Designing codes that combat the noise in a communication medium has remained a significant area of research in information theory as well as wireless communications. Asymptotically optimal channel codes have been developed by mathematicians for communicating under canonical models after over 60 years of research. On the other hand, in many non-canonical channel settings, optimal codes do not exist and the codes designed for canonical models are adapted via heuristics to these channels and are thus not guaranteed to be optimal. In this work, we make significant progress on this problem by designing a fully end-to-end jointly trained neural encoder and decoder, namely, Turbo Autoencoder (TurboAE), with the following contributions: (a) under moderate block lengths, TurboAE approaches state-of-the-art performance under canonical channels; (b) moreover, TurboAE outperforms the state-of-the-art codes under non-canonical settings in terms of reliability. TurboAE shows that the development of channel coding design can be automated via deep learning, with near-optimal performance.

In recent years, several papers have employed deep learning methods to decode various classes of codes (turbo codes, linear codes, polar codes). This work focuses on turbo codes, and has the more ambitious goal of providing joint training of the decoder and the encoder (which means that the resulting code will not be a turbo code in the traditional sense). The authors borrow some ideas from the turbo coding literature (e.g., interleaving) and use CNNs to design decoder and encoder (as opposed to RNNs used in several other papers). The proposed TurboAE algorithm achieves performance which is comparable to state-of-the-art codes (see Figure 1). This is quite impressive, even though the code length is quite short (i.e. 100 bits).

This paper considers a combination of encoder and decoder architecture, which is a serially-concatenated code with interleavers, as in the turbo codes, combined with turbo-like iterative decoding, and proposes implementing encoders and decoders of the constituent codes with 1D-CNN, which allow us to train the encoders and the decoders in an end-to-end and data-driven fashion. Two reviewers raised concern about the scalability issue of the proposal, and the authors admit in their rebuttal that it is a central question. Although the review scores exhibited a large split in the initial round of review, mainly due to the scalability issue as well as comparison in performance with other existing coding schemes, after the authors' rebuttal all the reviewers rated this paper above the acceptance threshold. I would therefore like to recommend acceptance of this paper.

Designing codes that combat the noise in a communication medium has remained a significant area of research in information theory as well as wireless communications. Asymptotically optimal channel codes have been developed by mathematicians for communicating under canonical models after over 60 years of research. On the other hand, in many non-canonical channel settings, optimal codes do not exist and the codes designed for canonical models are adapted via heuristics to these channels and are thus not guaranteed to be optimal. In this work, we make significant progress on this problem by designing a fully end-to-end jointly trained neural encoder and decoder, namely, Turbo Autoencoder (TurboAE), with the following contributions: (a) under moderate block lengths, TurboAE approaches state-of-the-art performance under canonical channels; (b) moreover, TurboAE outperforms the state-of-the-art codes under non-canonical settings in terms of reliability. TurboAE shows that the development of channel coding design can be automated via deep learning, with near-optimal performance.

Designing codes that combat the noise in a communication medium has remained a significant area of research in information theory as well as wireless communications. Asymptotically optimal channel codes have been developed by mathematicians for communicating under canonical models after over 60 years of research. On the other hand, in many non-canonical channel settings, optimal codes do not exist and the codes designed for canonical models are adapted via heuristics to these channels and are thus not guaranteed to be optimal. In this work, we make significant progress on this problem by designing a fully end-to-end jointly trained neural encoder and decoder, namely, Turbo Autoencoder (TurboAE), with the following contributions: (a) under moderate block lengths, TurboAE approaches state-of-the-art performance under canonical channels; (b) moreover, TurboAE outperforms the state-of-the-art codes under non-canonical settings in terms of reliability. TurboAE shows that the development of channel coding design can be automated via deep learning, with near-optimal performance.