Muutke küpsiste eelistusi

Towards Cognitive Autonomous Networks: Network Management Automation for 5G and Beyond [Kõva köide]

Edited by , Edited by
  • Formaat: Hardback, 560 pages, kõrgus x laius x paksus: 229x152x15 mm, kaal: 680 g
  • Ilmumisaeg: 22-Oct-2020
  • Kirjastus: John Wiley & Sons Inc
  • ISBN-10: 1119586380
  • ISBN-13: 9781119586388
Teised raamatud teemal:
Towards Cognitive Autonomous Networks: Network Management Automation for 5G and BeyondSuurem pilt
  • Formaat: Hardback, 560 pages, kõrgus x laius x paksus: 229x152x15 mm, kaal: 680 g
  • Ilmumisaeg: 22-Oct-2020
  • Kirjastus: John Wiley & Sons Inc
  • ISBN-10: 1119586380
  • ISBN-13: 9781119586388
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781119586388.html
Märksõnad:
Learn about the latest in cognitive and autonomous network management Towards Cognitive Autonomous Networks: Network Management Automation for 5G and Beyond delivers a comprehensive understanding of the current state-of-the-art in cognitive and autonomous network operation. Authors Mwanje and Bell fully describe todays capabilities while explaining the future potential of these powerful technologies. This book advocates for autonomy in new 5G networks, arguing that the virtualization of network functions render autonomy an absolute necessity. Following that, the authors move on to comprehensively explain the background and history of large networks, and how we come to find ourselves in the place were in now. Towards Cognitive Autonomous Networks describes several novel techniques and applications of cognition and autonomy required for end-to-end cognition including: • Configuration of autonomous networks • Operation of autonomous networks • Optimization of autonomous networks • Self-healing autonomous networks The book concludes with an examination of the extensive challenges facing completely autonomous networks now and in the future.
List of Contributors
xix
Foreword xxi
Foreword II xxv
Preface xxvii
1 The Need for Cognitive Autonomy in Communication Networks
1(28)
Stephen S. Mwanje
Christian Mannweiler
Henning Sanneck
1.1 Complexity in Communication Networks
2(9)
1.1.1 The Network as a Graph
2(2)
1.1.2 Planes, Layers, and Cross-Functional Design
4(2)
1.1.3 New Network Technology - 5G
6(3)
1.1.4 Processes, Algorithms, and Automation
9(1)
1.1.5 Network State Changes and Transitions
9(1)
1.1.6 Multi-RAT Deployments
10(1)
1.2 Cognition in Network Management Automation
11(8)
1.2.1 Business, Service and Network Management Systems
11(2)
1.2.2 The FCAPS Framework
13(2)
1.2.3 Classes/Areas of NMA Use Cases
15(2)
1.2.4 SON - The First Generation of NMA in Mobile Networks
17(1)
1.2.5 Cognitive Network Management - Second Generation NMA
18(1)
1.2.6 The Promise of Cognitive Autonomy
18(1)
1.3 Taxonomy for Cognitive Autonomous Networks
19(10)
1.3.1 Automation, Autonomy, Self-Organization, and Cognition
19(2)
1.3.2 Data Analytics, Machine Learning, and AI
21(1)
1.3.3 Network Autonomous Capabilities
22(1)
1.3.4 Levels of Network Automation
23(2)
1.3.5 Content Outline
25(1)
