Flexible-grid Networks Research Articles

Practical methods for allocating and assessing resources in flexgrid networks

This paper focuses on quantifying the efficiency of different methods used to allocate resources in flexgrid optical networks. These methods are based on a recently proposed integer linear programming formulation of the routing and spectrum assignment (RSA) problem that takes into account all possible paths and thus theoretically yields optimal solutions, whatever be the objective function. The paper advocates using a metric-based approach for assessing RSA methods preferably to the classical approach based on the blocking probability of dynamic demands because of the long lifetime of optical paths and of the necessity of evaluating an operational network’s state early enough before congestion. The main existing fragmentation metrics are extended to the network level, and a family of network remaining capacity metrics, more suited to assess congestion levels, is introduced. When demands are incrementally mapped, the latter decrease quite linearly, with a slope reflecting the quality of the RSA method. Remaining capacity values are used to compare several off-line methods where the demand sets are mapped either globally or one by one with a suitable ordering. In both on- and off-line cases, using the sum of the demands’ maximum spectrum slice index (an original objective function proposed here), provides the best performance. Finally a method to anticipate a possible congestion, based on a combination of metrics computed on the actual and a reference network instance, is presented.

1 Tbps Data Rate Transmission Using the Super Channel in the Flexi Grid Network

1 Tbps Data Rate Transmission Using the Super Channel in the Flexi Grid Network

QoT estimation using EGN-assisted machine learning for multi-period network planning

The rapidly growing traffic demands in fiber-optical networks require flexibility and accuracy in configuring lightpaths, for which fast and accurate quality of transmission (QoT) estimation is of pivotal importance. This paper introduces a machine learning (ML)-based QoT estimation approach that meets these requirements. The proposed gradient-boosting ML model uses precomputed per-channel self-channel-interference values as representative and condensed features to estimate non-linear interference in a flexible-grid network. With an enhanced Gaussian noise (GN) model simulation as the baseline, the ML model achieves a mean absolute signal-to-noise ratio error of approximately 0.1 dB, which is an improvement over the GN model. For three different network topologies and network planning approaches of varying complexities, a multi-period network planning study is performed in which ML and GN are compared as path computation elements (PCEs). The results show that the ML PCE is capable of matching or slightly improving the performance of the GN PCE on all topologies while reducing significantly the computation time of network planning by up to 70%.

Multi-step migration of optical connections to minimize disruption time in flex-grid networks

Migration of optical connections in flex-grid networks has become increasingly necessary as the arrivals and departures of optical connection requests are more dynamic. To mitigate disruptions of existing connections during migration, former research applies the resource dependency digraph (RDD) model to allow for sequential migration of optical connections. In this paper, we argue that, in this RDD model, each connection has a quota of one migration step only and that this one-step-quota constraint can be safely lifted to accomplish even lower disruption time due to the greater degree of migration freedom. We consider the whole migration process to be composed of multiple lightpath assignment states, each of which is to be determined individually. We show that this consideration can drastically reduce disruption time and hence the ensuing service level agreement violation penalties, even when it is already minimized under the one-step-quota constraint. For that, we propose two mathematical models. The first model, namely, multi-step optical connection migration (MOCM), adopts a layer-based abstraction in which each lightpath assignment state is represented as a layer, and the transition between two consecutive layers constitutes a migration step. In the second model, namely, MOCM2, to improve the solution efficiency, we readopt the concept of RDD and integrate multiple RDDs into MOCM to further enable even larger migration freedom. This way, the transition between two consecutive layers may involve multiple so-called meta-migration steps. As a use case, the proposed approaches are applied to perform connection migration for spectrum defragmentation. The results show that our multi-step migration scheme notably reduces disruption time as compared with the one-step-quota RDD-based approaches.

