KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning

Ashok V Makkuva, Xiyang Liu, Mohammad Vahid Jamali, Hessam Mahdavifar, Sewoong Oh, Pramod Viswanath
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:7368-7378, 2021.

Abstract

Landmark codes underpin reliable physical layer communication, e.g., Reed-Muller, BCH, Convolution, Turbo, LDPC, and Polar codes: each is a linear code and represents a mathematical breakthrough. The impact on humanity is huge: each of these codes has been used in global wireless communication standards (satellite, WiFi, cellular). Reliability of communication over the classical additive white Gaussian noise (AWGN) channel enables benchmarking and ranking of the different codes. In this paper, we construct KO codes, a computationally efficient family of deep-learning driven (encoder, decoder) pairs that outperform the state-of-the-art reliability performance on the standardized AWGN channel. KO codes beat state-of-the-art Reed-Muller and Polar codes, under the low-complexity successive cancellation decoding, in the challenging short-to-medium block length regime on the AWGN channel. We show that the gains of KO codes are primarily due to the nonlinear mapping of information bits directly to transmit symbols (bypassing modulation) and yet possess an efficient, high-performance decoder. The key technical innovation that renders this possible is design of a novel family of neural architectures inspired by the computation tree of the {\bf K}ronecker {\bf O}peration (KO) central to Reed-Muller and Polar codes. These architectures pave way for the discovery of a much richer class of hitherto unexplored nonlinear algebraic structures.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-makkuva21a, title = {KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning}, author = {Makkuva, Ashok V and Liu, Xiyang and Jamali, Mohammad Vahid and Mahdavifar, Hessam and Oh, Sewoong and Viswanath, Pramod}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {7368--7378}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/makkuva21a/makkuva21a.pdf}, url = {https://proceedings.mlr.press/v139/makkuva21a.html}, abstract = {Landmark codes underpin reliable physical layer communication, e.g., Reed-Muller, BCH, Convolution, Turbo, LDPC, and Polar codes: each is a linear code and represents a mathematical breakthrough. The impact on humanity is huge: each of these codes has been used in global wireless communication standards (satellite, WiFi, cellular). Reliability of communication over the classical additive white Gaussian noise (AWGN) channel enables benchmarking and ranking of the different codes. In this paper, we construct KO codes, a computationally efficient family of deep-learning driven (encoder, decoder) pairs that outperform the state-of-the-art reliability performance on the standardized AWGN channel. KO codes beat state-of-the-art Reed-Muller and Polar codes, under the low-complexity successive cancellation decoding, in the challenging short-to-medium block length regime on the AWGN channel. We show that the gains of KO codes are primarily due to the nonlinear mapping of information bits directly to transmit symbols (bypassing modulation) and yet possess an efficient, high-performance decoder. The key technical innovation that renders this possible is design of a novel family of neural architectures inspired by the computation tree of the {\bf K}ronecker {\bf O}peration (KO) central to Reed-Muller and Polar codes. These architectures pave way for the discovery of a much richer class of hitherto unexplored nonlinear algebraic structures.} }
Endnote
%0 Conference Paper %T KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning %A Ashok V Makkuva %A Xiyang Liu %A Mohammad Vahid Jamali %A Hessam Mahdavifar %A Sewoong Oh %A Pramod Viswanath %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-makkuva21a %I PMLR %P 7368--7378 %U https://proceedings.mlr.press/v139/makkuva21a.html %V 139 %X Landmark codes underpin reliable physical layer communication, e.g., Reed-Muller, BCH, Convolution, Turbo, LDPC, and Polar codes: each is a linear code and represents a mathematical breakthrough. The impact on humanity is huge: each of these codes has been used in global wireless communication standards (satellite, WiFi, cellular). Reliability of communication over the classical additive white Gaussian noise (AWGN) channel enables benchmarking and ranking of the different codes. In this paper, we construct KO codes, a computationally efficient family of deep-learning driven (encoder, decoder) pairs that outperform the state-of-the-art reliability performance on the standardized AWGN channel. KO codes beat state-of-the-art Reed-Muller and Polar codes, under the low-complexity successive cancellation decoding, in the challenging short-to-medium block length regime on the AWGN channel. We show that the gains of KO codes are primarily due to the nonlinear mapping of information bits directly to transmit symbols (bypassing modulation) and yet possess an efficient, high-performance decoder. The key technical innovation that renders this possible is design of a novel family of neural architectures inspired by the computation tree of the {\bf K}ronecker {\bf O}peration (KO) central to Reed-Muller and Polar codes. These architectures pave way for the discovery of a much richer class of hitherto unexplored nonlinear algebraic structures.
APA
Makkuva, A.V., Liu, X., Jamali, M.V., Mahdavifar, H., Oh, S. & Viswanath, P.. (2021). KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:7368-7378 Available from https://proceedings.mlr.press/v139/makkuva21a.html.

Related Material