1.3.5.1 Requirements Analysis
26(1)
1.3.5.2 Foundations
26(1)
1.3.5.3 Recent Cognitive Solutions
26(1)
1.3.5.4 Motivating the Future
26(1)
References
27(2)
2 Evolution of Mobile Communication Networks
29(64)
Christian Mannweiler
Cinzia Sartori
Bernhard Wegmann
Hannu Flinck
Andreas Maeder
Jurgen Goerge
Rudolf Winkelmann
2.1 Voice and Low-Volume Data Communications
30(8)
2.1.1 Service Evolution - From Voice to Mobile Internet
31(2)
2.1.2 2G and 3G System Architecture
33(2)
2.1.3 GERAN - 2G RAN
35(1)
2.1.4 UTRAN - 3G RAN
36(2)
2.2 Mobile Broadband Communications
38(4)
2.2.1 Mobile Broadband Services and System Requirements
38(1)
2.2.2 4G System Architecture
39(1)
2.2.3 E-UTRAN - 4G RAN
40(2)
2.3 Network Evolution - Towards Cloud-Native Networks
42(7)
2.3.1 System-Level Technology Enablers
42(4)
2.3.2 Challenges and Constraints Towards Cloud-Native Networks
46(1)
2.3.3 Implementation Aspects of Cloud-Native Networks
47(2)
2.4 Multi-Service Mobile Communications
49(20)
2.4.1 Multi-Tenant Networks for Vertical Industries
50(1)
2.4.2 5G System Architecture
51(3)
2.4.3 Service-Based Architecture in the 5G Core
54(2)
2.4.4 5G RAN
56(3)
2.4.5 5G New Radio
59(4)
2.4.6 5G Mobile Network Deployment Options
63(6)
2.5 Evolution of Transport Networks
69(3)
2.5.1 Architecture of Transport Networks
69(1)
2.5.2 Transport Network Technologies
70(2)
2.6 Management of Communication Networks
72(15)
2.6.1 Basic Principles of Network Management
72(4)
2.6.2 Network Management Architectures
76(1)
2.6.2.1 Legacy 3GPP Management Integration Architecture
77(1)
2.6.2.2 Service-Based Architecture in Network Management
78(1)
2.6.3 The Role of Information Models in Network Management
79(1)
2.6.4 Dimensions of Describing Interfaces
80(1)
2.6.4.1 Dimension 1: Hierarchy of the Management Function
80(1)
2.6.4.2 Dimension 2: Levels of Abstraction
81(1)
2.6.4.3 Dimension 3: Layers in Communication
81(1)
2.6.4.4 Dimension 4: Meta Data
81(1)
2.6.5 Network Information Models
82(1)
2.6.5.1 Model of the Dynamic Behaviour
82(2)
2.6.5.2 Format of the Data
84(1)
2.6.5.3 Semantical Part of the Model
85(1)
2.6.6 Limitations of Common Information Models
85(2)
2.1 Conclusion - Cognitive Autonomy in 5G and Beyond
87(6)
2.7.1 Management of Individual 5G Network Features
87(1)
2.7.2 End-to-End Operation of 5G Networks
88(1)
2.7.3 Novel Operational Stakeholders in 5G System Operations
88(1)
References
89(4)
3 Self-Organization in Pre-5G Communication Networks
93(52)
Muhammad Naseer-ul-lslam
Janne Ali-Tolppa
Stephen S. Mwanje
Guillaume Decarreau
3.1 Automating Network Operations
94(4)
3.1.1 Traditional Network Operations
94(1)
3.1.2 SON-Based Network Operations
95(1)
3.1.2.1 Centralized SON
95(1)
3.1.2.2 Distributed SON
96(1)
3.1.2.3 Hybrid SON
97(1)
3.1.3 SON Automation Areas and Use Cases
97(1)
3.2 Network Deployment and Self-Configuration
98(10)
3.2.1 Plug and Play
98(1)
3.2.1.1 Auto-Connectivity
99(1)
3.2.1.2 Auto-Commissioning
99(1)
3.2.1.3 Dynamic Radio Configuration
100(1)
3.2.2 Automatic Neighbour Relations (ANR)
101(1)
3.2.2.1 The ANR Procedure
101(2)
3.2.2.2 NRT and ANR Limitations
