ProductAE: Toward Training Larger Channel Codes based on Neural Product Codes
ProductAE: Toward Training Larger Channel Codes based on Neural Product Codes
复制标题
DOI：
10.1109/icc45855.2022.9839215
复制
发表时间：
2021-10
期刊：
影响因子：
--
通讯作者：
Mohammad Vahid Jamali;Hamid Saber;Homayoon Hatami;J. Bae
中科院分区：
文献类型：
--
作者：
Mohammad Vahid Jamali;Hamid Saber;Homayoon Hatami;J. Bae
近年来，已经进行了重要的研究活动，可以通过深度学习自动化渠道编码器和解码器的设计。由于通道编码的维度挑战，因此通过深度学习技术设计和训练相对较大的神经通道代码非常复杂。因此，文献中的大多数结果仅限于相对较短的代码，其信息位少于100个。在本文中，我们构建了Productaes，这是一个由深度学习驱动（编码器，解码器）对的计算高效家族，旨在以可管理的培训复杂性来培训相对较大的通道代码（编码器和解码器）。我们基于经典产品代码的想法，并建议使用较小的代码组件构建大型神经代码。更具体地说，我们提供了一个框架，而不是直接训练大型尺寸K和区块长度N的编码器和解码器，而是提供一个框架，需要培训有关代码参数（N1，K1）和（N2，K2）的神经编码器和解码器此类框架。 N1N2 = N和K1K2 = k。我们的训练结果表明，与连续取消（SC）相比，与参数守则（225,100）的所有信噪比（SNR）（SNR）（SNR）（SNR）（SNR）（SNR）（441,196）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）（SC）解码器。此外，我们的结果表明，涡轮自动编码器（Turboae）和最先进的经典代码有意义的收益。这是设计产品自动编码器和培训大型频道代码的开创性工作的第一项工作。
There have been significant research activities in recent years to automate the design of channel encoders and decoders via deep learning. Due the dimensionality challenge in channel coding, it is prohibitively complex to design and train relatively large neural channel codes via deep learning techniques. Consequently, most of the results in the literature are limited to relatively short codes having less than 100 information bits. In this paper, we construct ProductAEs, a computationally efficient family of deep-learning driven (encoder, decoder) pairs, that aim at enabling the training of relatively large channel codes (both encoders and decoders) with a manageable training complexity. We build upon the ideas from classical product codes, and propose constructing large neural codes using smaller code components. More specifically, instead of directly training the encoder and decoder for a large neural code of dimension k and blocklength n, we provide a framework that requires training neural encoders and decoders for the code parameters (n1,k1) and (n2,k2) such that n1n2 = n and k1k2 = k. Our training results show significant gains, over all ranges of signal-to-noise ratio (SNR), for a code of parameters (225,100) and a moderate-length code of parameters (441,196), over polar codes under successive cancellation (SC) decoder. Moreover, our results demonstrate meaningful gains over Turbo Autoencoder (TurboAE) and state-of-the-art classical codes. This is the first work to design product autoencoders and a pioneering work on training large channel codes.