Provisioning, Recovery, and In-Operation Planning in Elastic Optical Networks

LUIS VELASCO, PhD, is a full Professor at Universitat Politecnica de Catalunya, BarcelonaTech, Spain. He has devoted more than 25 years in the telecommunications industry for advanced research, development, and deployment of optical networks. MARC RUIZ, PhD, is an associate researcher at Universitat Politècnica de Catalunya, BarcelonaTech, Spain, and is highly skilled in operations research and statistics applied to communication networks.

• List of Contributors xiii1 Motivation 11.1 Motivation 11.2 Book Outline 81.3 Book Itineraries 11Acknowledgment 12Part I Introduction 132 Background 152.1 Introduction to Graph Theory 162.2 Introduction to Optimization 202.3 ILP Models and Heuristics for Routing Problems 222.3.1 ILP Formulations 222.3.2 Heuristics 252.3.3 Meta]Heuristics 272.4 Introduction to the Optical Technology 302.4.1 From Opaque to Transparent Optical Networks 312.4.2 Single]Layer and Multilayer Networks 322.4.3 EON Key Technologies 332.5 Network Life Cycle 352.5.1 Connection Provisioning 362.5.2 Connection Recovery 372.6 Conclusions 403 The Routing and Spectrum Allocation Problem 433.1 Introduction 443.2 The RSA Problem 453.2.1 Basic Offline Problem Statement 453.2.2 Notation 463.3 ILP Formulations Based On Slice Assignment 473.3.1 Starting Slice Assignment RSA (SSA]RSA) Formulation 473.3.2 Slice Assignment RSA (SA]RSA) Formulation 483.4 ILP Formulations Based On Slot Assignment 493.4.1 Slot Precomputation 493.4.2 Slot Assignment RSA (CA]RSA) Formulation 503.5 Evaluation of the ILP Formulations 513.5.1 Model Size Analysis 513.5.2 Performance Comparison 523.5.3 Evaluation in Real Scenarios 543.6 The RMSA Problem 563.6.1 Notation Extensions 563.6.2 Basic Offline Problem 563.6.3 Topology Design Problem as an RMSA Problem 573.7 Conclusions 604 Architectures for Provisioning and In]operation Planning 614.1 Introduction 624.2 Architectures for Dynamic Network Operation 644.2.1 Static versus Dynamic Network Operation 644.2.2 Migration toward In]operation Network Planning 654.2.3 Required Functionalities 674.2.4 The Front]end/Back]end PCE Architecture 684.3 In]operation Planning: Use Cases 734.3.1 VNT Reconfiguration after a Failure 734.3.2 Reoptimization 764.4 Toward Cloud]Ready Transport Networks 784.5 Conclusions 84Part II Provisioning in Single Layer Networks 855 Dynamic Provisioning of p2p Demands 875.1 Introduction 885.2 Provisioning in Transparent Networks 905.2.1 Problem Statement 905.2.2 Dynamic RSA Algorithm 905.2.3 Dynamic RMSA Algorithm 915.2.4 Bulk RSA Algorithm 925.2.5 Illustrative Results 935.3 Provisioning in Translucent Networks 995.4 Dynamic Spectrum Allocation Adaption 1025.4.1 Spectrum Allocation Policies 1035.4.2 Problem Statement 1045.4.3 Spectrum Adaption Algorithms 1055.4.4 Illustrative Results 1065.5 Conclusions 1106 Transfer]based Datacenter Interconnection 1136.1 Introduction 1146.2 Application Service Orchestrator 1166.2.1 Models for Transfer]based Connections 1176.2.2 Illustrative Results 1216.3 Routing and Scheduled Spectrum Allocation 1246.3.1 Managing Transfer]based Connections 1246.3.2 The RSSA Problem 1266.3.3 ILP Formulation 1276.3.4 Algorithms to Manage Transfer]based Requests 1306.3.5 Illustrative Results 1326.4 Conclusions 1387 Provisioning Multicast and Anycast Demands 1417.1 Introduction 1427.2 Multicast Provisioning 1437.2.1 P2MP]RSA Problem Statement 1457.2.2 ILP Formulation 1457.2.3 Heuristic Algorithm 1487.2.4 Illustrative Numerical Results 1507.2.5 Proposed Workflows and Protocol Issues 1527.2.6 Experimental Assessment 1547.3 Anycast Provisioning 1567.3.1 Optical Anycast (AC_RSA) Problem Statement 1577.3.2 Exact Algorithm for the AC_RSA Problem 1577.3.3 Illustrative Numerical Results 1587.3.4 Proposed Workflow 1597.3.5 Experimental Assessment 1617.4 Conclusions 162Part III Recovery and In]operation Planning in Single Layer Networks 1638 Spectrum Defragmentation 1658.1 Introduction 1668.2 Spectrum Reallocation and Spectrum Shifting 1688.3 Spectrum Reallocation: The SPRESSO Problem 1708.3.1 Problem Statement 1708.3.2 ILP Formulation 1708.3.3 Heuristic Algorithm 1728.4 Spectrum Shifting: The SPRING Problem 1788.4.1 Problem Statement 1788.4.2 ILP Formulation 1788.4.3 Heuristic Algorithm 1798.5 Performance