Multi-domain Elastic Optical Networks Research Articles

On the Game-Theoretic Analysis of Dynamic VNF Service Chaining in Edge-Cloud EONs

Network function virtualization (NFV) enables flexible, cost-effective, and timely deployment of new network services in elastic optical networks (EONs). Nowadays, the integration of cloud and edge computing becomes a prevailing trend, which has promoted the idea of edge-cloud EON (EC-EON) ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , a multi-domain EON that consists of a cloud domain and one or more edge domains). In this work, we study the dynamic provisioning of virtual network function based service chain (VNF-SC) requests in EC-EONs, and leverage game theory to explain why and how the cloud and edge domain managers (DMs) should collaborate when protecting their own interests. We first formulate the non-cooperative interactions between cloud and edge DMs as a two-stage Stackelberg game, prove the existence of Nash Equilibrium in the game, and propose an algorithm based on backward induction for the non-cooperative service provisioning. Next, we address the cooperative provisioning scheme where the DMs can reach an agreement for mutual benefit, and model it with Nash bargaining. Finally, we conduct extensive simulations to compare the non-cooperative and cooperative provisioning schemes with the traditional centralized approach in EC-EONs in-depth. Simulation results confirm the effectiveness of our non-cooperative and cooperative provisioning schemes on protecting the interests of DMs, and suggest that the cooperative provisioning scheme can outperform the traditional centralized approach in terms of blocking probability, when inter- and intra-domain VNF-SC requests both exist in an EC-EON.

Hierarchical Reinforcement Learning in Multi-Domain Elastic Optical Networks to Realize Joint RMSA

To improve the network scalability, a large elastic optical network is typically segmented into multiple autonomous domains, where each domain possesses high autonomy and privacy. This architecture is referred to as the multi-domain elastic optical network (MDEON). In the MDEON, the routing, modulation, and spectrum allocation (RMSA) for the inter-domain service requests are challenging. As a result, deep reinforcement learning (DRL) has been introduced recently where the RMSA policies are learned during the interaction of the DRL agents with the MDEON environment. Due to the autonomy of each domain in the MDEON, joint RMSA is essential to improve the overall performance. To realize the joint RMSA, we propose a hierarchical reinforcement learning (HRL) framework which consists of a high-level DRL module and multiple low-level DRL modules (one for each domain), with the collaboration of DRL modules. More specifically, for the inter-domain service requests, the high-level module obtains some abstracted information from the low-level DRL modules and generates the inter-domain RMSA decision for the low-level modules. Then the low-level DRL module gives the intra-domain RMSA decision and feeds back to the high-level module. The proposed HRL framework preserves the autonomy of each single domain while delivering effective overall network performance through the cooperation of the high-level and low-level DRL modules. Simulation results demonstrate that our proposed method outperforms previous approaches.

Performance studies of evolutionary transfer learning for end-to-end QoT estimation in multi-domain optical networks [Invited

This paper proposes an evolutionary transfer learning approach (Evol-TL) for scalable quality-of-transmission (QoT) estimation in multi-domain elastic optical networks (MD-EONs). Evol-TL exploits a broker-based MD-EON architecture that enables cooperative learning between the broker plane (end-to-end) and domain-level (local) machine learning functions while securing the autonomy of each domain. We designed a genetic algorithm to optimize the neural network architectures and the sets of weights to be transferred between the source and destination tasks. We evaluated the performance of Evol-TL with three case studies considering the QoT estimation task for lightpaths with (i) different path lengths (in terms of the numbers of fiber links traversed), (ii) different modulation formats, and (iii) different device conditions (emulated by introducing different levels of wavelength-specific attenuation to the amplifiers). The results show that the proposed approach can reduce the average amount of required training data by up to13×while achieving an estimation accuracy above 95%.

