[Paper Review] Deep Learning in Mobile and Wireless Networking: A Survey

Chaoyun Zhang, Paul Patras+1 more|arXiv (Cornell University)|Mar 12, 2018

IoT and Edge/Fog Computing335 references128 citations

TL;DR

This survey reviews how deep learning is applied to mobile and wireless networking, covering models, deployment platforms, applications, tailoring approaches, and open challenges.

ABSTRACT

The rapid uptake of mobile devices and the rising popularity of mobile applications and services pose unprecedented demands on mobile and wireless networking infrastructure. Upcoming 5G systems are evolving to support exploding mobile traffic volumes, agile management of network resource to maximize user experience, and extraction of fine-grained real-time analytics. Fulfilling these tasks is challenging, as mobile environments are increasingly complex, heterogeneous, and evolving. One potential solution is to resort to advanced machine learning techniques to help managing the rise in data volumes and algorithm-driven applications. The recent success of deep learning underpins new and powerful tools that tackle problems in this space. In this paper we bridge the gap between deep learning and mobile and wireless networking research, by presenting a comprehensive survey of the crossovers between the two areas. We first briefly introduce essential background and state-of-the-art in deep learning techniques with potential applications to networking. We then discuss several techniques and platforms that facilitate the efficient deployment of deep learning onto mobile systems. Subsequently, we provide an encyclopedic review of mobile and wireless networking research based on deep learning, which we categorize by different domains. Drawing from our experience, we discuss how to tailor deep learning to mobile environments. We complete this survey by pinpointing current challenges and open future directions for research.

Motivation & Objective

• Bridge the gap between deep learning and mobile/wireless networking and motivate its use in future networks.
• Provide an up-to-date encyclopedic review of deep learning techniques with potential networking applications.
• Discuss practical enablers for deploying deep learning on mobile systems (software/hardware, platforms).
• Offer guidelines for tailoring deep learning models to mobile networking tasks and identify open challenges and directions.

Proposed method

• Categorize and review state-of-the-art deep learning models relevant to mobile and wireless networking.
• Explain core DL concepts and their relevance to network analysis and management.
• Compare architectures and provide model selection guidelines for networking problems.
• Survey DL-enabled applications across domains such as traffic analytics, security, and management.
• Discuss how to tailor DL models to mobile networking problems and outline open research directions.

Experimental results

Research questions

• RQ1 Why is deep learning promising for solving mobile networking problems?
• RQ2 What cutting-edge deep learning models are relevant to mobile and wireless networking?
• RQ3 What are the most recent successful deep learning applications in the mobile networking domain?
• RQ4 How can researchers tailor deep learning to specific mobile networking problems?
• RQ5 What are the most important and promising directions for further study?

Key findings

• Deep learning automates feature extraction, reducing the need for hand-crafted features in heterogeneous mobile data.
• DL scales with big data and leverages SGD for scalable training, mitigating overfitting in large datasets.
• DL enables learning from unlabeled or semi-labeled data using unsupervised or semi-supervised methods.
• Representations learned by DL can be shared across tasks, enabling multi-task learning with reduced retraining needs.
• Specialized DL architectures (e.g., for geometric data like graphs and point clouds) suit mobile data representations.
• The survey also identifies limitations and open challenges, outlining future research directions.