Brown-Field Migration Aware Routing and Spectrum Assignment in Backbone Optical Networks

With the development of optical networks technology, broad attention has been paid to flexible grid technology in optical networks due to its ability to carry large-capacity information as well as provide flexible and fine-grained services through on-demand spectrum resource allocation. However, a one-time green-field deployment of a flexible grid network may not be practical. The transition technology called the fixed/flex-grid optical networks is more applicable and highly pragmatic. In such network, many nodes would likely be upgraded from a fixed-grid to flex-grid. In fact, dynamic service provisioning during the process of a node upgrade in fixed/flex-grid optical networks have become a challenge because the service connection can be easily interrupted, which leads to considerable data loss because of node upgrade. To overcome this challenge, we propose a brown-field migration aware routing and spectrum assignment (BMA-RSA) algorithm in fixed/flex-grid optical networks. The aim is to construct a probabilistic migration label (PML) model. The well-designed label setting of PML can balance the relationship between distance and node-upgrade probability. Dynamic service provisioning operations are undertaken based on the PML model to achieve a migration-aware dynamic connection before network migration occurs. We also evaluate the performance of different service provisioning strategies under different traffic models. The simulation results show that the BMA-RSA algorithm can achieve: (1) the tradeoff between distance and node upgrade probability during the process of service provisioning; (2) lower service interruption compared with the traditional non-migration aware K-shortest-path routing and spectrum assignment algorithm.

Resource Distribution Equilibrium for Virtual Network Embedding Over Flexi-Grid Optical Networks

Virtual network embedding (VNE) in network virtualization over flexi-grid networks is one of the main challenges up to now. Many models and algorithms based on reducing the cost or keeping integrity and continuity of spectrum slots have been already proposed. In this paper, we employ the concept of distribution equilibrium of computation resources to improve performance of VNE. An integer linear programming (ILP) model and an algorithm based on the concept of distribution equilibrium for virtual node embedding are developed. The experimental results show that this algorithm has some significant advantages. And then, an ILP model based on resource distribution equilibrium and adaptive modulation for VNE over flexi-grid optical networks is established. Since the ILP problem is NP-complete, a heuristic algorithm to manipulate VNE on table is developed. The simulation results reveal the superiority in terms of acceptance ratios and system revenues.

Spectral processing techniques for efficient monitoring in optical networks

Having ubiquitous optical monitors in dense wavelength-division multiplexing (DWDM) or flex-grid networks allows the estimation in real time of crucial parameters. Such monitoring would be even more important in disaggregated optical networks, to inspect performance issues related to inter-vendor interoperability. Several important parameters can be retrieved using optical spectrum analyzers (OSAs). However, omnipresent OSAs represent an infeasible solution. Nevertheless, the advent of new, relatively cheap, compact and medium-resolution optical channel monitors (OCMs) enable a more intensive deployment of these devices. In this paper, we identify two main scenarios for the placement of such monitors: at the ingress and at the egress of the optical nodes. In the ingress scenario, we can directly estimate the parameters related to the signals, but not those related to the filters. On the contrary, in the egress scenario, the filter-related parameters can be easily detected, but not those related to amplified spontaneous emission. Therefore, we present two methods that, leveraging a curve fitting and a machine learning regression algorithm, allow detection of the missing parameters. We verify the proposed solutions with spectral data acquired in simulation and experimental setups. We obtained good estimation accuracy for both setups and for both studied placement scenarios. It is noteworthy that in the experimental assessment of the ingress scenario, we achieved a maximum absolute error (MAE) lower than 1 GHz in filter bandwidth estimation and a MAE lower than 0.5 GHz in filter frequency shift estimation. In addition, by comparing the relative errors of the considered parameters, we identified the ingress scenario as the more beneficial. In particular, we estimated the filter central frequency shift with 84% and the filter 6 dB bandwidth with 75% higher accuracy, with respect to datasheet/reference values. This translates into a total reduction of the estimated signal-to-noise ratio (SNR) penalty, introduced by a single optical filter, of 0.24 dB.