Routing, spectrum and core assignment for multi-domain elastic optical networks with multi-core fibers

Abstract Routing, spectrum assignment are key issues and the hot focus area in multi-domain elastic optical networks (EONs). When the EONs with multi-cores, more challenges will be faced, such as core allocation, crosstalk, etc. To tackle these challenges, we establish a mixed integer linear programming model with the maximum index used frequency slots (MIUFS) to be minimized, which is used to determine the optimal schemes of routing, core allocation and spectrum assignment for the multi-domain EONs with multi-cores. To solve the model efficiently, an efficient genetic algorithm with tailor-made crossover, mutation and local search operators is designed. To evaluate the model and the proposed algorithm, several simulation experiments are conducted on two different multi-domain EONs, and the experimental results show that the proposed algorithm has a higher performance than the compared algorithms.

Demonstration of distributed collaborative learning with end-to-end QoT estimation in multi-domain elastic optical networks.

This paper proposes a distributed collaborative learning approach for cognitive and autonomous multi-domain elastic optical networking (EON). The proposed approach exploits a knowledge-defined networking framework which leverages a broker plane to coordinate the operations of multiple EON domains and applies machine learning (ML) to support autonomous and cognitive inter-domain service provisioning. By employing multiple distributed ML blocks learning domain-level features and working with broker plane aggregation ML blocks (through the chain rule-based training), the proposed approach enables to develop cognitive networking applications that can fully exploit the multi-domain EON states while obviating the need for the raw and confidential intra-domain data. In particular, we investigate end-to-end quality-of-transmission estimation application using the distributed learning approach and propose three estimator designs incorporating the concepts of multi-task learning (MTL) and transfer learning (TL). Evaluations with experimental data demonstrate that the proposed designs can achieve estimation accuracies very close to (with differences less than 0.5%) or even higher than (with MTL/TL) those of the baseline models assuming full domain visibility.

Building Autonomic Elastic Optical Networks with Deep Reinforcement Learning

Conventional schemes for service provisioning in next-generation elastic optical networks (EONs) rely on rule-based policies that suffer from scalability issues and can lead to poor resource utilization efficiency due to the lack of knowledge about the essential characteristics of EONs (e.g., traffic profiles, physical-layer impairments). This article discusses the application of emerging deep reinforcement learning (DRL) techniques in EONs for enabling an autonomic (self-driving) and cognitive networking framework. This new framework achieves self-learning-based service provisioning capabilities by employing DRL agents to learn policies from dynamic network operations. Such capabilities can remarkably reduce the amount of human effort invested in developing effective service provisioning policies for emerging applications, and thus, can facilitate fast network evolutions. Based on the framework, we first present DeepRMSA, a DRL-based routing, modulation, and spectrum assignment (RMSA) agent for EONs. Then, as today's networks are often composed of multiple autonomous systems, we extend the autonomic networking framework to multi-domain EONs by applying multi-agent DRL (where multiple autonomous DRL agents learn through jointly interacting with their environments). Comparisons of the results from numerical simulations show significant advantages of the proposed framework over the existing rule-based heuristic designs.

Experimental Demonstration of Machine-Learning-Aided QoT Estimation in Multi-Domain Elastic Optical Networks with Alien Wavelengths

In multi-domain elastic optical networks with alien wavelengths, each domain needs to consider intradomain and interdomain alien traffic to estimate and guarantee the required quality of transmission (QoT) for each lightpath and perform provisioning operations. This paper experimentally demonstrates an alien wavelength performance monitoring technique and machine-learning-aided QoT estimation for lightpath provisioning of intradomain/ interdomain traffic. Testbed experiments demonstrate modulation format recognition, QoT monitoring, and cognitive routing for a 160 Gbaud alien multi-wavelength light- path. By using experimental training datasets from the testbed and an artificial neural network, we demonstrated an accurate optical-signal-to-noise ratio prediction with an accuracy of ~95% when using 1200 data points.

