| Autor: |
Seyyed Ali Hashemi, Nghia Doan, Elie Ngomseu Mambou, Warren J. Gross |
| Rok vydání: |
2019 |
| Předmět: |
|
| Zdroj: |
Machine Learning for Future Wireless Communications |
| DOI: |
10.1002/9781119562306.ch15 |
| Popis: |
This chapter provides the background and motivation for the use of deep learning (DL) in various forward error correction schemes used for wireless communication systems. Polar codes are a recent breakthrough in the field of channel coding, as they were proven to achieve channel capacity with efficient encoding and decoding algorithms. Successive cancellation and belief propagation decoding algorithms are first introduced to decode polar codes. The chapter provides some basic knowledge about polar codes and conventional polar decoders. It then discusses several DL‐based decoding algorithms and their variants for polar codes, followed by a detailed evaluation concerning error‐correction performance and decoding latency of state‐of‐the‐art DL‐aided decoders for a 5G polar code. The chapter also describes the use of DL in decoding polar codes with emphasis on off‐the‐shelf DL decoders and DL‐aided decoders by addressing their working principles, algorithm details, and performance evaluations. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
