Mobile Microservices: Building Flexible Pervasive Applications

Cover
Half Title
Title Page
Copyright Page
Dedication
Contents
Preface
Contributors
Chapter 1: Introduction
	1.1. BOOK STRUCTURE
Chapter 2: Design Concepts for Pervasive Applications
	2.1. MOTIVATING SCENARIO: A SMART PUBLIC SPACE
		2.1.1. Challenges
		2.1.2. Possible Solutions
	2.2. LOCATING A PROVIDER
		2.2.1. Reactive Discovery
		2.2.2. Proactive Discovery
		2.2.3. Concept 1: Planning-Based Composition Announcement
	2.3. REQUEST ROUTING
		2.3.1. Controlled Flooding
		2.3.2. Directory Based
		2.3.3. Overlay Based
		2.3.4. Concept 2: Dynamic Controlled Flooding
	2.4. COMPOSITION PLANNING
		2.4.1. Open Service Discovery
		2.4.2. Goal-Oriented Planning
		2.4.3. Concept 3: Decentralised Flexible Backward Planning
	2.5. SERVICE BINDING
		2.5.1. QoS-Based Selection
		2.5.2. Adaptable Binding
		2.5.3. On-Demand Binding
		2.5.4. Concept 4: Path Reliability-Driven Selection
		2.5.5. Concept 5: Bind Microservices On-Demand
	2.6. SERVICE INVOCATION
		2.6.1. Fragments Distribution
		2.6.2. Process Migration Approaches
		2.6.3. Concept 6: Runtime Service Announcement
	2.7. FAULT TOLERANCE
		2.7.1. Preventive Adaptation
		2.7.2. Composition Recovery
		2.7.3. Concept 7: Local Execution Path Maintenance
	2.8. CHAPTER SUMMARY
Chapter 3: Microservices Deployment in Edge/Fog Computing Environments
	3.1. EDGE COMPUTING: PERVASIVE APPLICATION’S NEW
		3.1.1. Edge Computing and Fog Computing
	3.2. FEATURES IN EDGE ENVIRONMENTS
		3.2.1. Latency-Sensitive
		3.2.2. Mobility is Everywhere
		3.2.3. Openness of Network Systems
		3.2.4. Constantly Changing Environment
		3.2.5. Limited Power Supply
	3.3. FOG-AS-A-SERVICE MODEL
	3.4. EDGE AND FOG COMPUTING ARCHITECTURE
	3.5. FOG NODE OVERLAY NETWORK
	3.6. HIERARCHICAL MICROSERVICES MANAGEMENT
		3.6.1. Fog Services and Service Composition
		3.6.2. Proxy Fog Nodes
		3.6.3. Seamless Service Invocation
	3.7. ADAPTABILITY AT EDGE
		3.7.1. Monitoring Environmental Changes
		3.7.2. Adaptation Analysis Based on Deep Learning
		3.7.3. Adaptation Planning Based on Reinforcement Learning
		3.7.4. Strategy Execution and Knowledge Base Utilisation
		3.7.5. Extension of MAPE-K Framework
	3.8. MICROSERVICES DEPLOYMENT AND DYNAMIC REDEPLOYMENT
	3.9. EXAMPLES OF PERVASIVE APPLICATIONS AT EDGE
		3.9.1. Mobile Video
		3.9.2. Smart Home
		3.9.3. Computational Offloading
	3.10. OPEN ISSUES TO EDGE-ENABLED PERVASIVE APPLICATIONS
		3.10.1. End-to-End Security
		3.10.2. Distributed Run-Time Management
		3.10.3. Scalability and Reconfigurability
		3.10.4. Predictive Fault-Tolerant
		3.10.5. Intelligent Edge Computing for 6G
	3.11. CHAPTER SUMMARY
Chapter 4: Microservices Composition Model
	4.1. MICROSERVICE MODEL
	4.2. DYNAMIC GOAL-DRIVEN COMPOSITION PLANNING
		4.2.1. Local Service Planning
		4.2.2. Complex Service Flows
	4.3. HEURISTIC SERVICE DISCOVERY
	4.4. EXECUTION FRAGMENTS SELECTION AND INVOCATION