103(1)
3.2.3 LTE Physical Cell Identity (PCI) Assignment
103(1)
3.2.3.1 PCI Assignment Objectives
104(2)
3.2.3.2 PCI Assignment Strategies
106(1)
3.2.3.3 PCI Assignment Challenges
107(1)
3.3 Self-Optimization
108(16)
3.3.1 Mobility Load Balancing (MLB)
108(1)
3.3.1.1 Scenarios for Load Balancing and Traffic Steering
109(1)
3.3.1.2 Standardization Support for Load Balancing and Traffic Steering
109(2)
3.3.2 Mobility Robustness Optimization (MRO)
111(1)
3.3.2.1 Optimization Objectives for MRO
111(3)
3.3.2.2 Standardization Support for MRO
114(1)
3.3.3 Energy Saving Management
115(1)
3.3.3.1 Scenarios for Energy Saving
116(1)
3.3.3.2 Standardization Support for Energy Saving
117(1)
3.3.4 Coverage and Capacity Optimization (CCO)
117(1)
3.3.4.1 Scenarios for CCO
118(1)
3.3.4.2 Solution Ideas for CCO
119(1)
3.3.5 Random Access Channel (RACH) Optimization
120(2)
3.3.6 Inter-Cell Interference Coordination (ICIC)
122(2)
3.4 Self-Healing
124(5)
3.4.1 The General Self-Healing Process
125(1)
3.4.2 Cell Degradation Detection
125(2)
3.4.3 Cell Degradation Diagnosis
127(1)
3.4.4 Cell Outage Compensation
128(1)
3.5 Support Function for SON Operation
129(7)
3.5.1 SON Coordination
129(1)
3.5.1.1 SON Function Conflicts
129(2)
3.5.1.2 SON Function Coordination
131(2)
3.5.2 Minimization of Drive Test (MDT)
133(1)
3.5.2.1 Scenarios and Use Cases for Drive Tests
133(2)
3.5.2.2 Standardization Support for MDT
135(1)
3.6 5G SON Support and Trends in 3GPP
136(4)
3.6.1 Critical 5G RAN Features
136(1)
3.6.2 SON Standardization for 5G
137(3)
3.7 Concluding Remarks
140(5)
References
141(4)
4 Modelling Cognitive Decision Making
145(28)
Stephen S. Mwanje
Henning Sanneck
4.1 Inspirations from Bio-Inspired Autonomy
146(2)
4.1.1 Distributed, Efficient Equilibria
146(1)
4.1.2 Distributed, Effective Management
147(1)
4.1.3 Robustness Amidst Self-Organization
147(1)
4.1.4 Adaptability
147(1)
4.1.5 Natural Stochasticity
148(1)
4.1.6 From Simplicity Emerges Complexity
148(1)
4.2 Self-Organization as Visible Cognitive Automation
148(6)
4.2.1 Attempts at Definition
149(1)
4.2.2 Bio-Chemical Examples of Self-Organizing Systems
149(2)
4.2.3 Human Social-Economic Examples of Self-Organizing Systems
151(1)
4.2.4 Features of Self-Organization - As Evidenced by Ant Foraging
152(2)
4.2.5 Self-Organization or Cognitive Autonomy? - The Case of Ants
154(1)
4.3 Human Cognition
154(5)
4.3.1 Basic Cognitive Processes
155(1)
4.3.2 Higher, Complex Cognitive Processes
156(1)
4.3.2.1 Thought
156(1)
4.3.2.2 Language
157(1)
4.3.2.3 Problem-Solving
157(1)
4.3.2.4 Intelligence
157(1)
4.3.3 Cognitive Processes in Learning
158(1)
4.4 Modelling Cognition: A Perception-Reasoning Pipeline
159(8)
4.4.1 Conceptualization
160(1)
4.4.2 Contextualization
160(1)
4.4.3 Organization
161(1)
4.4.4 Inference
161(1)
4.4.5 Memory Operations
162(1)
4.4.6 Concurrent Processing and Actioning
162(1)
4.4.7 Attention and the Higher Processes
163(1)
4.4.8 Comparing Models of Cognition
164(3)
4.5 Implications for Network Management Automation
167(2)
4.5.1 Complexity of the PRP Processes
167(1)
