Edge/Fog Computing Paradigm: The Concept, Platforms and Applications

Front Cover
Edge/Fog Computing Paradigm: The Concept, Platforms and Applications
Copyright
Contents
Contributors
Preface
Chapter One: Exploring the edge AI space: Industry use cases
1. The proliferation of IoT devices and sensors
2. Activating on-device intelligence
3. The artificial intelligence (AI) processing at the edge
4. Machine learning (ML) at the edge
5. Deep learning at the edge
6. Digging into the paradigm of edge AI
6.1. The blend of edge AI and 5G rekindles state-of-the-art applications
6.2. Person re-identification (Re-ID) (https://towardsdatascience.com/why-we-need-person-re-identification-3a45d170098b)
7. Edge AI for next-generation retail experiences
8. Edge AI for smarter cities
9. Edge AI for telecommunication
10. Conclusion
Further reading
Chapter Two: Edge computing: Types and attributes
1. Introduction
2. Internet of Things (IoT) edge
2.1. Device type and mobility
2.2. Service capabilities
2.3. Security and privacy
2.4. Latency requirements
3. On-premises edge
3.1. Disaggregation
3.2. Network requirements
3.3. Scalability
3.4. Security
3.5. Life cycle management
4. Wireless Access Edge
4.1. Scalability
4.2. Distributed RAN functions
4.3. Service models
4.4. Intelligent RAN
5. Network edge
5.1. Next Generation Central Office
5.2. Wireline fixed access edge
5.3. Physical locations
6. Challenges in edge computing
7. Multi-Access Edge Computing
References
Chapter Three: Industry initiatives across edge computing
1. Linux Foundation Edge
1.1. Akraino
1.2. EdgeXFoundary
1.3. EVE
1.4. Fledge
1.5. Home Edge
1.6. Open Horizon
1.7. State of the Edge
1.8. Baetyl
1.9. eKupier
1.10. Secure Device Onboard
2. Linux Foundation for Networking
2.1. Open Network Automation Project (ONAP)
2.2. Anuket
3. O-RAN alliance
4. Open Network Foundation
4.1. VOLTHA
4.2. SEBA
4.3. Aether
4.4. SD-RAN
4.5. SD-CORE
5. 3GPP
6. Small Cell Forum
7. Broadband Forum
7.1. Connected Home
7.2. 5G
8. 5G Alliance for connected industry and automation (5G-ACIA)
9. 5G Automotive Association (5GAA)
10. Automotive Edge Computing Consortium (AECC)
11. Telecom Infra Project
11.1. OpenRAN
11.2. Connected City Infrastructure
11.3. 5G Private Networks
12. IEEE International Network Generations Roadmap Edge Services Platform (ESP)
13. KubeEdge
14. StarlingX
15. Open Edge Computing Initiative
16. Smart Edge Open
17. Edge Multi Cluster Orchestrator (EMCO)
18. Global Systems for Mobile Association (GSMA)
References
Chapter Four: IoT-edge analytics for BACON-assisted multivariate health data anomalies