Cross-layer static resource provisioning for dynamic traffic in flexible grid optical networks

Flexible grid networks need rigorous resource planning to avoid network over-dimensioning and resource over-provisioning. The network must provision the hardware and spectrum resources statically, even for dynamic random bandwidth demands, due to the infrastructure of flexible grid networks, hardware limitations, and reconfiguration speed of the control plane. We propose a flexible online–offline probabilistic (FOOP) algorithm for the static spectrum assignment of random bandwidth demands. The FOOP algorithm considers the probabilistic nature of random bandwidth demands and balances hardware and control plane pressures with spectrum assignment efficiency. The FOOP algorithm uses the probabilistic spectrum Gaussian noise (PSGN) model to estimate the physical-layer impairment (PLI) for random bandwidth traffic. Compared to a benchmark spectrum assignment algorithm and a widely applied PLI estimation model, the proposed FOOP algorithm using the PSGN model saves up to 49% of network resources.

Connection setup and traffic aggregation in IP-based Elastic Optical Networks (EON)

To overcome the challenges of the increased IP traffic volume on the Internet, the IP/MPLS-based fixed grid DWDM network structure was proposed, which makes the interaction between both the IP layer and the optical wavelength routed fixed DWDM layer applicable. However, the deployment of this architecture was not only expensive, but power-consuming. The emerging optical Flexgrid networks technology is nowadays suggested to be the potential candidate for future optical transport networks. This is because it provides extremely flexible and spectrally effective resources utilisation. Therefore, it has attracted the attention of network operators, optical networks equipment manufacturers and the wider research and development society. In this research work, we propose a novel multilayer routing scheme for an IP/MPLS-based optical Flex-grid network, which is based on the protocol suite defined by the GMPLS architecture. Furthermore, we evaluate its behaviour in an online dynamic traffic scenario.

On solving the capacitated routing and spectrum allocation problem for flexgrid optical networks

The elastic optical networks, a.k.a., flexgrid networks, have flexible spectrum allocation. Such flexibility is possible due to the division of the spectrum into narrow frequency slots and the use of bandwidth-variable transponders. This enables the demand, in terms of band request, to be contiguously allocated starting from any available slot of the spectrum. In this context, the RSA (Routing and Spectrum Allocation) was proposed as an offline optimization problem of allocating the optical transmission resources in order to meet a predefined set of demands. This work deals with the capacitated version of the RSA problem. To solve it, we develop a multi-commodity flow formulation based on ILP (Integer Linear Programming) and a load balancing procedure that generate promising paths for an ILP formulation from the literature. Extensive computational experiments show that the proposed solution approaches are capable of obtaining high quality solutions on a variety of scenarios.

Effects of all-optical wavelength conversion on signal quality under variable-channel spacing in elastic optical networks

All-optical wavelength converters (AOWCs) based on parametric amplification in highly nonlinear fibers are investigated and evaluated, aiming at application flex-grid optical networks. A multichannel transmission system with a multilevel modulation format at 56-Gb / s bit rate is adopted. Parameters such as symbol error rate and signal bandwidth allocation are evaluated considering the optical signal-to-noise ratio, the power reception level, the channel spacing, and the impact of several cascade conversions—series of multiple AOWCs—on the signal quality. This work was carried out with an analytical modeling and computer simulations. Results reveal consistent modeling and adequate operation of the proposed AOWCs for flex-grid wavelength-division multiplexing networks and indicate that the number of conversions is limited by the size of the converter operating band and the allowed flex-grid slot size.

Gradual Migration of Co-Existing Fixed/Flexible Optical Networks for Cloud-Fog Computing

Fog computing emerges as a great candidate to mitigate the unsolved problems of cloud computing. Data migration between fog nodes and datacenters is placing huge bandwidth requirements on substrate optical networks. To increase network capacity and to improve network flexibility, broad attention has been given to flexible-grid technology. This paper addresses the issue of how to migrate substrate optical networks from the fixed-grid network era to the flexible-grid network era gradually, due to the unbearable capital expenditure and service stability. First, we model the network migration and elucidate the upgrade probability by calculating the node demands for upgrading. Then, based on the upgrade probability, three schemes are proposed to construct Potential Upgrade Nodes Group (PUNG). Besides, to maintain the traffic stability, a Migration-aware Service Provisioning (MSP) scheme is proposed based on PUNG. Numerical results illustrate that the proposed MSP scheme can effectively enhance the traffic stability compared to Non-migration-aware Service Provisioning (NSP) scheme, and the connection interrupted ratio is excessively reduced.

