Communication Algorithms via Deep Learning
Authors: Hyeji Kim, Yihan Jiang, Ranvir B. Rana, Sreeram Kannan, Sewoong Oh, Pramod Viswanath
ICLR 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this paper we study whether it is possible to automate the discovery of decoding algorithms via deep learning. We study a family of sequential codes parametrized by recurrent neural network (RNN) architectures. We show that creatively designed and trained RNN architectures can decode well known sequential codes such as the convolutional and turbo codes with close to optimal performance on the additive white Gaussian noise (AWGN) channel, which itself is achieved by breakthrough algorithms of our times (Viterbi and BCJR decoders, representing dynamic programing and forward-backward algorithms). We show strong generalizations, i.e., we train at a speciļ¬c signal to noise ratio and block length but test at a wide range of these quantities, as well as robustness and adaptivity to deviations from the AWGN setting. |
| Researcher Affiliation | Academia | Hyeji Kim , Yihan Jiang , Ranvir Rana*, Sreeram Kannan , Sewoong Oh*, Pramod Viswanath* University of Illinois at Urbana Champaign*, University of Washington |
| Pseudocode | No | The paper does not include pseudocode or clearly labeled algorithm blocks. It describes the architectures and procedures in text and diagrams. |
| Open Source Code | Yes | Source codes available in https://github.com/yihanjiang/Sequential-RNN-Decoder |
| Open Datasets | No | The paper states, 'We generate N training examples (y(i), b(i)) for i [N] according to this joint distribution to train our model.' This indicates data was generated rather than using a specific publicly available dataset with concrete access information. |
| Dataset Splits | No | The paper discusses 'training data' and 'test data' but does not explicitly mention a separate 'validation' set or specify explicit train/validation/test splits (e.g., percentages or counts). |
| Hardware Specification | No | The paper does not specify any particular hardware used for running the experiments (e.g., GPU model, CPU type, memory, or cloud instance names). |
| Software Dependencies | No | The paper mentions 'a python implementation of BCJR in Taranalli (2015)' but does not provide specific version numbers for Python, the 'Commpy' library, or any deep learning frameworks (like TensorFlow or PyTorch) that would have been used. |
| Experiment Setup | Yes | We trained the decoder with Adam optimizer with learning rate 1e-3, batch size 200, and total number of examples is 12,000, and we use clip norm... We train with 10 epochs and ADAM optimizer with learning rate 0.001. |