1. Introduction
2. Related works
3. System design
3.1. BACON algorithm for selection of multivariate outliers nomination
3.2. Initial basic subset of regression data algorithm
3.3. BACON robust regression algorithm
3.4. IoT-BACON-EEM algorithm
4. Results
4.1. Robust distance wise analysis
4.1.1. Robust distance
4.2. Robust linear regression analysis for HRV
4.2.1. Residual-fitted
4.2.2. Normal Q–Q
4.2.3. Scale location
4.2.4. Robust Mahalanobis distance
4.3. Coefficients
5. Conclusion
References
Chapter Five: The edge AI paradigm: Technologies, platforms and use cases
1. Introduction
2. Delineating the two paradigms
3. Tending toward the digital era
4. The key connectivity technologies
5. The 5G use cases and benefits
6. About edge computing
7. Edge computing architecture
8. Edge cloud infrastructures
9. Edge analytics
10. The key benefits of edge computing
11. Tending toward edge AI
12. Artificial intelligence (AI) chips for edge devices
13. The noteworthy trends toward edge AI
14. Why edge processing?
15. Edge-based AI solutions: The advantages
16. Applications that can be performed on edge devices
17. Edge AI use cases
18. Conclusion
Chapter Six: Microservices architecture for edge computing environments
1. Introduction
2. Need for edge and fog computing
3. Nature and requirements in edge and fog computing environment
3.1. Evolving or changing needs of the IoT
3.2. Heterogeneous nature of the IoT layer
3.3. Mobility and low power
3.4. Distributed nature of the layer
3.5. Low computing/processing power
4. Why microservices architecture for edge/fog computing applications?
5. How the unique features of MSA fits as a natural choice for edge and fog layers?
6. Overview about elements of microservices
6.1. Design principle of microservices
6.2. Communication protocols of microservices
6.3. Application programing interface and microservices
6.4. Microservice registration and discovery
6.5. API gateway
6.6. Polyglot support for development
6.7. MSA and transaction support
6.8. MSA and design patterns
6.9. MSA and security
7. MSA for edge/fog computing
7.1. Fog computing environment
8. Challenges
9. Conclusion
References
Chapter Seven: Edge data analytics technologies and tools
1. Introduction to edge data analytics and benefits
1.1. Benefits of edge data analytics
2. Edge data analytics versus server-based data analytics
3. Architecture and methodology of edge data analytics
4. Edge data analytics technologies and solutions
4.1. An extended AWS to the edge devices for function with data generated while using the cloud for necessary resources i ...
4.1.1. Benefits
4.1.2. Working of AWS IoT Greengrass
4.1.3. Scenarios
4.2. CSA-A network assessment tool by Cisco named Cisco SmartAdvisor (Cisco discovery service) (Fig. 4)
4.2.1. Features and benefits
4.2.2. Working of CDS (Fig. 5)
4.2.3. Requirements for the application of CSA
4.3. An analytics software package named Dell Statistica (Fig. 6)
4.4. An edge system providing high performance, low latency data processing by Hewlett Packard enterprise named HPE edgel ...
4.4.1. Significance of converging OT and IT: Connecting workers with data and insights at the edge
4.4.2. Benefits
4.4.3. Use cases
4.5. An edge analytics agent by IBM named IBM Watson IoT edge analytics (Fig. 8)
4.6. An IoT hub to apply analytics at edge devices named microsoft azure IoT edge (Fig. 9)
4.6.1. Benefits
4.7. A solution by Oracle for event processing is named Oracle edge analytics (OEA) (Fig. 10)
4.7.1. Benefits
4.7.2. Use cases
4.8. A solution to handle complex analytical processes is named PTC ThingWorx analytics (Fig. 11)
4.8.1. Benefits
4.9. Components and use cases
4.10. A to-the-edge component to collect, process, and transmit data is named streaming lite by SAP HANA (Fig. 12)
4.10.1. Use cases and benefits
5. Working principles and feature comparisons
6. Some of the other use cases of edge analytics [19,20]
References
Chapter Eight: Edge platforms, frameworks and applications
1. Introduction to cloud computing
2. Cloud computing to edge computing
3. Edge computing: A brief overview
4. Essential of edge computing
5. Advantages of edge computing
5.1. Latency reduction