Virtualization of disaggregated optical networks with open data models in support of network slicing

Network slicing has been a major selling point of 5G networks, where a slice is roughly defined as a self-contained logical network on top of a shared infrastructure tailored for a specific service or vertical industry, composed of heterogeneous resources (computing, storage, bandwidths) and typically involving specific application and/or networking functions. The relationship between the network slicing and the underlying network virtualization is an open research topic, allowing multiple deployment models. In particular, a specific model of interest involves the virtualization of the optical infrastructure, where optical virtual networks are instantiated in support of network slicing, so the virtual network underlies and provides connectivity to component functions of a network slice. In this paper, we propose and implement a network virtualization architecture for open optical (partially) disaggregated networks, based on the concept of a device hypervisor relying on OpenROADM data models, in support of 5G network slicing over interconnected network function virtualization infrastructure points of presence (NFVI-PoPs). The architecture is experimentally validated, showing the provisioning of ITU-T flexigrid network media channels across a virtualized network.

Latency-Aware Resource Orchestration in SDN-Based Packet Over Optical Flexi-Grid Transport Networks

In the upcoming 5G networks and following the emerging software-defined networking/network function virtualization (SDN/NFV) paradigm, demanded services will be composed of a number of virtual network functions that may be spread across the whole transport infrastructure and allocated in distributed data centers (DCs). These services will impose stringent requirements such as bandwidth and end-to-end latency that the transport network will need to fulfill. In this paper, we present an orchestration system devised to select and allocate virtual resources in distributed DCs connected through a multi-layer (packet over flexi-grid optical) network. Three different on-line orchestration algorithms are conceived to accommodate the incoming requests by satisfying computing, bandwidth, and end-to-end latency constraints, setting up multi-layer connections. We addressed end-to-end latency requirements by considering both network (due to propagation delay) and processing delay components. The proposed algorithms have been extensively evaluated and assessed (via a number of figures of merit) through experimental tests carried out in a packet over optical flexi-grid network available in the ADRENALINE testbed with emulated DCs connected to it.

Performance of flexi-grid networks with dynamic and contiguous resource allocation

Abstract Flexi-grid networks with orthogonal frequency-division multiplexing has recently been proposed in optical networks, which can provide flexible capacity granularity. We investigate the network performance under dynamic subcarrier allocation methods. The performance measurements, including blocking probability and average link efficiency, are considered by simulations. The results demonstrate the performance profiles of three scenarios, and show that the performance of the network with first fit method is better than that with next fit one in a single fiber link; the traffic requests in one connection can influence the performance of related connections; the blocking probability of connection with multiple alternative paths is smaller than that of connection with just one path.

Robust Regenerator Allocation in Nonlinear Flexible-Grid Optical Networks With Time-Varying Data Rates

Predeployment of regenerators in a selected subset of network nodes allows service providers to achieve rapid provisioning of traffic demands, high utilization, and reduced network operational costs, while still guaranteeing lightpath quality of transmission. Enabled by bandwidth-variable transceivers in flexible-grid optical networks, optical channel bandwidths are no longer fixed but constantly changing according to real-time communication requirements. Consequently, the data-rate-variable traffic together with other new network features introduced by flexible-grid networks will render the regenerator allocation very difficult due to the complicated network states. In this paper, we investigate how to allocate regenerators robustly in flexible-grid optical networks to combat physical-layer impairments when the data rates of traffic demands are random variables. The Gaussian noise model and a modified statistical network assessment process framework are used to characterize the probabilistic distributions of physical-layer impairments for each demand, based on which a heuristic algorithm is proposed to select a set of regenerator sites with minimum blocking probabilities. Our method achieves the same blocking probabilities with on average 10% less regenerator sites compared with the greedy constrained-routing regenerator allocation method, and obtains blocking probabilities two orders of magnitude lower than that of the routing and reach method with the same number of regenerator sites.

