Bookstore

As an important enabler for changing people's lives, advances in artificial intelligence (AI)-based applications and services are on the rise, despite being hindered by efficiency and latency issues. By focusing on deep learning as the most representative technique of AI, this book provides a comprehensive overview of how AI services are being applied to the network edge near the data sources, and demonstrates how AI and edge computing can be mutually beneficial. To do so, it introduces and discusses: 1) edge intelligence and intelligent edge; and 2) their implementation methods and enabling technologies, namely AI training and inference in the customized edge computing framework. Gathering essential information previously scattered across the communication, networking, and AI areas, the book can help readers to understand the connections between key enabling technologies, e.g. a) AI applications in edge; b) AI inference in edge; c) AI training for edge; d) edge computing for AI; and e) using AI to optimize edge. After identifying these five aspects, which are essential for the fusion of edge computing and AI, it discusses current challenges and outlines future trends in achieving more pervasive and fine-grained intelligence with the aid of edge computing.

Edge AI: Convergence of Edge Computing and Artificial Intelligence

1: Introduction

2: Fundamentals of Edge Computing

        2.1 Paradigms of Edge Computing

                Cloudlet and Micro Data Centers

                Fog Computing

                Mobile (Multi-access) Edge Computing (MEC)

                Definition of Edge Computing Terminologies

                Collaborative End-Edge-Cloud Computing

        2.2 Hardware for Edge Computing

                AI Hardware for Edge Computing

                Integrated Commodities Potentially for Edge Nodes

                Edge Computing Frameworks

        2.3 Virtualizing the Edge

                Virtualization Techniques

                Network Virtualization

                Network Slicing

3. Fundamentals of Deep Learning

        3.1 Neural Networks in Deep Learning

                Fully Connected Neural Network (FCNN)

                Auto-Encoder (AE)

                Convolutional Neural Network (CNN)

                Generative Adversarial Network (GAN)

                Recurrent Neural Network (RNN)

                Transfer Learning (TL)

        3.2 Deep Reinforcement Learning (DRL)

                Value-based DRL

                Policy-gradient-based DRL

        3.3 Distributed DL Training

        3.4 Potential DL Libraries for Edge

4. Deep Learning Applications on Edge

                4.1 Real-time Video Analytic

                4.2 Autonomous Internet of Vehicles (IoVs)

                4.3 Intelligent Manufacturing

                4.4 Smart Home and City

5. Deep Learning Inference in Edge

                5.1 Optimization of DL Models in Edge

                                General Methods for Model Optimization

                                Model Optimization for Edge Devices

                5.2 Segmentation of DL Models

                5.3 Early Exit of Inference (EEoI)

                5.4 Sharing of DL Computation

6. Edge Computing for Deep Learning

                6.1 Edge Hardware for DL

                                Mobile CPUs and GPUs

        FPGA-based Solutions

6.2 Communication and Computation Modes for Edge DL

        Integral Offloading

        Partial Offloading

        Vertical Collaboration

        Horizontal Collaboration

6.3 Tailoring Edge Frameworks for DL

6.4 Performance Evaluation for Edge DL

7. Deep Learning Training at Edge

                7.1 Distributed Training at Edge

                7.2 Vanilla Federated Learning at Edge

                7.3 Communication-efficient FL

                7.4 Resource-optimized FL

                7.5 Security-enhanced FL

8. Deep Learning for Optimizing Edge

                8.1 DL for Adaptive Edge Caching

                8.2 DL for Optimizing Edge Task Offloading

                8.3 DL for Edge Management and Maintenance

                                Edge Communication

                                Edge Security

                                Joint Edge Optimization

9. Lessons Learned and Open Challenges

        9.1 More Promising Applications

        9.2 General DL Model for Inference

        9.3 Complete Edge Architecture for DL

        9.4 Practical Training Principles at Edge

        9.5 Deployment and Improvement of Intelligent Edge