Deep Learning for Intelligent Wireless Networks: A Comprehensive Survey
基于智能无线网络的深度学习:全面调查
摘要
As a promising machine learning tool to handle the accurate pattern recognition from complex raw data, deep learning (DL) is becoming a powerful method to add intelligence to wireless networks with large-scale topology and complex radio conditions.
作为一种很有前途的机器学习工具,机器学习可以从复杂的原始数据中处理精确的模式识别,机器学习成为一种强有力的方法来为具有大规模拓扑结构和复杂无线电条件的无线网络添加人工智能。
- wireless networks 无线网络
DL uses many neural network layers to achieve a brain-like acute feature extraction from high-dimensional raw data.
深度学习使用许多神经网络层来实现从高维原始数据中提取类似大脑的确切特征。
It can be used to find the network dynamics (such as hotspots, interference distribution, congestion points, traffic bottlenecks, spectrum availability, etc.) based on the analysis of a large amount of network parameters (such as delay, loss rate, link signal-to-noise ratio, etc.).
基于分析大量的网络参数,例如如延时、损耗率、链路信噪比等,可以发现网络的动态特性,例如例如热点、干扰分布、拥塞点、交通单元、频谱可用性等。
- network parameters 网络参数
- dynamics 动态特性
Therefore, DL can analyze extremely complex wireless networks with many nodes and dynamic link quality.
因此,深度学习可以分析多节点,动态链接质量及其复杂的无线网络。
This paper performs a comprehensive survey of the applications of DL algorithms for different network layers, including physical layer modulation/coding, data link layer access control/resource allocation, and routing layer path search, and traffic balancing.
本文对不同网络层的深度学习算法做了全面的研究,包括物理层调制编码,数据链路层访问控制资源分配,路由层路径搜索和流量均衡。
The use of DL to enhance other network functions, such as network security, sensing data compression, etc., is also discussed.
文中还讨论了如何用深度学习增加其他网络功能,例如网络安全,传感数据压缩等。
Moreover, the challenging unsolved research issues in this field are discussed in detail, which represent the future research trends of DL-based wireless networks.
更多的是对这个领域尚未解决的具有挑战性的问题也被详细讨论了,这代表了基于深度学习无线网络的未来研究趋势。
This paper can help the readers to deeply understand the state-of-the-art of the DL-based wireless network designs, and select interesting unsolved issues to pursue in their research.
这篇论文帮助阅读者深入理解目前基于深度学习神经网络的现状,并选择自己感兴趣的未解决问题进行研究。
关键词
Wireless networks, deep learning (DL), deep reinforcement learning (DRL), protocol layers, performance optimization.
无线网络,深度学习,深度强化学习,协议层,性能优化
第一章 介绍
P1
Human…etc.
P2
DL is a subclass of machine learning which uses cascaded layers to extract features from the input data and eventually forms a decision.
深度学习是机器学习的一个子类,他使用层叠层从输入数据中提取特征,并形成最终决策。
- cascaded layers 层叠层
The application of DL should consider four aspects:
(1) How to represent the state of the environment in suitable numerical formats, which will be taken as the input layer of the DL network;
(2) How to represent/interpret the recognition results, i.e., the physical meaning of the output layer of the DL network;
(3) How to compute/update the reward value, and what is the proper reward function that can guide the iterative weight updating in each neural layer;
(4) The structure of the DL system, including how many hidden layers, the structure of each layer, and the connections between layers.
深度学习的应用需要考虑4个方面:
(1)如何用合适的数值格式表示环境的状态,将其作为深度学习神经网络的输入层;
(2)如何表示和解释识别结果,即DL网络输出层的物理意义;
(3)如何计算/更新奖励值,以及在每个神经层中,什么样的迭代函数可以指导迭代权值的更新;
(4)DL的系统结构,包括隐藏层的数量,每一层的结构,以及层之间的联系。
P3
**Currently, many DL systems are tied with Reinforcement Learning (RL) models , which comprises three parts: **
1) an environment which can be described by some features,
**2) an agent which takes actions to change the environment, **
and 3) an interpreter which announces the current state and the action the agent takes.
Meanwhile, the interpreter announces the reward after the action takes effect in the environment, as shown in Fig. 1. The goal of the RL is to train the agent in such a way that for a given environment state, it chooses the optimal action that yields the highest reward. Therefore, one of the main differences between DL and RL is that the former usually learns from examples (e.g., training data) to create a model to classify data, however, the latter trains the model by maximizing the reward associated with different actions.
目前,许多DL系统都与强化学习模型绑定,它目前包括三个部分:
1)一个可以用一些特征描述的环境,
2)一个代理采取行动来改变环境,
3)一个解释器,宣布当前状态和代理采取的行动。
同时,当动作在环境中生效后,解释器会公布奖励信息,如图1所示。RL的目标是训练agent在给定的环境状态下选择最高回报的最优行动。因此,DL和RL的主要区别是,深度学习通常从例子中创造一个模型来分类数据,而强化学习通过最大化不同行动的回报来训练模型。
P4
DL has already shown astonishing capabilities in dealing with many real-world scenarios, such as the success of Alpha Go, the face recognition on mobile phones, etc.
Researchers in computer network areas also cast strong interests in DL applications. By using DL model the complex network environment can be represented, abstract features can be obtained, and a better decision can be achieved finally for the computer network nodes to achieve improved network quality-of-service (QoS) and quality-of-experience (QoE).
深度学习在处理真实场景时已经展现出惊人的能力,例如Alpha Go的成功以及手机上的人脸识别等。计算机网络领域的研究人员也对DL应用产生了浓厚的兴趣。使用DL模型可以表示复杂的网络环境,获得抽象特征,最终为计算机网络结点做出更好的决策,以实现改进的网络质量服务和体验质量。
P5
Wireless networks yield complex features, such as communication signal characteristics, channel quality, queueing state of each node, path congestion situation, etc.
On the other hand, there are many network control targets having significant impacts on the communication performances, such as resource allocation, queue management, congestion control, etc.
To handle the complicated situations, machine learning technique has been extensively explored . Chen et al. presented a comprehensive summary towards the ML applications in wireless networks, including wireless communications and networking using Unmanned Aerial Vehicles (UAVs), wireless virtual reality, mobile edge caching and computing, spec- trum management and co-existence of multiple radio access, Internet of Things, etc. The applications of ML in these areas present astonishi