4.5.2 How Cognitive Is SON?
168(1)
4.5.3 Expectations from Cognitive Autonomous Networks
168(1)
4.6 Conclusions
169(4)
References
170(3)
5 Classic Artificial Intelligence: Tools for Autonomous Reasoning
173(30)
Stephen Mwanje
Marton Kajo
Benedek Schultz
Kimmo Hatonen
Ilaria Malanchini
5.1 Classical AI: Expectations and Limitations
174(3)
5.1.1 Caveat: The Common-Sense Knowledge Problem
174(1)
5.1.2 Search and Planning for Intelligent Decision Making
175(1)
5.1.3 The Symbolic AI Framework
176(1)
5.2 Expert Systems
177(3)
5.2.1 System Components
177(1)
5.2.2 Cognitive Capabilities and Application of Expert Systems
177(1)
5.2.3 Rule-Based Handover-Events Root Cause Analysis
178(1)
5.2.4 Limitations of Expert Systems
179(1)
5.3 Closed-Loop Control Systems
180(2)
5.3.1 The Controller
180(1)
5.3.2 Cognitive Capabilities and Application of Closed-Loop Control
181(1)
5.3.3 Example: Handover Optimization Loop
181(1)
5.4 Case-Based Reasoning
182(4)
5.4.1 The CBR Execution Cycle
183(1)
5.4.2 Cognitive Capabilities and Applications of CBR Systems
184(1)
5.4.3 CBR Example for RAN Energy Savings Management
185(1)
5.4.4 Limitations of CBR Systems
185(1)
5.5 Fuzzy Inference Systems
186(6)
5.5.1 Fuzzy Sets and Membership Functions
186(1)
5.5.2 Fuzzy Logic and Fuzzy Rules
187(1)
5.5.3 Fuzzy Interference System Components
188(1)
5.5.4 Cognitive Capabilities and Applications of FIS
189(1)
5.5.5 Example Application: Selecting Handover Margins
190(1)
5.5.5.1 Step 1: Fuzzification
190(1)
5.5.5.2 Step 2: Apply Fuzzy Operators)
191(1)
5.5.5.3 Step 3: Apply Weighted Implication
192(1)
5.5.5.4 Step 4: Aggregate All Outputs
192(1)
5.5.5.5 Step 5: Defuzzify
192(1)
5.6 Bayesian Networks
192(4)
5.6.1 Definitions
193(1)
5.6.2 Example Application: Diagnosis in Mobile Networks
193(1)
5.6.3 Selecting and Training Bayesian Networks
194(1)
5.6.4 Cognitive Capabilities and Applications of Bayesian Networks
195(1)
5.7 Time Series Forecasting
196(3)
5.7.1 Time Series Modelling
196(2)
5.7.2 Auto Regressive and Moving Average Models
198(1)
5.7.3 Cognitive Capabilities and Applications of Time Series Models
198(1)
5.8 Conclusion
199(4)
References
199(4)
6 Machine Learning: Tools for End-to-End Cognition
203(52)
Stephen Mwanje
Marton Kajoa
Benedek Schultz
6.1 Learning from Data
204(15)
6.1.1 Definitions
205(2)
6.1.2 Training Using Numerical Optimization
207(2)
6.1.3 Over- and Underfilling, Regularization
209(2)
6.1.4 Supervised Learning in Practice - Regression
211(1)
6.1.5 Supervised Learning in Practice - Classification
212(1)
6.1.6 Unsupervised Learning in Practice - Dimensionality Reduction
213(1)
6.1.6.1 Factor Analysis
213(1)
6.1.6.2 Principal Components Analysis
214(1)
6.1.6.3 Independent Components Analysis
214(1)
6.1.6.4 Implementations
215(1)
6.1.7 Unsupervised Learning in Practice - Clustering Using K-Means
215(1)
6.1.8 Cognitive capabilities and Limitations of Machine Learning
216(2)
6.1.9 Example Application: Temporal-Spatial Load Profiling
218(1)
6.2 Neural Networks
219(8)
6.2.1 Neurons and Activation Functions
220(1)
6.2.2 Neural Network Computational Model
221(1)
6.2.3 Training Through Gradient Descent and Backpropagation
222(2)
6.2.4 Overfitting and Regularization
224(2)
6.2.5 Cognitive Capabilities of Neural Networks
226(1)
6.2.6 Application Areas in Communication Networks
226(1)
6.3 A Dip into Deep Neural Networks
227(14)
6.3.1 Deep Learning
227(1)
6.3.2 The Vanishing Gradients Problem
228(1)
6.3.3 Drivers, Enablers, and Computational Constraints
229(1)
6.3.3.1 Computational Power
