The concept of edge caching provision in emerging 5G and beyond mobile networks is a promising method to deal both with the traffic congestion problem in the core network as well as reducing latency to access popular content. In that respect end user demand for popular content can be satisfied by proactively caching it at the network edge, i.e, at close proximity to the users. In addition to model based caching schemes learning-based edge caching optimizations has recently attracted significant attention and the aim hereafter is to capture these recent advances for both model based and data driven techniques in the area of proactive caching. This paper summarizes the utilization of deep learning for data caching in edge network. We first outline the typical research topics in content caching and formulate a taxonomy based on network hierarchical structure. Then, a number of key types of deep learning algorithms are presented, ranging from supervised learning to unsupervised learning as well as reinforcement learning. Furthermore, a comparison of state-of-the-art literature is provided from the aspects of caching topics and deep learning methods. Finally, we discuss research challenges and future directions of applying deep learning for caching
A Survey on Reinforcement Learning-Aided Caching in Mobile Edge Networks
IRIDA Research Centre for Communication Technologies（IRIDA通信技术研究中心）；University of Cyprus（塞浦路斯大学）；Ionian University（爱奥尼亚大学）；Aalto University（阿尔托大学）；Rutgers University（罗格斯大学）
Distributed Edge Caching via Reinforcement Learning in Fog Radio Access Networks
Southeast University（东南大学）；Xidian University（西安电子科技大学）；Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences（中国科学院上海微系统与信息技术研究所）；University of Oulu（奥卢大学）；University of Manchester（曼彻斯特大学）；Harbin Institute of Technology（哈尔滨工业大学）
Deep Reinforcement Learning and Permissioned Blockchain for Content Caching in Vehicular Edge Computing and Networks
University of Electronic Science and Technology of China（电子科技大学）；Simula Metropolitan Center for Digital Engineering（Simula都市数字工程中心）；University of Oslo（奥斯陆大学）
Attention-Enhanced Prioritized Proximal Policy Optimization for Adaptive Edge Caching
Lassonde School of Engineering at York University（约克大学拉松德工程学院）；School of Electronic and Computer Engineering at the Hong Kong University of Science and Technology（香港科技大学电子与计算机工程学院）；School of Engineering and Applied Science at the University of Pennsylvania（宾夕法尼亚大学工程与应用科学学院）；School of Electrical and Computer Engineering at the University of Tehran（德黑兰大学电气与计算机工程学院）
Cooperative Edge Caching Based on Elastic Federated and Multi-Agent Deep Reinforcement Learning in Next-Generation Network
School of Internet of Things Engineering, Jiangnan University（江南大学物联网工程学院）；State Key Laboratory of Integrated Services Networks (Xidian University)（西安电子科技大学综合业务网理论及关键技术国家重点实验室）；Qualcomm（高通公司）；School of Engineering, University of Kent（肯特大学工程学院）；School of Engineering, Hong Kong University of Science and Technology（香港科技大学工程学院）
Two-Timescale Model Caching and Resource Allocation for Edge-Enabled AI-Generated Content Services
Xiamen University（厦门大学）；University of Hong Kong（香港大学）；Tsinghua University（清华大学）；University at Buffalo–SUNY（纽约州立大学布法罗分校）；Nanyang Technological University（南洋理工大学）；The Hong Kong University of Science and Technology（香港科技大学）
Applying Machine Learning Techniques for Caching in Edge Networks: A Comprehensive Survey
COMSATS University Islamabad（COMSATS大学伊斯兰堡校区）；Umm Al-Qura University（乌姆·阿尔-库拉大学）；College of Computer and Information Systems, Umm Al-Qura University（乌姆·阿尔-库拉大学计算机与信息系统学院）
PA-Cache: Evolving Learning-Based Popularity-Aware Content Caching in Edge Networks
Chongqing University（重庆大学）；Binghamton University, State University of New York（纽约州立大学宾汉姆顿大学）；Swinburne University of Technology（斯威本科技大学）；Harbin Institute of Technology（哈尔滨工业大学）
随着无处不在和个性化服务的蓬勃发展，大量移动设备在网络上产生了越来越多的流量。因此，内容缓存逐渐延伸到网络边缘，以提供低延迟服务，提高服务质量，并减少冗余数据流量。与传统的内容分发网络相比，边缘网络中具有较小尺寸的高速缓存通常必须容纳更多突发请求。在本文中，我们提出了一个不断发展的学习为基础的内容缓存策略，命名为PA缓存在边缘网络。它自适应地学习随时间变化的内容流行度，并确定当缓存已满时应该替换哪些内容。与传统的深度神经网络（DNN）不同，DNN使用具有高计算复杂性的整个训练数据集来学习微调但可能过时或有偏见的预测模型，PA-Cache对大量内容特征集进行加权，并在随着时间的推移有更多请求到达时从浅到深地训练多层递归神经网络。我们广泛评估我们提出的PA缓存的性能从一个大型的在线视频点播服务提供商在现实世界中的痕迹。\rb{The results show that PA-Cache outperforms existing popular caching algorithms and approximates the optimal algorithm with only a 3.8\% performance gap when the cache percentage is 1.0\%}。与传统的基于DNN的方法相比，PA-Cache还显着降低了计算成本。
Digital Twin-Enabled Mobility-Aware Cooperative Caching in Vehicular Edge Computing
Guangxi Normal University（广西师范大学）；Key Laboratory of Education Blockchain and Intelligent Technology, Ministry of Education（教育部教育区块链与智能技术重点实验室）；Guangxi Key Laboratory of Brain-inspired Computing and Intelligent Chips（广西脑启发计算与智能芯片重点实验室）；Guangxi Key Laboratory of Multi-Source Information Mining and Security（广西多源信息挖掘与安全重点实验室）；University Engineering Research Center of Educational Intelligent Technology（教育智能技术大学工程研究中心）；School of Computer Science and Engineering（计算机科学与工程学院）；PCA Lab（PCA实验室）；Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education（教育部高维信息智能感知与系统重点实验室）
Mobility-aware Caching Scheduling for Fog Computing in mmWave Band
作为云计算的延伸，网络边缘的雾计算提供了低延迟、位置感知和实时交互等特性。同时，毫米波（mmWave）通信能够提供具有大可用带宽的定向多千兆比特传输速率。基于用户在一个区域内的移动轨迹，可以获得用户经常经过的几个活动热点。通过在热点附近的边缘节点缓存热门内容，用户可以在短距离内直接下载缓存内容，用户体验得到显著提升。考虑到一个区域内存在多个热点，如何有效地调度传输以满足边缘节点的缓存需求成为一个关键问题。在本文中，我们专注于在热点附近的边缘节点的缓存的移动性感知的传输调度问题，并利用多跳中继和并发传输，以实现更好的性能。在将最优调度问题表示为随机非线性混合整数规划之后，我们提出了一种移动性感知的缓存调度方案，称为MHRC（Multi-Hop Relaying based Caching），其中为边缘节点建立多跳D2 D路径，并且在边缘节点处的缓存调度中利用并发传输。广泛的性能评估表明，MHRC实现了超过1倍的预期缓存数据量相比，国家的最先进的计划。
Fine-Grained AI Model Caching and Downloading With Coordinated Multipoint Broadcasting in Multi-Cell Edge Networks
State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, School of Electrical and Electronic Engineering, North China Electric Power University（替代电力系统与可再生能源电源国家重点实验室，电气与电子工程学院，华北电力大学）；China Satellite Network Group Shanghai Research Institute（中国卫星网络集团上海研究院）
