The deployment of 5G wireless communication systems is projected to begin in 2020. With new scenarios, new technologies, and new network architectures, the traffic management for 5G networks will present significant technical challenges. In recent years, AI technologies, especially ML technologies, have demonstrated significant success in many application domains, suggesting their potential to help solve the problem of 5G traffic management. In this article, we investigate the new characteristics of 5G wireless network traffic and discuss challenges they present for 5G traffic management. Potential solutions and research directions for the management of 5G traffic, including distributed and lightweight ML algorithms and a novel AI assistant content retrieval algorithm framework, are discussed.

中文翻译：

---

人工智能管理5G无线网络的网络流量

预计5G无线通信系统的部署将于2020年开始。随着新场景，新技术和新网络架构的出现，5G网络的流量管理将面临重大的技术挑战。近年来，人工智能技术（尤其是机器学习技术）在许多应用领域中均取得了巨大成功，这表明它们有潜力帮助解决5G流量管理问题。在本文中，我们研究了5G无线网络流量的新特征，并讨论了它们对5G流量管理提出的挑战。讨论了5G流量管理的潜在解决方案和研究方向，包括分布式和轻量级ML算法以及新颖的AI助手内容检索算法框架。