230(1)
6.3.3.2 Timing Constraints
230(1)
6.3.3.3 Quantity of Data
231(1)
6.3.4 Convolutional Networks for Image Recognition
231(2)
6.3.4.1 Convolution Layers
233(1)
6.3.4.2 Max Pooling
234(1)
6.3.5 Recurrent Neural Networks for Sequence Processing
235(1)
6.3.5.1 Long Short-Term Memory
236(1)
6.3.6 Combining LSTMs with Convolutional Networks
237(1)
6.3.7 Autoencoders for Data Compression and Cleaning
238(2)
6.3.8 Cognitive Capabilities and Application of Deep Neural Networks
240(1)
6.4 Reinforcement Learning
241(12)
6.4.1 Learning Through Exploration
241(1)
6.4.2 RL Challenges and Framework
242(1)
6.4.3 Value Functions
243(1)
6.4.4 Model-Based Learning Through Value and Policy Iteration
244(1)
6.4.5 Q-Learning Through Dynamic Programming
245(1)
6.4.6 Linear Function Approximation
246(1)
6.4.7 Generalized Approximators and Deep Q-Learning
247(1)
6.4.8 Policy Gradient and Actor-Critic Methods
248(1)
6.4.8.1 Reinforce Algorithm
249(1)
6.4.8.2 Reducing Variance
250(1)
6.4.8.3 Policy Gradient Algorithm
251(1)
6.4.8.4 Actor-Critic
251(1)
6.4.9 Cognitive Capabilities and Application of Reinforcement Learning
252(1)
6.5 Conclusions
253(2)
References
253(2)
7 Cognitive Autonomy for Network Configuration
255(46)
Stephen S. Mwanje
Rashid Mijumbi
Lars Christoph Schmelz
7.1 Context Awareness for Auto-Configuration
256(11)
7.1.1 Environment, Network, and Function Contexts
257(2)
7.1.2 NAF Context-Aware Configuration
259(1)
7.1.3 Objective Model
260(3)
7.1.4 Context Model - Context Regions and Classes
263(2)
7.1.5 Deriving the Context Model
265(1)
7.1.6 Deriving Network and Function Configuration Policies
266(1)
7.2 Multi-Layer Co-Channel PCI Auto-Configuration
267(7)
7.2.1 Automating PCI Assignment in LTE and 5G Radio
268(1)
7.2.2 PCI Assignment Objectives
269(1)
7.2.3 Blind PCI Auto Configuration
270(1)
7.2.4 Initial Blind Assignment
271(1)
7.2.5 Learning Pico-Macro NRs
272(1)
7.2.6 Predicting Macro-Macro NRs
272(1)
7.2.7 PCI Update/Optimization and New Cells Configuration
273(1)
7.2.8 Performance Expectations
273(1)
7.3 Energy Saving Management in Multi-Layer RANs
274(11)
7.3.1 The HetNet Energy Saving Management Challenge
275(1)
7.3.2 Power Saving Groups
276(1)
7.3.3 Cell Switch-On Switch-Off Order
277(1)
7.3.4 PSG Load and ESM Triggering
278(1)
7.3.5 Static Cell Activation and Deactivation Sequence
279(1)
7.3.6 Reference-Cell-Based ESM
280(1)
7.3.7 ESM with Multiple Reference Cells
281(2)
7.3.8 Distributed Cell Activation and Deactivation
283(2)
7.3.9 Improving ESM Solutions Through Cognition
285(1)
7.4 Dynamic Baselines for Real-Time Network Control
285(12)
7.4.1 DARN System Design
286(2)
7.4.2 Data Pre-Processing
288(1)
7.4.3 Prediction
288(1)
7.4.4 Decomposition
289(1)
7.4.4.1 Adaptation
290(1)
7.4.4.2 Baseline Generation
290(1)
7.4.5 Learning Augmentation
290(1)
7.4.5.1 Knowledge Base
291(1)
7.4.5.2 Alarm Generation
292(1)
7.4.5.3 Metric Clustering
293(1)
7.4.6 Evaluation
294(1)
7.4.6.1 Accuracy of Generated Baselines and Clusters
294(1)
7.4.6.2 Effect of Baseline Adaptation
294(1)
7.4.6.3 Effect of Learning Augmentation
295(2)
7.5 Conclusions
297(4)
References
298(3)
8 Cognitive Autonomy for Network-Optimization
301(44)
Stephen S. Mwanje
Mohammad Naseer Ul-Islam
Qi Liao
8.1 Self-Optimization in Communication Networks
302(4)
8.1.1 Characterization of Self-Optimization
302(2)
8.1.2 Open-and Closed-Loop Self-Optimization
304(1)
8.1.3 Reactive and Proactive Self-Optimization
305(1)
8.1.4 Model-Based and Statistical Learning Self-Optimization
306(1)