5.2. Safer data processing
5.3. Inexpensive scalability
5.4. Simple expansions to new markets
5.5. Consistent user experience
5.6. Speed
5.7. Edge computing technologies
5.8. Cloudlets: An overview
6. Significance of cloudlets
6.1. MEC benefits
6.2. FOG computing
6.3. Benefits
6.3.1. Confidentiality
6.3.2. Efficiency
6.3.3. Safety and security
6.3.4. Bandwidth
6.3.5. Latency
6.4. Edge computing applications
6.4.1. Smart systems
6.4.2. Video streaming
6.4.3. Remote monitoring and predictive analysis
6.4.4. Gaming-as-a-service
6.4.5. 5G communications
6.4.6. 5G smart health care
6.4.7. Security monitoring
6.4.8. AR and VR
6.5. Edge computing and future
6.6. Shortcomings of edge computing
7. Conclusion
References
Chapter Nine: Edge computing challenges and concerns
1. Introduction
2. Cloud, fog and edge computing
2.1. Cloud computing
2.2. Fog computing
2.3. Edge computing
3. Implications and challenges in adopting edge computing
3.1. Accessibility
3.2. Control and management
3.3. Scalability
3.4. Privacy and security
3.5. Data storage
3.6. Latency
3.7. Performance
4. Concerns with edge computing
4.1. Cost
4.2. Data
4.3. Response time requirement
4.4. Security concerns
4.5. Security aspects
5. Security and privacy attacks on edge computing enabled devices
5.1. Physical attacks
5.2. Sniffing
5.3. Unauthorized access
5.4. Routing table attack
5.5. Distributed denial of service attack (DDoS)
5.6. Malicious hardware and software injections
5.7. Integrity attack
5.8. Privacy leakage
5.9. Logging attacks
5.10. Data storage and protection
6. Countermeasures to security and privacy attacks in edge infrastructure
6.1. Solution to physical attack
6.2. Solution to sniffing
6.3. Solution to unauthorized access
6.4. Solution to routing table attack
6.5. Solution to distributed denial of service attack (DDoS)
6.6. Solution to malicious hardware and software injections
6.7. Solution to integrity attack
6.8. Solution to privacy leakage
6.9. Solution to logging attacks
6.10. Solution to data storage and protection
6.11. Embedding blockchain on edge infrastructure
7. Future of edge computing
8. Conclusion
References
Further reading
Chapter Ten: A smart framework through the Internet of Things and machine learning for precision agriculture
1. Introduction
2. Existing infrastructure in agriculture
2.1. Limitations in current agriculture infrastructure
2.1.1. Social limitations
2.1.2. Environmental limitations
2.1.3. Technical limitations
2.1.4. Inherent limitation
2.1.5. Organizational limitation
2.1.5.1. We still get food and the necessary things to make our living. Does that mean the agricultural sector is functio ...
2.1.5.2. What could be the cause?
2.2. Intelligent workforce infrastructure
3. IoT ecosystem—A complete view
3.1. IoT devices
3.2. Communication technology
3.3. Data storage and processing
3.4. Knowledge layer
3.5. Service layer
4. Agricultural monitoring system based on sensors
4.1. Field assessment
4.2. Cattle behavior control
4.3. Traditional agricultural monitoring
4.4. Intelligent IoT based irrigation system
5. Difficulties in sensor-based agribusiness observing frameworks
5.1. Cost
5.2. Reliability
5.3. Resources
6. Factors affecting climatic changes in savvy agribusiness
6.1. Climate-brilliant horticulture
6.2. Key highlights of atmosphere shrewd rural scenes
6.3. Climate-keen practices at field and homestead scale
6.4. Diversity of land use over the scene
6.4.1. Reduce hazard
6.4.2. Provide key nourishment and feed saves
6.4.3. Sustaining seasonal forest as a carbon resource
6.4.4. Effective functions of the ecosystem
6.4.5. Improve the advantages of environmentally smart policies on the ground
7. AI in agriculture—An introduction
8. Machine learning techniques for smart agriculture
8.1. Wide division of machine learning algorithms
9. Artificial neural network (ANN)
9.1. Artificial neural networks in agriculture
10. Automation and wireless system networks in agriculture
10.1. Smart agriculture
10.2. Smart farming (SF)
10.3. Smart IoT agricultural revolution
10.4. Smart system design methodologies
11. Hardware components in the smart agriculture system
11.1. Humidity sensor
11.2. CO2 sensor
11.3. Moisture sensor
