Advanced persistent threat, network intrusions and anomalies, and invasions of location and data privacy are constantly violating edge networks, disturbing guaranteed services, and further destroying critical end devices and cyberinfrastructure. Edge gateway defense against malicious cyber threats and privacy leakage for a multitude of connected physical devices remains a significant problem. Edge computing empowered network intelligence will become a practical solution to tackle this challenge by building gateway-focused protection mechanisms, dynamically extending the edge network, and establishing a flexible and intelligent platform to perform rapid responses with the cloud and wide area network. In this article, we present the research in edge intelligence assisted gateway defense in cyber security. Specifically, we first provide the overview of edge computing, and then discuss the motivations of developing and deploying a robust and effective edge intelligence. We also present the details of edge intelligence assisted gateway defense, including major cyber threats and defense mechanisms at gateways. We further offer detailed descriptions of the emerging data-driven approaches for gateway defense. Finally, we perform an experimental study, and we demonstrate that robust edge intelligence achieves high detection accuracy and effective computational performance.

中文翻译：

---

边缘智能辅助网络安全中的网关防御

高级持续威胁，网络入侵和异常以及位置和数据隐私的入侵不断违反边缘网络，扰乱了有保证的服务，并进一步破坏了关键的终端设备和网络基础设施。边缘网关针对大量已连接物理设备的恶意网络威胁和隐私泄漏的防御仍然是一个重大问题。支持边缘计算的网络智能将通过构建以网关为中心的保护机制，动态扩展边缘网络以及建立灵活的智能平台来对云和广域网执行快速响应，从而成为应对这一挑战的实用解决方案。在本文中，我们介绍了网络安全中边缘智能辅助网关防御的研究。特别，我们首先提供边缘计算的概述，然后讨论开发和部署强大有效的边缘智能的动机。我们还介绍了边缘智能辅助网关防御的详细信息，包括网关的主要网络威胁和防御机制。我们进一步提供了新兴的数据驱动网关防御方法的详细描述。最后，我们进行了一项实验研究，并证明了强大的边缘智能功能可实现较高的检测精度和有效的计算性能。我们进一步提供了新兴的网关驱动数据驱动方法的详细说明。最后，我们进行了一项实验研究，并证明了强大的边缘智能功能可实现较高的检测精度和有效的计算性能。我们进一步提供了新兴的数据驱动网关防御方法的详细描述。最后，我们进行了一项实验研究，并证明了强大的边缘智能功能可实现较高的检测精度和有效的计算性能。