8.2 Q-Learning Framework for Self-Optimization
306(8)
8.2.1 Self-Optimization as a Learning Loop
307(1)
8.2.2 Homogeneous Multi-Agent Q-Learning
308(1)
8.2.3 The Heterogeneous Multi-Agent Q-Learning SO Framework
309(1)
8.2.4 Fuzzy Q-Learning
310(4)
8.3 QL for Mobility Robustness Optimization
314(8)
8.3.1 HO Performance and Parameters Sensitivity
314(1)
8.3.2 Q-Learning Based MRO (QMRO)
315(2)
8.3.3 Parameter Search Strategy
317(1)
8.3.4 Optimization Algorithm
318(1)
8.3.5 Evaluation
318(4)
8.4 Fuzzy Q-Learning for Tilt Optimization
322(7)
8.4.1 Fuzzy Q-Learning Controller (FQLC) Components
322(1)
8.4.1.1 State and Action Fuzzy Variables
322(1)
8.4.1.2 Rule-Based Fuzzy Inference System
323(1)
8.4.1.3 Instantaneous Reward
324(1)
8.4.2 The FQLC Algorithm
324(1)
8.4.3 Homogeneous Multi-Agent Learning Strategies
325(2)
8.4.4 Coverage and Capacity Optimization
327(1)
8.4.5 Self-Healing and eNB Deployment
327(2)
8.5 Interference-Aware Flexible Resource Assignment in 5G
329(11)
8.5.1 Muting in Wireless Networks
330(1)
8.5.2 Notations, Definitions, and Preliminaries
331(1)
8.5.3 System Model and Problem Formulation
332(2)
8.5.4 Optimal Resource Allocation and Performance Limits
334(1)
8.5.5 Successive Approximation of Fixed Point (SAFP)
335(1)
8.5.6 Partial Resource Muting
335(1)
8.5.6.1 Triggering the Resource Muting Scheme
336(1)
8.5.6.2 Detection of Bottleneck Users
336(1)
8.5.7 Evaluation
337(3)
8.6 Summary and Open Challenges
340(5)
References
341(4)
9 Cognitive Autonomy for Network Self-Healing
345(40)
Janne Ali-Totppa
Morton Kajo
Borislava Gajic
Ilaria Malanchini
Benedek Schultz
Qi Liao
9.1 Resilience and Self-Healing
346(3)
9.1.1 Resilience by Design
347(1)
9.1.2 Holistic Self-Healing
348(1)
9.2 Overview on Cognitive Self-Healing
349(9)
9.2.1 The Basic Building Blocks of Self-Healing
350(1)
9.2.2 Profiling and Anomaly Detection
351(2)
9.2.3 Diagnosis
353(1)
9.2.4 Remediation Action
354(1)
9.2.5 Advanced Self-Healing Concepts
354(2)
9.2.6 Feature Reduction and Context Selection for Anomaly Detection
356(1)
9.2.6.1 Feature Reduction
356(1)
9.2.6.2 Context Selection
357(1)
9.2.6.3 Feature Reduction and Context Selection in the Future
358(1)
9.3 Anomaly Detection in Radio Access Networks
358(8)
9.3.1 Use Cases
359(1)
9.3.2 An Overview of the RAN Anomaly Detection Process
360(1)
9.3.3 Profiling the Normal Behaviour
361(1)
9.3.4 The New Normal - Adapting to Changes
362(2)
9.3.5 Anomaly-Level Calculation
364(1)
9.3.6 Anomaly Event Detection
365(1)
9.4 Diagnosis and Remediation in Radio Access Networks
366(5)
9.4.1 Symptom Collection
367(1)
9.4.2 Diagnosis
367(2)
9.4.3 Augmented Diagnosis
369(2)
9.4.4 Deploying Corrective Actions
371(1)
9.5 Knowledge Sharing in Cognitive Self-Healing
371(8)
9.5.1 Information Sharing in Mobile Networks
371(2)
9.5.2 Transfer Learning and Self-Healing for Mobile Networks
373(1)
9.5.3 Applying Transfer Learning to Self-Healing
374(1)
9.5.4 Prognostic Cross-Domain Anomaly Detection and Diagnosis
374(1)
9.5.5 Cognitive Slice Lifecycle Management
375(1)
9.5.6 Diagnosis Knowledge Cloud
376(1)
9.5.7 Diagnosis Cloud Components
377(1)
9.5.8 Diagnosis Cloud Evaluation
378(1)
9.6 The Future of Self-Healing in Cognitive Mobile Networks
379(6)
9.6.1 Predictive and Preventive Self-Healing
379(1)
9.6.2 Predicting the Black Swan - Ludic Fallacy and Self-Healing
380(2)
References
382(3)
10 Cognitive Autonomy in Cross-Domain Network Analytics
385(34)
Szabolcs Novaczki
Peter Szilagyi
Csaba Vulkan
10.1 System State Modelling for Cognitive Automation
386(10)