Evaluation 1808.5.1 SPRESSO Heuristics Tuning 1808.5.2 Heuristics versus the ILP Model 1828.5.3 Performance of the SPRESSO Algorithm 1828.6 Experimental Assessment 1848.6.1 Proposed Workflow and Algorithm 1848.6.2 PCEP Issues 1868.6.3 Experiments 1888.7 Conclusions 1919 Restoration in the Optical Layer 1939.1 Introduction 1949.2 Bitrate Squeezing and Multipath Restoration 1959.2.1 The BATIDO Problem 1979.2.2 ILP Formulation 1979.2.3 Heuristic Algorithm 2009.2.4 Numerical Results 2029.3 Modulation Format]Aware Restoration 2079.3.1 The MF]Restoration Problem 2109.3.2 Algorithm for MF]Restoration 2119.3.3 Protocol Extensions and Proposed Workflows 2139.3.4 Experimental Assessment 2169.4 Recovering Anycast Connections 2169.4.1 ILP Formulations and Algorithm 2179.4.2 Proposed Workflow 2209.4.3 Validation 2219.5 Conclusions 22310 After]Failure]Repair Optimization 22510.1 Introduction 22610.2 The AFRO Problem 22810.2.1 Problem Statement 23010.2.2 Optimization Algorithm 23010.2.3 ILP Formulation 23110.2.4 Heuristic Algorithm 23310.2.5 Disruption Considerations 23410.2.6 Performance Evaluation 23610.3 Restoration and AFRO with Multiple Paths 24010.3.1 Problem Statement 24210.3.2 MILP Formulation 24210.3.3 Heuristic Algorithm 24410.3.4 MP]AFRO Performance Evaluation 24510.4 Experimental Validation 24610.4.1 Proposed Reoptimization Workflow 24610.4.2 Experimental Assessment 24910.5 Conclusions 252Part IV Multilayer Networks 25511 Virtual Network Topology Design and Reconfiguration 25711.1 Introduction 25811.2 VNT Design and Reconfiguration Options 25911.3 Static VNT Design 26211.3.1 The VNT Design Problem 26211.3.2 MILP Formulation 26211.4 VNT Reconfiguration Based on Traffic Measures 26411.4.1 The VENTURE Problem 26411.4.2 ILP Formulation 26511.4.3 Heuristic Algorithm 26711.4.4 Proposed Workflow 27211.5 Results 27311.5.1 Simulation Results 27311.5.2 Experimental Assessment 27511.6 Conclusions 27812 Recovery in Multilayer Networks 27912.1 Introduction 28012.2 Path Restoration in GMPLS]Controlled Networks 28112.2.1 The DYNAMO Problem 28512.2.2 MP Formulation 28512.2.3 Heuristic Algorithm 29012.2.4 DYNAMO Numerical Results 29012.2.5 PCE Architecture 29712.2.6 Experimental Results 29912.3 Survivable VNT for DC Synchronization 30212.3.1 Mathematical Formulations and Algorithms 30412.3.2 Workflows and Protocol Extensions 30912.3.3 Experimental Assessment 31012.4 Conclusions 312Part V Future Trends 31313 High Capacity Optical Networks Based on Space Division Multiplexing 31513.1 Introduction 31613.2 SDM Fibers 31913.2.1 Uncoupled/Weakly Coupled Spatial Dimensions 32013.2.2 Strongly Coupled Spatial Dimensions 32013.2.3 Subgroups of Strongly Coupled Spatial Dimensions 32113.3 SDM Switching Paradigms 32213.4 Resource Allocation in SDM Networks 32513.5 Impact of Traffic Profile on the Performance of Spatial Sp]Ch Switching in SDM Networks 33213.5.1 Illustrative Results 33313.6 Impact of Spatial and Spectral Granularity on the Performance of SDM Networks Based on Spatial Sp]Ch Switching 33613.6.1 Illustrative Results 33813.7 Conclusions 34214 Dynamic Connectivity Services in Support of Future Mobile Networks 34514.1 Introduction 34614.2 C]RAN Requirements and CVN Support 34814.2.1 C]RAN Architecture Model 34914.2.2 Backhaul Requirements in C]RAN 34914.2.3 CVN Reconfiguration 35114.3 The CUVINET Problem 35414.3.1 Problem Statement 35414.3.2 MILP Formulation 35514.3.3 Heuristic Algorithm 35914.4 Illustrative Numerical Results 36114.4.1 Network Scenario 36114.4.2 Heuristic Algorithm Validation 36214.4.3 Approaches to Support CVNs 36214.4.4 Performance Evaluation 36314.5 Conclusions 36715 Toward Cognitive In]operation Planning 36915.1 Introduction 37015.2 Data Analytics for Failure Localization 37115.2.1 Algorithm for Failure Identification/Localization 37215.2.2 Experiments and Results 37515.2.3 Generic Modules to Implement the OAA Loop 37715.3 Data Analytics to Model Origin–Destination Traffic 37815.3.1 Generic Modules for VNT Reconfiguration Based on Traffic Modeling 37815.3.2 Machine Learning Procedure for Traffic Estimation 38015.3.3 Use Case I: Anomaly Detection 38315.3.4 Use Case II: VNT Reconfiguration Triggered by Anomaly Detection 39015.4 Adding Cognition to the ABNO Architecture 39315.5 Conclusions 395List of Acronyms 397References 403Index 419