• Corpus ID: 239024894

Wideband and Entropy-Aware Deep Soft Bit Quantization

  title={Wideband and Entropy-Aware Deep Soft Bit Quantization},
  author={Marius Arvinte and Jonathan I. Tamir},
Deep learning has been recently applied to physical layer processing in digital communication systems in order to improve end-to-end performance. In this work, we introduce a novel deep learning solution for soft bit quantization across wideband channels. Our method is trained end-to-end with quantizationand entropy-aware augmentations to the loss function and is used at inference in conjunction with source coding to achieve near-optimal compression gains over wideband channels. To efficiently… 

Figures from this paper


Deep Log-Likelihood Ratio Quantization
A deep learning-based method for log-likelihood ratio (LLR) lossy compression and quantization is proposed, with emphasis on a single-input single-output uncorrelated fading communication setting, and the method is competitive with state-of-the-art approaches.
Learning to Quantize Deep Networks by Optimizing Quantization Intervals With Task Loss
This work proposes a trainable quantizer that can be trained on a heterogeneous dataset, and thus can be used to quantize pretrained networks without access to their training data, and outperforms existing methods to achieve the state-of-the-art accuracy.
Soft-to-Hard Vector Quantization for End-to-End Learning Compressible Representations
We present a new approach to learn compressible representations in deep architectures with an end-to-end training strategy. Our method is based on a soft (continuous) relaxation of quantization and
Practical Full Resolution Learned Lossless Image Compression
The first practical learned lossless image compression system, L3C, is proposed and it outperforms the popular engineered codecs, PNG, WebP and JPEG 2000, and finds that learning the auxiliary representation is crucial and outperforms predefined auxiliary representations such as an RGB pyramid significantly.
Quantization for soft-output demodulators in bit-interleaved coded modulation systems
The quantizer is proposed to design such that the quantizer outputs become equiprobable, and bit error rate simulations for BICM systems with LLR quantization using a rate 1/2 low-density parity-check code are shown.
Generating Diverse High-Fidelity Images with VQ-VAE-2
It is demonstrated that a multi-scale hierarchical organization of VQ-VAE, augmented with powerful priors over the latent codes, is able to generate samples with quality that rivals that of state of the art Generative Adversarial Networks on multifaceted datasets such as ImageNet, while not suffering from GAN's known shortcomings such as mode collapse and lack of diversity.
Quantization of Log-Likelihood Ratios to Maximize Mutual Information
  • W. Rave
  • Computer Science
    IEEE Signal Processing Letters
  • 2009
A quantization scheme for log-likelihood ratios which optimizes the trade-off between rate and accuracy in the sense of rate distortion theory and is slightly superior to the previously proposed idea of applying the Lloyd-Max algorithm to the dasiasoft bitpsila density associated to the L-values.
Neural Discrete Representation Learning
Pairing these representations with an autoregressive prior, the model can generate high quality images, videos, and speech as well as doing high quality speaker conversion and unsupervised learning of phonemes, providing further evidence of the utility of the learnt representations.
Dynamics and performance analysis of analog iterative decoding for low-density parity-check (LDPC) codes
This paper investigates the dynamics of a continuous-time (asynchronous) analog implementation of iterative decoding, and shows that it can be approximated as the application of the well-known successive relaxation (SR) method for solving the fixed-point problem.
On quantization of log-likelihood ratios for maximum mutual information
  • A. Winkelbauer, G. Matz
  • Computer Science
    2015 IEEE 16th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC)
  • 2015
The proposed algorithm is reminiscent of the famous Lloyd-Max algorithm and is not restricted to any particular LLR distribution and can be used to design LLR quantizers during data transmission.