Skip to main content
main-content

Über dieses Buch

This book mainly discusses the most important issues in artificial intelligence-aided future networks, such as applying different ML approaches to investigate solutions to intelligently monitor, control and optimize networking. The authors focus on four scenarios of successfully applying machine learning in network space. It also discusses the main challenge of network traffic intelligent awareness and introduces several machine learning-based traffic awareness algorithms, such as traffic classification, anomaly traffic identification and traffic prediction. The authors introduce some ML approaches like reinforcement learning to deal with network control problem in this book.

Traditional works on the control plane largely rely on a manual process in configuring forwarding, which cannot be employed for today's network conditions. To address this issue, several artificial intelligence approaches for self-learning control strategies are introduced. In addition, resource management problems are ubiquitous in the networking field, such as job scheduling, bitrate adaptation in video streaming and virtual machine placement in cloud computing. Compared with the traditional with-box approach, the authors present some ML methods to solve the complexity network resource allocation problems. Finally, semantic comprehension function is introduced to the network to understand the high-level business intent in this book.

With Software-Defined Networking (SDN), Network Function Virtualization (NFV), 5th Generation Wireless Systems (5G) development, the global network is undergoing profound restructuring and transformation. However, with the improvement of the flexibility and scalability of the networks, as well as the ever-increasing complexity of networks, makes effective monitoring, overall control, and optimization of the network extremely difficult. Recently, adding intelligence to the control plane through AI&ML become a trend and a direction of network development

This book's expected audience includes professors, researchers, scientists, practitioners, engineers, industry managers, and government research workers, who work in the fields of intelligent network. Advanced-level students studying computer science and electrical engineering will also find this book useful as a secondary textbook.

Inhaltsverzeichnis

Frontmatter

Chapter 1. Introduction

Abstract
The current Internet architecture established from TCP/IP has gained huge success and been one of the indispensable infrastructures for our daily life, economic operation and society. However, burgeoning megatrends in the information and communication technology (ICT) domain are urging the Internet for pervasive accessibility, broadband connection and flexible management, which call for potential new Internet architectures. The original design tactic of the Internet, which is “Leaving the complexity to hosts while maintaining the simplicity of network”, leads to the almost insurmountable challenge known as “Internet ossification”: software in the application layer has developed rapidly, and abilities in the application layer have been drastically enriched. By contrast, protocols in the network layer lack scalability and the core architecture is hard to modify, which means that new functions have to be implemented through myopic and clumsy ad hoc patches in the existing architecture. For example, the transition from IPv4 to IPv6 is difficult to deploy in practice.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 2. Intelligence-Driven Networking Architecture

Abstract
Network architecture provides a full picture of the established network with a detailed view of all the resources accessible. The architecture of software defined network facilitates separation of the control plane from the forwarding plane. However, the work on the control plane largely relies on a manual process in configuring forwarding strategies. To address this issue, in this chapter, we propose NetworkAI, a new network architecture exploiting software-defined networking, network monitor technologies and reinforcement learning technologies for controlling networks in an intelligent way. NetworkAI implements a network state upload link and a decision download link to accomplish a closed-loop control of the network and builds a centralized intelligent agent aiming at learning the policy by interacting with the whole network.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 3. Intelligent Network Awareness

Abstract
In the network, different applications produce various traffic types with diverse features and service requirements. Therefore, in order to better manage and control networking, the intelligent awareness of network traffic plays a significant role. Network traffic information mainly includes service-level information (e.g., QoS/QoE), anomaly traffic detection information, etc. In this chapter, we first present a multi-level intrusion detection model framework named MSML to address these issues. The MSML framework includes four modules: pure cluster extraction, pattern discovery, fine-grained classification and model updating. Then, we propose a novel IDS framework called HMLD to address these issues, which is an exquisitely designed framework based on Hybrid Multi-Level Data Mining. In addition, we propose a new model based on big data analysis, which can avoid the influence brought by adjustment of network traffic distribution, increase detection accuracy and reduce the false negative rate. Finally, we propose an end-to-end IoT traffic classification method relying on deep learning aided capsule network for the sake of forming an efficient classification mechanism that integrates feature extraction, feature selection and classification model. Our proposed traffic classification method beneficially eliminates the process of manually selecting traffic features, and is particularly applicable to smart city scenarios.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 4. Intelligent Network Control