11.4. Rain drop sensor
11.5. Ultrasonic sensor
11.5.1. Implementation using raspberry PI 2 model B
12. Use cases
12.1. Solar fertigation: Internet of Things architecture for smart agriculture
12.2. Wireless sensor based crop monitoring system for agriculture using Wi-Fi network dissertation
12.3. Secure smart agriculture monitoring technique through isolation
12.4. Realizing social-media-based analytics for smart agriculture
13. Conclusion
References
Chapter Eleven: 5G Communication for edge computing
1. Introduction
2. Architectures of edge computing
2.1. Edge computing reference frame
2.2. The architecture of edge computing
3. 5G and edge computing
3.1. Importance of edge computing
3.2. Taxonomy of edge computing
3.2.1. Goals
3.2.2. Computational platforms
3.2.3. Characteristics
3.2.4. 5G Functions
3.2.5. The functioning of edge computing in 5G
3.2.6. Integration of 5G and edge computing
3.2.7. Security and privacy
4. 5G and edge computing use cases
4.1. Industry 4.0
4.2. 5G Communication technology in Industry 4.0
5. Challenges during the deployment of edge computing in 5G
6. Conclusion
References
Chapter Twelve: The future of edge computing
1. Introduction
2. Emergence of edge computing
3. Drawbacks of out-of-date cloud computing
4. Significance of edge computing
5. Edge computing technologies
5.1. Fog
5.2. Cloudlets
5.3. Mobile edge computing
6. Possible advancements in digitization using edge computing
6.1. Edge computing in network architecture
6.2. Remote monitoring
6.3. Healthcare
6.4. E-Commerce
6.5. Markets/business
6.6. Security
6.7. 5G Communications
6.8. Smartphone advancements
7. Opportunities for edge in future
7.1. Multimedia and edge computing
7.2. Energy efficiency and edge
7.3. Smart living
7.4. Communication efficiency
7.5. Resource management
7.6. Environment monitoring
8. Conclusion
References
Chapter Thirteen: Edge computing security: Layered classification of attacks and possible countermeasures
1. Introduction
2. Four layer architecture of edge computing
3. Security attacks in edge computing: Layered classification and analysis
3.1. Physical layer
3.2. Data link layer
3.3. Network layer
3.4. Transport layer
3.5. Application layer
4. Edge based existing solutions for the security issues present in real world IoT applications
4.1. Smart city
4.2. Industrial environment
4.3. Smart campus
4.4. Vehicular network
4.5. Healthcare system
5. Discussion
5.1. Analysis of existing defense mechanisms
5.2. Open research challenges
5.2.1. Rapid increase in the number of network components
5.2.2. Heterogeneous nature
5.2.3. Possibilities of identity based attacks
6. Conclusion and future works
References
Chapter Fourteen: Blockchain technology for IoT edge devices and data security
1. Introduction
1.1. What is IoT?
1.2. Basics of edge devices
2. IoT layered architecture
2.1. The perception layer
2.2. The network layer
2.3. The application layer
2.4. The service support layer
3. IoT security threats and attacks
3.1. Classification of attacks based on IoT-architecture
3.2. Attacks—sensing (or) perception layer
3.2.1. Node attack
3.2.2. Sinkhole attack
3.2.3. Selective-forwarding attack
3.2.4. Witch attack
3.2.5. HELLO flood attacks
3.2.6. Physical damage
3.3. Attacks—Network and service support layers
3.3.1. Man-in-the-middle (MITM) attack
3.3.2. Replay attack
3.3.3. Denial of service attack
3.3.4. Sybil attack
3.3.5. Sinkhole attack
3.3.6. Sniffing attack
3.4. Attacks—Middle-ware layer (or) service support layer
3.4.1. Flooding attack in cloud
3.4.2. Malware injection
3.4.3. Signature wrapping attack
3.4.4. Web browser attack
3.4.5. SQL injection attack
3.5. Attacks—Application layer
3.5.1. Code injection
3.5.2. Buffer overflow
3.5.3. Sensitive data permission and manipulation
3.5.4. Phishing attack
3.5.5. Authentication and authorization
4. IoT—Edge computing
4.1. Functions
4.2. Three-Tier edge computing model
4.3. Edge vs cloud
4.4. Attacks on edge nodes
4.4.1. Network attack
4.4.2. Port attack
4.4.3. Side-channel attack
4.4.4. Physical attack
5. Requirements for integration of blockchain and edge computing
5.1. Authentication
5.2. Adaptability
5.3. Data integrity