On the Efficiency of Survivable Flex-Grid SDM Networks

In this paper, we focus on the performance of a survivable flex-grid network with space division multiplexing (SDM). To provide network survivability, we use dedicated path protection and shared backup path protection. For both schemes, we also consider squeezed protection that allocates backup resources only for a part of the primary flow. The squeezed protection is considered for the protection level changing from 0% (no flows are protected) up to 100% (full network protection) with a step of 10%. Moreover, we consider two SDM switching policies—independent ( SDM-indsw ) and joint ( SDM-jsw ). In this paper, we define all considered problem versions using integer linear programming technique as well as propose a dedicated heuristic algorithm. Then, we perform extensive simulations focused on the performance of a survivable SDM network in terms of the spectrum usage. The results show that the path protection in an SDM network brings additional spectrum consumption; however, the amount of additional resources depends on various factors such as required protection level, applied protection method, switching policy, and traffic pattern. Moreover, the results also reveal that SDM-indsw is much more spectrum-efficient than SDM-jsw considering all protection methods and levels. It uses, on average, half of the spectrum required for SDM-jsw . Additionally, the investigation shows that SDM-jsw is especially spectrum-inefficient when realizing variable-size demands (sets including both small and large demands).

Investigation of cost, power, and spectral efficiency in fixed- and flexi-grid networks

With the steady growth of traffic volume in core networks, it is predicted that future optical network communication will be constrained mainly by cost and power consumption. Thus, for Internet sustainability, it will be necessary to ensure cost and power efficiency in optical networks. The aims of this study are (i) to identify the main sources of cost and power consumption in fixed-grid (SLR and MLR) and flexi-grid (OFDM) optical networks, and (ii) to compare techniques for improving cost and power efficiency in SLR/MLR- and OFDM-based networks. To this end, we conducted a comparative analysis of cost and power efficiency for the OFDM- and MLR/SLR-based networks, and considering realistic networks, evaluated the cost and power consumed by different components in the optical layer. Our results show that (i) OFDM-based networks outperform those based on MLR/SLR in terms of both cost and power-efficiency, (ii) the extra equipment cost incurred due to under-utilization of spectrum is reduced by switching to a flexi-grid network, (iii) lower power consumption per bit is obtained when the networking solution ensures a finer bit-rate granularity, and (iv) there exists a power and spectrum minimization trade-off that is network characteristic dependent.

YANG Models for Vendor-Neutral Optical Networks, Reconfigurable through State Machine

Multi-vendor interoperability can be achieved at node and network levels by relying on standard data modeling. YANG represents an attractive data modeling solution for network component definition. This article reports on the work done on YANG models for optical networks with particular reference to flexible-grid networks. In addition to a YANG model description for link, node, and media channels, YANG for a sliceable transponder is introduced given the importance of such a data plane device for the next generation backbone. Then a contribution is provided in proposing YANG models for events and state machine to further extend and increase the programmability of networks. This latter contribution is particularly relevant in the case of faults or physical layer degradation in a network. Finally, YANG models are validated in an experimental control plane testbed.

Real-Time ROADM Filtering Penalty Characterization and Generalized Precompensation for Flexible Grid Networks

Concatenated filtering is an important impairment in high-spectral density transmission systems enabled by flexible-grid reconfigurable optical add-drop multiplexers (ROADMs). In this paper, an analytical model for wavelength selective switches (WSS) modules is used to determine the statistical distribution of experimentally measured filter shapes and to obtain the cascaded filter characteristics. A generalized approach for filtering penalty assessment is proposed and verified using a commercial 200G 16-quadrature amplitude modulation (16QAM) transponder and WSS modules. A practical method for filter penalty reduction is described, whereby a single, optimized digital pre-equalization filter is used to compensate filtering effects of multiple concatenated WSSs. This pre-equalization of cascaded ROADM filters is demonstrated with a 100G quadrature phase shift keying (QPSK) C-form factor pluggable (CFP) transponder operating on the 37.5-GHz grid, resulting in a doubling of the number of WSSs, which can be supported per link from three to six.