• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IEEE Workshop on Signal Processing Systems >Principles and Techniques to Optimize Digital Communication Systems from End to End
【24h】

Principles and Techniques to Optimize Digital Communication Systems from End to End

机译：从一端到另一端优化数字通信系统的原理和技术

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Digital communication systems, serving to reliably transmit information in binary form, exhibit a complex input to output response that is part stochastic, part deterministic, which depends not only on the communication channel but also on the underlying hardware. The amount and type of stochasticity is typically unspecified, and the deterministic in-to-out response is most often unknown, where both phenomena are only coarsely and improperly modeled in practice. Common methods for communication rely on the use generic textbook modulation formats and demodulation schemes to imprint and extract bits to and from waveforms, yielding modulation and detection schemes that are neither optimized nor tailored for the system at hand. As a result, conventional bits-to-waveform and waveform-to-bits mapping and demapping methods leave performance on the table and do not exploit the full communication capabilities of the system. Learning to communicate through the physical layer presents several challenges. In this paper we review the fundamentals of end-to-end learning for digital communication systems. We review different techniques to learn to recover at the system's output the symbols or bits presented at its input as well as different metrics to minimize to achieve this goal, with interpretations of the underlying signals. Techniques to train the transmitter are addressed as well as other challenges of end-to-end learning for digital communication systems.
机译：用于可靠地以二进制形式传输信息的数字通信系统展现了复杂的输入到输出响应，该响应是部分随机的，部分确定的，这不仅取决于通信通道，而且还取决于底层硬件。随机性的数量和类型通常是不确定的，并且确定性的由内而外的响应通常是未知的，在实践中，这两种现象都只是粗略地和不正确地建模。常用的通信方法依赖于使用通用的教科书调制格式和解调方案来在波形上压印和从波形中提取位，从而产生既未针对现有系统进行优化也未针对其定制的调制和检测方案。结果，传统的位到波形和波形到位的映射和解映射方法使性能无法发挥作用，并且无法利用系统的全部通信功能。学会通过物理层进行通信提出了一些挑战。在本文中，我们回顾了数字通信系统端到端学习的基础。我们回顾了不同的技术，以了解如何在系统的输出处恢复在其输入处出现的符号或位以及不同的度量，以最小化实现此目标以及对基本信号的解释。解决了训练发射机的技术以及数字通信系统端到端学习的其他挑战。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号