5.4. Verifiable computation
5.5. Low latency
5.6. Network security
6. Integration of blockchain and edge computing
6.1. Blockchain role in edge computing
6.2. Mixing—Blockchain and edge computing
6.2.1. Edge computing—Inadequate security
6.2.2. Challenges and restrictions of blockchain
7. IoT framework: Secure edge computing with blockchain technology
7.1. Design overview
7.2. Blockchain framework layered architecture
7.3. Distributed-IoT device layer
7.4. Point-to-point edge servers network
7.5. Decentralized resources of cloud
8. Factors to be addressed in secure edge computing
8.1. Low latency
8.2. Longer battery life for IoT devices
8.3. More efficient data management
8.4. Access to data analytics and AI
8.5. Resilience
8.6. Scalability
9. Advantages—Integration of blockchain and edge computing
10. Use cases—Blockchain with edge computing
10.1. Smart city
10.2. Smart transportation
10.3. Industrial IoTs
10.4. Smart home
10.5. Smart grid
11. Further challenges and recommendations
11.1. Technical threats
11.2. Interoperability and standardization
11.3. Blockchain framework
11.4. Administration, authority, controlling and legal characteristics
11.5. Rapid field testing
12. Conclusion
References
Chapter Fifteen: EDGE/FOG computing paradigm: Concept, platforms and toolchains
1. Introduction
2. Machine learning (ML) in FC
3. Classes of service for fog applications
4. Clusters for lightweight edge clouds
4.1. FAP´s machine learning algorithms
4.2. Machine learning for the protection of security and privacy
5. IoT Application with fog real time application
6. Safeguarding data consistency at the edge
6.1. Data embedding on computing device with IoT on fog environment
7. Cloud-fog-edge-IoT collaborative framework
8. Edge computing with machine learning
8.1. Resource management in fog computing
9. Security challenges in fog computing
10. Conclusion
Reference
Chapter Sixteen: Artificial intelligence in edge devices
1. Introduction
2. Primer on artificial intelligence
2.1. Artificial intelligence
2.2. Deep learning and deep neural networks
2.3. From deep learning to model training and inference
2.4. Popular deep learning models
2.4.1. Convolutional neural networks
2.4.2. Recurrent neural networks
2.4.3. Generative adversarial networks
2.4.4. Deep reinforcement learning
3. Edge intelligence
3.1. Motivation and benefits of edge intelligence
4. Edge intelligence model training
4.1. Architectures
4.2. Key performance indicators
4.3. Enabling technologies
4.4. Summary of the existing systems and frameworks
5. Edge intelligence model interface
5.1. Architectures
5.2. Key performance indicators (KPIs)
5.3. Enabling technologies
5.4. Summary of the existing systems and frameworks
6. Future research directions
6.1. Programming and software platforms
6.2. Resource-friendly edge AI model design
6.3. Computation-aware networking techniques
6.4. Trade-off design with various DNN performance metrics
6.5. Resource management and smart services
6.6. Security and privacy issues
6.7. The EI ecosystem is a large open collaboration that focuses on incentives and business models
7. Conclusions
References
Further reading
Chapter Seventeen: 5G—Communication in HealthCare applications
1. Introduction
2. 5G—IOT for E-healthcare
3. 5G—Industrial Internet of Thongs (IIoT)
4. 5G—Network requirements for healthcare
5. 5G—Virtual HealthCare
6. TeleHealth vs. virtual health
7. 5G—Remote HealthCare monitoring
8. 5G—Remote surgery
9. 5G—Futures and robotics in healthcare
10. 5G—Impact on HealthCare
11. Conclusion
References
Chapter Eighteen: The integration of blockchain and IoT edge devices for smart agriculture: Challenges and use cases
1. Introduction
1.1. Requirement of IoT in agriculture
1.2. Requirement of blockchain in agriculture
2. Blockchain technology: An overview
3. Working of blockchain
4. IoT: An overview
5. Working of IoT
6. Edge computing: An overview
7. A proposed model for smart agriculture using blockchain and IoT
8. Advantages of blockchain, edge computing and IoT based agriculture
9. Summary of the research for applying blockchain and IoT in agriculture (Table 2)
10. Challenges and open issues
11. Conclusion
References
Back Cover