Artificial Neural Networks-Based Machine Learning for Wireless Networks: A Tutorial

[sharksmaht]

In order to effectively provide ultra reliable lowlatency communications and pervasive connectivity for Internetof Things (IoT) devices, next-generation wireless networks canleverage intelligent, data-driven functions enabled by the integration of machine learning notions across the wireless core andedge infrastructure. In this context, this paper provides a comprehensive tutorial that overviews how artificial neural networks(ANNs)-based machine learning algorithms can be employed forsolving various wireless networking problems. For this purpose,we first present a detailed overview of a number of key types ofANNs that include recurrent, spiking, and deep neural networks,that are pertinent to wireless networking applications. For eachtype of ANN, we present the basic architecture as well as specificexamples that are particularly important for wireless networkdesign. Such examples include echo state networks, liquid statemachine, and long short term memory. And then, we providean in-depth overview on the variety of wireless communicationproblems that can be addressed using ANNs, ranging fromcommunication using unmanned aerial vehicles to virtual realityapplications over wireless networks and edge computing andcaching. For each individual application, we present the mainmotivation for using ANNs along with the associated challengeswhile we also provide a detailed example for a use case scenarioand outline future works that can be addressed using ANNs. Ina nutshell, this article constitutes the first holistic tutorial on thedevelopment of ANN-based machine learning techniques tailoredto the needs of future wireless networks.

https://sci-hub.se/10.1109/COMST.2019.2926625