10.1.1 Cognitive Context-Aware Assessment and Actioning
386(1)
10.1.2 State Modelling and Abstraction
387(2)
10.1.3 Deriving the System-State Model
389(1)
10.1.3.1 The Static-State Model
390(1)
10.1.3.2 State Trajectory Modelling and State Clustering
391(1)
10.1.4 Symptom Attribution and Interpretation
392(2)
10.1.5 Remediation and Self-Monitoring of Actions
394(2)
10.2 Real-Time User-Plane Analytics
396(9)
10.2.1 Levels of User Behaviour and Traffic Patterns
396(2)
10.2.2 Monitoring and Insight Collection
398(2)
10.2.3 Sources of U-Plane Insight
400(1)
10.2.4 Insight Analytics from Correlated Measurements
401(1)
10.2.5 Insight Analytics from Packet Patterns
402(3)
10.3 Real-Time Customer Experience Management
405(6)
10.3.1 Intent Contextualization and QoE Policy Automation
406(2)
10.3.2 QoE Descriptors and QoE Target Definition
408(2)
10.3.3 QoE Enforcement
410(1)
10.4 Mobile Backhaul Automation
411(6)
10.4.1 The Opportunities of MBH Automation
412(1)
10.4.2 Architecture of the Automated MBH Management
413(3)
10.4.3 MBH Automation Use Cases
416(1)
10.5 Summary
417(2)
References
418(1)
11 System Aspects for Cognitive Autonomous Networks
419(50)
Stephen S. Mwanje
Janne Ali-Tolppa
Ilaria Malanchini
11.1 The SON Network Management Automation System
420(3)
11.1.1 SON Framework for Network Management Automation
420(1)
11.1.2 SON as Closed-Loop Control
421(1)
11.1.3 SON Operation - The Rule-Based Multi-Agent Control
422(1)
11.2 NMA Systems as Multi-Agent Systems
423(3)
11.2.1 Single-Agent System (SAS) Decomposition
423(1)
11.2.2 Single Coordinator or Multi-Agent Team Learning
424(1)
11.2.3 Team Modelling
425(1)
11.2.4 Concurrent Games/Concurrent Learning
425(1)
11.3 Post-Action Verification of Automation Functions Effects
426(10)
11.3.1 Scope Generation
427(1)
11.3.2 Performance Assessment
428(1)
11.3.3 Degradation Detection, Scoring and Diagnosis
429(2)
11.3.4 Deploying Corrective Actions - The Deployment Plan
431(2)
11.3.5 Resolving False Verification Collisions
433(3)
11.4 Optimistic Concurrency Control Using Verification
436(4)
11.4.1 Optimistic Concurrency Control in Distributed Systems
436(1)
11.4.2 Optimistic Concurrency Control in SON Coordination
437(1)
11.4.3 Extending the Coordination Transaction with Verification
437(3)
11.5 A Framework for Cognitive Automation in Networks
440(6)
11.5.1 Leveraging CFs in the Functional Decomposition of CAN Systems
440(2)
11.5.2 Network Objectives and Context
442(1)
11.5.3 Decision Applications (DApps)
443(1)
11.5.4 Coordination and Control
444(1)
11.5.4.1 Configuration Management Engine (CME)
444(1)
11.5.4.2 Coordination Engine (CE)
445(1)
11.5.5 Interfacing Among Functions
446(1)
11.6 Synchronized Cooperative Learning in CANs
446(10)
11.6.1 The SCL Principle
448(1)
11.6.2 Managing Concurrency: Spatial-Temporal Scheduling (STS)
449(2)
11.6.3 Aggregating Peer Information
451(1)
11.6.4 SCL for MRO-MLB Conflicts
452(1)
11.6.4.1 QMRO Rewards
453(1)
11.6.4.2 QLB Reward Function
453(1)
11.6.4.3 Performance Evaluation
454(1)
11.6.4.4 Observed Performance
454(2)
11.6.4.5 Challenges and Limitations
456(1)
11.7 Inter-Function Coopetition - A Game Theoretic Opportunity
456(8)
11.7.1 A Distributed Intelligence Challenge
457(1)
11.7.2 Game Theory and Bayesian Games
458(1)
11.7.2.1 Formal Definitions
459(1)
11.7.2.2 Bayesian Games
460(1)
11.7.3 Learning in Bayesian Games
461(2)
11.7.4 CF Coordination as Learning Over Bayesian Games
463(1)
11.8 Summary and Open Challenges
464(5)
11.8.1 System Supervision
464(1)
11.8.2 The New Paradigm
465(1)
