Generative Adversarial Networks (GANs) in networking: A comprehensive survey & evaluation
It is demonstrated how this branch of machine learning can benefit multiple aspects of computer and communication networks, including mobile networks, network analysis, internet of things, physical layer, and cybersecurity.
Abstract
Despite the recency of their conception, Generative Adversarial Networks\n(GANs) constitute an extensively researched machine learning sub-field for the\ncreation of synthetic data through deep generative modeling. GANs have\nconsequently been applied in a number of domains, most notably computer vision,\nin which they are typically used to generate or transform synthetic images.\nGiven their relative ease of use, it is therefore natural that researchers in\nthe field of networking (which has seen extensive application of deep learning\nmethods) should take an interest in GAN-based approaches. The need for a\ncomprehensive survey of such activity is therefore urgent. In this paper, we\ndemonstrate how this branch of machine learning can benefit multiple aspects of\ncomputer and communication networks, including mobile networks, network\nanalysis, internet of things, physical layer, and cybersecurity. In doing so,\nwe shall provide a novel evaluation framework for comparing the performance of\ndifferent models in non-image applications, applying this to a number of\nreference network datasets.\n