Knowledge-Based Autonomous Service Provisioning in Multi-Domain Elastic Optical Networks

Optimization of end-to-end service provisioning across multi-domain EONs is critically important due to the restricted intra-domain visibility and the agile spectrum allocation mechanism of EONs. Previously demonstrated schemes apply fixed service provisioning policies (e.g., heuristic routing and spectrum assignment and threshold- based anomaly detection) and fail to fully address the inherent principle and dynamic conditions of multi-domain EONs. This article brings in data analytics and presents a knowledge-based autonomous service provisioning framework for multi-domain EONs. The proposed framework employs a new broker plane working with domain managers while supporting the autonomy of each management domain. This can be seen as a hierarchical multi-domain network control and management architecture in support of enhanced resource efficiency and network scalability where each hierarchy employs an observe-analyze-act cycle. Specifically, the broker and domain managers can observe the network state through comprehensive network monitoring (monitoring of optical paths' performance, resource utilization, and traffic dynamics), analyze the network behavior and form the knowledge repository of network rules using machine learning techniques, and act intelligently according to the obtained knowledge. We assess the proposed framework by evaluating two key knowledge-based service provisioning use cases: proactive inter-domain traffic-prediction- based autonomous traffic engineering and quality-of-transmission-aware inter-domain lightpath provisioning. Numerical results demonstrate remarkable benefits of the proposed framework over conventional solutions.

Leveraging Game Theory to Achieve Efficient Attack-Aware Service Provisioning in EONs

Multidomain elastic optical networks (MD-EONs) help to improve network scalability, extend service coverage, and facilitate good interoperability to orchestrate administrative domains managed by different carriers. Since the users in other domains can launch cross-domain physical-layer attacks to a domain, this paper studies the problem of attack-aware service provisioning in one domain of an MD-EON. We consider a realistic scenario that does not treat all the interdomain lightpaths as malicious ones, and try to arrange the lightpaths' routing and spectrum assignment (RSA) schemes with the help of the game theory to balance the spectrum utilization and security-level of the domain well. Specifically, we define a two-player Bayesian game to represent the provisioning procedure for each interdomain request, and design the game strategies and utility functions for the players (i.e., the domain manager and the user from other domains). Then, we formulate a nonlinear programming model, solve the game with it to obtain a Bayesian Nash equilibrium (BNE), and determine the best strategies for the players based on the BNE. Finally, with the game model, we propose a game-assisted RSA algorithm to achieve attack-aware service provisioning efficiently. The proposed algorithm is evaluated with extensive simulations and the results confirm its effectiveness.

Optical communications networks

The field of optical communications continues to evolve to meet the ever increasing demands for capacity, flexibility, scalability, and security. In this second installment of the Optical Communications Networks Series in 2016, we have selected six contributions that address physical layer security in optical networks, resource optimization in cloud-based radio over fiber networks, transceiver configuration in flex-grid optical networks, control plane architecture in multi-domain elastic optical networks, cost assessment of FTTdp networks with G.Fast, and hybrid ROADM architectures for scalable C/DWDM metro networks. Diverse topics addressed in this issue are indicative of the pervasive role of optical communications technologies in modern communications networks

A control plane architecture for multi-domain elastic optical networks: the view of the IDEALIST project

A key objective of the IDEALIST project included the design and implementation of a GMPLS and PCE-based control plane for multi-vendor and multi-domain flexi-grid EON, leveraging the project advances in optical switching and transmission technology, an enabling interoperable deployment. A control plane, relying on a set of entities, interfaces, and protocols, provides the automation of the provisioning, recovery, and monitoring of end-to-end optical connections. This article provides an overview of the implemented architecture. We present the macroscopic system along with the core functional blocks, control procedures, message flows, and protocol extensions. The implemented end-toend architecture adopted active stateful hierarchical PCE, under the control and orchestration of an adaptive network manager, interacting with a parent PCE, which first coordinates the selection of domains and the end-to-end provisioning using an abstracted view of the topology, and second, delegates the actual computation and intra-domain provisioning to the corresponding children PCEs. End-to-end connectivity is obtained by either a single LSP, or by the concatenation of multiple LSP segments, which are set up independently by the underlying GMPLS control plane at each domain. The architecture and protocol extensions have been implemented by several partners, assessing interoperability in a multi-partner testbed and adoption by the relevant Internet SDO (standards development organization).