Abstract
Finding the near-optimal control strategy is the most critical and ubiquitous problem in a network. Examples include routing decision, load balancing, QoS-enable load scheduling, and so on. However, the majority solutions of these problems are largely relying on a manual process. To address this issue, in this chapter, we apply several artificial intelligence approaches for self-learning control strategies in networks. In this chapter, we first present an energy-aware multi-controller placement scheme as well as a latency-aware resource management model for the SDWN. Moreover, the particle swarm optimization (PSO) is invoked for solving the multi-controller placement problem, and a deep reinforcement learning (DRL) algorithm aided resource allocation strategy is conceived. Then, we present a novel controller mind (CM) framework to implement automatic management among multiple controllers and propose a novel Quality of Service (QoS) enabled load scheduling algorithm based on reinforcement learning to solve the problem of complexity and pre-strategy in the networks. In addition, we present a Wireless Local Area Networks (WLAN) interference self-optimization method based on a Self-Organizing Feature Map (SOM) neural network model to suppress the interference in local area networks. Finally, we propose a BC-based consensus protocol in distributed SDIIoT, where BC works as a trusted third party to collect and synchronize network-wide views between different SDN controllers. In addition, we use a novel dueling deep Q-learning approach to solve this joint problem.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 5. Intelligent Network Resource Management

Abstract
Resource management problems are ubiquitous in the networking field, such as job scheduling, bitrate adaptation in video streaming and virtual machine placement in cloud computing. In this chapter, we propose a reinforcement learning based dynamic attribute matrix representation (RDAM) algorithm for virtual network embedding. The RDAM algorithm decomposes the process of node mapping into the following three steps: (1) static representation of substrate physical network. (2) dynamic update of substrate physical network. (3) Reinforcement-Learning-Based algorithm. Then, we design and implement a policy network based on reinforcement learning to make node mapping decisions. We use policy gradient to achieve optimization automatically by training the policy network with the historical data based on virtual network requests.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 6. Intention Based Networking Management

Abstract
Compared to traditional networking which using command-line interfaces, intent-based networking abstracts network complexity and improves automation by eliminating manual configurations. It allows a user or administrator to send a simple request—using natural language—to plan, design and implement/operate the physical network which can improve network availability and agility. For example, an IT administrator can request improved voice quality for its voice-over-IP application, and the network can respond. For intent-based networking, the translation and validation system take a higher-level business policy (what) as input from end users and converts it to the necessary network configuration (how) by natural language understanding technology. In this chapter, we focus on how artificial intelligence technology can be used in the natural language understanding in translation and validation system. We firstly propose an effective model for the similarity metrics of English sentences. In the model, we first make use of word embedding and convolutional neural network (CNN) to produce a sentence vector and then leverage the information of the sentence vector pair to calculate the score of sentence similarity. Then, we propose the SM-CHI feature selection method based on the common method used in Chinese text classification. Besides, the improved CHI formula and synonym merging are used to select feature words so that the accuracy of classification can be improved and the feature dimension can be reduced. Finally, we present a novel approach which considers both the semantic and statistical information to improve the accuracy of text classification. The proposed approach computes semantic information based on HowNet and statistical information based on a kernel function with class-based weighting.
Haipeng Yao, Chunxiao Jiang, Yi Qian

Chapter 7. Conclusions and Future Challenges

Abstract
This book mainly discusses the architectures of intelligent networks as well as the possible techniques and challenges. We first introduced the concept of NetworkAI, which is a novel paradigm that applying machine learning to automatically control a network. NetworkAI employs reinforcement learning and incorporates network monitoring technologies such as the in-band network telemetry to dynamically generate control policies and produces a near-optimal decision. We employ the SDN and INT to implement a network state upload link and a decision download link to accomplish a closed-loop control of a network and build a centralized intelligent agent aiming at learning the policy by interaction with a whole network.
Then, we discuss the possible machine learning methods for network awareness. With the rapid development of compelling application scenarios of the networks, such as 4K/8K, IoT, it becomes substantially important to strengthen the management of data traffic in networks. As a critical part of massive data analysis, traffic awareness plays an important role in ensuring network security and defending traffic attacks. Moreover, the classification of different traffic can help to improve their work efficiency and quality of service (QoS).
Haipeng Yao, Chunxiao Jiang, Yi Qian
Weitere Informationen