Recent advances in communications and networking have enabled the support of real-life applications within several Internet of Things (IoT) verticals including smart city, smart transportation, smart utilities and E-health. In order to support the ever-increasing number of IoT devices and heterogeneous applications, it is crucial to design resource-efficient and scalable upcoming 5G and beyond systems in a way that they can operate in the complex wireless environment while supporting the Machine-Type Communications (MTC)/IoT traffic. Towards enabling the effective design and operation of these IoT systems, the application of Artificial Intelligence (AI) and Machine Learning (ML) techniques has recently caught the attention of various research communities and the related stakeholders. On one hand, some researchers argue that employing ML in communication systems offers little benefits because communication systems were primarily designed for bandwidth, power and complexity optimization. On the other hand, big data-driven solutions, including Deep Learning (DL) can be highly advantageous for data-driven prediction, analysis and performance improvement by utilizing time-dependent properties of network elements. This could be achieved by adapting the ML/DL models to the dynamicity of the underlying communication environment along with the training data generated by the IoT devices. In addition, the DL techniques may provide significant benefits towards automating the conventional acquisition/processing/computing tasks in communication networks supporting heterogeneous IoT applications. However, there are several issues to be addressed for the applications of ML/DL techniques in the complex IoT environment involving resource-constrained IoT/MTC devices, including heterogeneity, limited computational capability, distributed nature, and distinct traffic characteristics.

This workshop focuses on recent research activities related to AI and ML in communication and networking tasks for various IoT verticals, and aims to open a debate on the applications of AI and ML/DL techniques in emerging IoT networks and applications. In this direction, we invite researchers from the academia, industry, and governmental organizations to submit their novel works on system architectures, learning models, theoretical models/algorithms, system level simulations/experimental results, hardware demonstration results, and standardization activities in the related areas.

The main topics of interest for this special issue include, but not limited to the following:

· Supervised, unsupervised and reinforcement learning for IoT systems

· Deep learning for IoT Applications

· Active learning in wireless IoT systems

· Learning-assisted resource allocation and management for IoT

· ML for access congestion management in edge IoT networks

· Bayesian networks for IoT applications

· Fuzzy logic based learning for IoT applications

· ML/DL for bioinformatics/E-healthcare/smart city/smart home

· Machine learning for traffic management in IoT

· ML-assisted decision automation in Industrial IoT

· Knowledge acquisition, discovery and learning for IoT

· Machine learning for information retrieval in IoT

· Machine learning for web navigation and data mining in IoT

· Learning through mobile data mining in IoT systems

· Distributed and parallel learning algorithms in IoT applications

· Feature extraction and classification in IoT Applications

· Computational learning theory for IoT

· Hybrid learning algorithms for IoT systems

· ML-assisted security enhancement for IoT systems

· ML-assisted intrusion and malware detection for IoT

· Learning-assisted privacy preservation and trust for IoT

· Adversial ML for IoT systems

Accepted and presented papers will appear in the conference proceedings of IEEE SECON 2019 and will be submitted for inclusion in IEEE Xplore®.

Important Dates:

· Review paper submission: 15 April 2019 (Extended to 25 April 2019)

· Notification of acceptance: 8 May 2019

· Camera-ready submission: 18 May 2019