Autoencoder-Based Unequal Error Protection Codes
Artikel i vetenskaplig tidskrift, 2021

We present a novel autoencoder-based approach for designing codes that provide unequal error protection (UEP) capabilities. The proposed approach, based on a generalization of an autoencoder loss function, provides a versatile framework for the design of message-wise and bit-wise UEP codes. Using an associated weight vector, the generalized loss function can be used to trade off error probabilities corresponding to different importance classes and to explore the region of achievable error probabilities. For message-wise UEP, we compare the proposed autoencoder-based UEP codes with a union of random coset codes. For bit-wise UEP, the proposed codes are compared with UEP rateless spinal codes and the superposition of random Gaussian codes. In all cases, the autoencoder-based codes show superior performance while providing design simplicity and flexibility in trading off error protection among different importance classes.

deep learning


Error correction codes



Neural networks



Error probability

unequal error protection


Vukan Ninkovic

University of Novi Sad

Dejan Vukobratovic

University of Novi Sad

Christian Häger

Chalmers, Elektroteknik, Kommunikations- och antennsystem, Kommunikationssystem

Henk Wymeersch

Chalmers, Elektroteknik, Kommunikations- och antennsystem, Kommunikationssystem

Alexandre Graell I Amat

Chalmers, Elektroteknik, Kommunikations- och antennsystem, Kommunikationssystem

IEEE Communications Letters

1089-7798 (ISSN)

Vol. 25 11 3575-3579

INnovation and excellence in massive-scale COMmunications and information processING - INCOMING

Europeiska kommissionen (EU) (856967-INCOMING), 2020-01-01 -- 2022-12-31.



Inbäddad systemteknik

Sannolikhetsteori och statistik



Mer information

Senast uppdaterat