection schemes based on Bayesian message passing are designed and trade-offs between cloud and edge processing are assessed.
Comments: submitted for publication
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
DOI: 10.1109/LSP.2021.3072278
Cite as: arXiv:2101.11309 [cs.IT]
  (or arXiv:2101.11309v1 [cs.IT] for this version)

Submission history

From: Johannes Dommel [view email]
[v1] Wed, 27 Jan 2021 10:41:36 GMT (3313kb,D)

Link back to: arXiv, form interface, contact.