11.8.3 Old Problems with New Faces?
466(1)
References
466(3)
12 Towards Actualizing Network Autonomy
469(48)
Stephen S. Mwanje
Jurgen Goerge
Janne Ali-Tolppa
Kimmo Hatonen
Harald Bender
Csaba Rotter
Ilaria Malanchini
Henning Sanneck
12.1 Cognitive Autonomous Networks - The Vision
470(16)
12.1.1 Cognitive Techniques in Network Automation
471(1)
12.1.1.1 Matching Problem Requirements
471(1)
12.1.1.2 Development Effort vs Required Data
472(1)
12.1.1.3 Training Time vs Development Effort
473(2)
12.1.2 Success Factors in Implementing CAN Projects
475(1)
12.1.3 Implications on KPI Design and Event Logging
476(1)
12.1.4 Network Function Centralization and Federation
477(1)
12.1.5 CAN Outlook on Architecture and Technology Evolution
478(5)
12.1.6 CAN Outlook on NM System Evolution
483(3)
12.2 Modelling Networks: The System View
486(20)
12.2.1 System Description of a Mobile Network
486(2)
12.2.2 Describing Performance
488(1)
12.2.3 Implications on Automation
489(1)
12.2.4 Control Strategies
490(1)
12.2.4.1 Configuration vs Goal-Based Control
490(1)
12.2.4.2 Command-Based vs State-Based Configuration Management
491(2)
12.2.4.3 Benefits and Limitations of Configuration- and Goal-Based Control
493(1)
12.2.4.4 Implicit Mix of Strategies
494(1)
12.2.5 Two-Dimensional Continuum of Control
495(2)
12.2.6 Levels of Policy Abstraction
497(3)
12.2.7 Implications on Optimization
500(1)
12.2.7.1 Modelling Optimization
500(1)
12.2.7.2 Dealing with Uncertainty
501(1)
12.2.8 The Promise of Intent-Based Network Control
502(2)
12.2.8.1 Definition
504(1)
12.2.8.2 Intent-Based Cognitive Autonomy
505(1)
12.3 The Development - Operations Interface in CANs
506(4)
12.3.1 The DevOps Paradigm
506(2)
12.3.2 Requirements for Successful Adoption of DevOps
508(1)
12.3.3 Benefits of DevOps for CAN
509(1)
12.4 CAN as Data Intensive Network Operations
510(7)
12.4.1 Network Data: A New Network Asset
510(1)
12.4.2 From Network Management to Data Management
511(1)
12.4.3 Managing Failure in CANs
512(2)
References
514(3)
Index 517
Editors

Stephen S. Mwanje is a Senior Research Engineer in the Cognitive Network Management research team at Nokia Bell Labs in Munich, Germany. He leads the team's research on the system aspects related to the development and application of Cognition in network management, with a special focus on the management of the radio access networks. Prior to Bell Labs, he worked for 8 years in network operations, planning and deployment for radio access, microwave and fiber optic systems.

Christian Mannweiler is the Head of the Core Network Architecture & Security Research department at Nokia Bell Labs in Munich, Germany. His research focuses on network automation for 5G mobile communications systems, architectures for cloudified networks, and integration of cellular and private industrial networks. Prior to working at Bell Labs, he has been a Senior Researcher at the German Research Center for Artificial Intelligence (DFKI GmbH) in Kaiserslautern, Germany.