Effects of Deep Learning Technologies on Employment in the Field of Digital Communication Systems

2021 ◽  
Vol 12 (4) ◽  
pp. 35-42
Author(s):  
Thomas Alan Woolman ◽  
Philip Lee

There are significant challenges and opportunities facing the economies of the United States in the coming decades of the 21st century that are being driven by elements of technological unemployment. Deep learning systems, an advanced form of machine learning that is often referred to as artificial intelligence, is presently reshaping many aspects of traditional digital communication technology employment, primarily network system administration and network security system design and maintenance. This paper provides an overview of the current state-of-the-art developments associated with deep learning and artificial intelligence and the ongoing revolutions that this technology is having not only on the field of digital communication systems but also related technology fields. This paper will also explore issues and concerns related to past technological unemployment challenges, as well as opportunities that may be present as a result of these ongoing technological upheavals.

2021 ◽  
Author(s):  
Simon Bos ◽  
Evgenii Vinogradov ◽  
Sofie Pollin

Recently, deep learning is considered to optimize the end-to-end performance of digital communication systems. The promise of learning a digital communication scheme from data is attractive, since this makes the scheme adaptable and precisely tunable to many scenarios and channel models. In this paper, we analyse a widely used neural network architecture and show that the training of the end-to-end architecture suffers from normalization errors introduced by an average power constraint. To solve this issue, we propose a modified architecture: shifting the batch slicing after the normalization layer. This approach meets the normalization constraints better, especially in the case of small batch sizes. Finally, we experimentally demonstrate that our modified architecture leads to significantly improved performance of trained models, even for large batch sizes where normalization constraints are more easily met.<br>


2021 ◽  
Author(s):  
Simon Bos ◽  
Evgenii Vinogradov ◽  
Sofie Pollin

Recently, deep learning is considered to optimize the end-to-end performance of digital communication systems. The promise of learning a digital communication scheme from data is attractive, since this makes the scheme adaptable and precisely tunable to many scenarios and channel models. In this paper, we analyse a widely used neural network architecture and show that the training of the end-to-end architecture suffers from normalization errors introduced by an average power constraint. To solve this issue, we propose a modified architecture: shifting the batch slicing after the normalization layer. This approach meets the normalization constraints better, especially in the case of small batch sizes. Finally, we experimentally demonstrate that our modified architecture leads to significantly improved performance of trained models, even for large batch sizes where normalization constraints are more easily met.<br>


2019 ◽  
Vol 23 (1) ◽  
pp. 192-195 ◽  
Author(s):  
Jules M. Moualeu ◽  
Daniel B. da Costa ◽  
Walaa Hamouda ◽  
Ugo S. Dias ◽  
Rausley A. A. de Souza

Sign in / Sign up

Export Citation Format

Share Document