		4.4.1. Microservice Composite Selection and Invocation
		4.4.2. Service Execution and Guidepost Adaptation
	4.5. DISCUSSION
		4.5.1. Quantitative Analysis
		4.5.2. Service Flows
		4.5.3. Privacy and Security
		4.5.4. Semantic Matchmaking
		4.5.5. High Composition Demand
	4.6. CHAPTER SUMMARY
Chapter 5: Cooperative Microservices Provisioning
	5.1. COOPERATIVE CACHING AND SELFISH CACHING
		5.1.1. Social Behaviours in Caching
		5.1.2. Social Selfishness of Service Providers
	5.2. LOCAL OPTIMAL CACHING ALGORITHM WITH SOCIAL SELFISHNESS
	5.3. COOPERATIVE DEVICES
	5.4. SOCIAL SELFISHNESS-BASED UTILITY
		5.4.1. Access Admission Mechanism
		5.4.2. Social Group Utility Mechanism
	5.5. SERVICE DEPLOYMENT AND PROVISIONING GAME
	5.6. OPTIMAL LOCAL SERVICE DEPLOYMENT
		5.6.1. Algorithm for the Caching Model
	5.7. CHAPTER SUMMARY
Chapter 6: Implementation I: Service Middleware
	6.1. SERVICE COMPOSITION ARCHITECTURE
	6.2. CLIENT AND PROVIDER
		6.2.1. Client Engine
		6.2.2. Microservices Provider
	6.3. ROUTING CONTROLLER
	6.4. GUIDEPOST MANAGER
		6.4.1. Adapting a Guidepost
		6.4.2. Guidepost Data in Service Execution
	6.5. MESSAGE HELPER
	6.6. PROTOTYPES
		6.6.1. Prototype on Android
		6.6.2. Prototype on Ns-3
	6.7. IMPLEMENTATION SUMMARY
Chapter 7: Implementation II: Artificial Intelligence Services
	7.1. SERVICE PROVISIONING FRAMEWORKS
		7.1.1. Spring Cloud
		7.1.2. Service Configuration
		7.1.3. Service Registration at Edge
		7.1.4. Service Gateway
	7.2. DEPLOY AI MODELS
		7.2.1. Packed as a Microservice
		7.2.2. Microservice Deployment
		7.2.3. Platforms for AI Services
	7.3. CHALLENGES FOR AI-BASED SERVICES COMPOSITION
		7.3.1. Feature Heterogeneity
		7.3.2. High-Dimensional Data
		7.3.3. Dynamic Raw Data
	7.4. CHAPTER SUMMARY
Chapter 8: Evaluation
	8.1. EVALUATION METHODS AND CRITERIA
	8.2. PROTOTYPE CASE STUDY
		8.2.1. Case Study Configurations
		8.2.2. Samples and Results
			8.2.2.1. Composition Planning Case
			8.2.2.2. Adaptation Case
	8.3. SIMULATION STUDIES
		8.3.1. Environment Configurations
			8.3.1.1. General Settings
			8.3.1.2. Evaluation Scenarios
		8.3.2. Baseline Approach
		8.3.3. Simulation Results and Analysis
			8.3.3.1. Flexibility of Service Planning
			8.3.3.2. Adaptability of Composite Services
			8.3.3.3. Impact of Heuristic Service Discovery
			8.3.3.4. Planning Complex Service Flows
			8.3.3.5. Availability of Cooperative Microservices Provisioning
	8.4. EVALUATION SUMMARY
		8.4.1. Service Composition
		8.4.2. Cooperative Service Provisioning
Chapter 9: Discussion and Conclusion
	9.1. ACHIEVEMENTS
Appendix A: Further Implementation Detail: Prototypes
	A.1. GOCOMO APP
	A.2. GOCOMO-N.3
Appendix B: Evaluation Results’ Validity
	B.1. RESULTS’ VALIDITY USING.2-SAMPLE Z-TEST
		B.1.1. CoopC and GoCoMo’s Service Discovery Delay
		B.1.2. CoopC and GoCoMo’s Service Discovery Traffic
		B.1.3. CoopC and GoCoMo’s Response Time
		B.1.4. CoopC and GoCoMo’s Composition Traffic
Appendix C: Glossary of Terms
Bibliography