Attack-Aware Service Provisioning to Enhance Physical-Layer Security in Multi-Domain EONs

It is known that with the multi-domain scenario, elastic optical networks (EONs) can improve network scalability, extend service coverage, and handle the multi-carrier situation. However, as a malicious client can launch cross-domain attacks in the physical layer, the security issues in multi-domain EONs should not be overlooked. In this paper, we consider how to improve the physical-layer security-level of multi-domain EONs. Specifically, we propose to differentiate the routing and spectrum assignment (RSA) schemes of intra- and inter-domain requests with security considerations. To achieve this, we review the physical-layer vulnerabilities due to different clients (especially trusted and untrusted ones) sharing optical components in EONs, analyze the potential attack scenarios to different RSA arrangements, and quantify the corresponding security threats with an attack factor. Then, we define the problem of multi-domain attack-aware RSA and formulate an integer linear programming model to solve it exactly. To reduce the time complexity, a heuristic algorithm is also proposed. The proposed algorithms are evaluated with extensive simulations using both the offline and online provisioning scenarios, and the simulation results verify its effectiveness.

Multipartner Demonstration of BGP-LS-Enabled Multidomain EON Control and Instantiation With H-PCE [Invited

The control of multidomain elastic optical networks (EONs) is possible by combining Hierarchical Path Computation Element (H-PCE)-based computation, Border Gateway Protocol with Extensions for Traffic Engineering Link State Information (BGP-LS) topology discovery, remote instantiation via Path Computation Element Communication Protocol (PCEP), and signaling via Resource Reservation Protocol with Extensions for Traffic Engineering (RSVP-TE). Two evolutionary architectures are considered, one based on stateless H-PCE, PCEP instantiation, and end-to-end RSVP-TE signaling (SL-E2E), and a second one based on stateful active H-PCE with perdomain instantiation and stitching. This paper presents the first multiplatform demonstration that fully validates both control architectures achieving multiprotocol interoperability. SL-E2E leads to slightly faster provisioning but needs to keep the state of the stitching of the end-to-end label-switched paths in the parent PCE.

Proactive H-PCE Architecture With BGP-LS Update for Multidomain Elastic Optical Networks [Invited

The hierarchical path computation element (H-PCE) architecture has been proposed to perform effective end-to-end path computation in multidomain elastic optical networks (EONs). In the H-PCE architecture a single parent PCE is responsible for inter-domain path computations, while a dedicated child PCE performs intra-domain path computations within each domain. In this scenario, effective inter-domain path computation is achieved only if detailed and updated intra-domain resource availability information is locally stored or retrievable by the parent PCE. Border Gateway Protocol has been recently extended to transport link state information (BGP-LS) that can be used to update the resource availability information at the parent PCE, thus improving scalability and effectiveness of the H-PCE architecture. Specifically, BGP-LS speakers located at each child PCE have been demonstrated to update the parent PCE with per-link spectrum slices availability information triggered by the reception of local Interior Gateway Protocol (IGP) advertisements. This paper proposes a proactive scheme to update the parent PCE, in which the generation of BGP-LS updating messages is not dependant on the IGP advertisements but automatically triggered by path computation requests at the child PCEs. The proposed scheme is applied during provisioning and restoration phases in a multidomain EON scenario and compared against two reference schemes. Simulation results show a significant reduction in terms of blocking probability and amount of exchanged control plane messages.

Ant colony optimization based multi-faults localization mechanism in elastic optical networks

In order to withstand and recovery from multi-faults in elastic optical networks, we propose a novel multi-fault localization mechanism based on ant colony optimization and mixed line-rates. Multi-faults localization has been proved to be a NP-complete problem in wavelength switched optical networks, and all existing multi-faults localization algorithms require time that is super polynomial in the input size. Furthermore, multi-faults localization in elastic optical networks gets new features that the affected high-bit-rate services will play a greater role than the affected low-bit-rate services. In order to handle the mixed line-rates, we introduce the dependency metric which is used to describe dependency between alarms and likely causes. We establish the linear programming model for multi-faults localization and propose an objective function while considering the mixed line-rates. We implement the ant colony optimization based multi-faults localization mechanism on the stateful PCE-based multi-domain elastic optical networks test bed. The numerical results show that ant colony optimization based multi-faults localization mechanism has low flooding time and alarm packets, high success rate compared with the existing localization algorithms. We choose the best configuration of ant colony optimization based multi-faults localization by adjusting the parameters.