Federico Pederzolli

Comparison of Spectral and Spatial Super-Channel Allocation Schemes for SDM Networks

We evaluate the advantages of using the extra dimension introduced by space-division multiplexing (SDM) for dynamic bandwidth-allocation purposes in a flexible optical network. In that respect, we aim to compare spectral and spatial super-channel (Sp-Ch) allocation policies in an SDM network based on bundles of SMFs (to eliminate coupling between spatial dimensions from the study) and to investigate the role of modulation format selection in the blocking probability performance with an emphasis on the spectral efficiency (SE)/reach tradeoff for different multiline-rate scenarios, created either by changing the number of sub-channels (Sb-Ch), or by employing different modulation formats. Our network-performance results show that DP-8QAM -in a multichannel (MC) single-modulation-format system assuming ITU-T 50-GHz WDM Sb-Ch spectrum occupation-offers the best compromise between SE and optical reach for both spectral and spatial Sp-Ch allocation policies. They also reveal that an MC multimodulation-format system improves the network performance, particularly for spectral Sp-Ch allocation with Sb-Ch spectrum occupation of 37.5 GHz on the 12.5-GHz grid. Additionally, as another important contribution of the paper, we investigate, for spatial Sp-Ch allocation, the performance of several SDM switching options: independent switching (InS), which offers highest flexibility, joint-switching (JoS), which routes all spatial modes as a single entity, and fractional-joint switching, which separates out the spatial modes into sub-sets of spatial modes which are routed independently. JoS is proved to offer a similar performance to that of InS for particular network load profiles, while allowing a significant reduction in the number of wavelength-selective switches.

Resource allocation policies in SDM optical networks (Invited paper)

The capacity benefits of flexi-grid optical networks are limited by the finite amount of bandwidth of standard fibers, and will not be able to scale indefinitely. Space Division Multiplexing (SDM) has emerged as a promising technology to overcome this limitation. In this preliminary work we give an overview of current SDM technologies, covering coupled and non-coupled SDM networks, and of the possibilities and limitations they bring to the problem of Routing, Space and Spectrum Allocation (RSSA). We focus on the trade-offs between spectral efficiency and amount of transmission devices enabled by SDM technologies, presenting a number of preliminary heuristic policies to solve RSSA, and evaluating them in the context of dynamic traffic by means of simulations. We show that different heuristics can optimize different aspects of RSSA, and that one strikes a reasonable balance.

Improving performance of spatially joint- switched space division multiplexing optical networks via spatial group sharing

Space division multiplexing (SDM) is a promising approach to overcome the looming fiber capacity crunch. Joint and fractional joint switching (JoS and FJoS) architectures, which switch multiple spatial dimensions together as atomic units, have been proposed to reduce the cost of implementing SDM and cater to strongly coupled transmission media. However, they have been shown to impose significant penalties to overall network performance, at least when many relatively small connections are used in the context of uncoupled spatial dimensions, while performing nearly as well as the independent switching (InS) architecture if the average connection size is sufficiently large to efficiently exploit their increased granularity. In this work we tackle the performance penalty that JoS and FJoS architectures impose on SDM networks with uncoupled spatial dimensions compared to InS, by proposing two novel routing, space, and spectrum allocation (RSSA) heuristics that effectively share jointly switched resources, adapted to use different types of SDM reconfigurable optical add/drop multiplexer architectures requiring only a small additional cost. We investigate the performance of these new RSSA algorithms compared to an existing one under several traffic conditions and models using simulations. We find that our proposals can achieve significant improvements in the performance of SDM networks carrying large numbers of relatively small connections, up to parity of performance with InS designs, at a lower implementation cost, under all traffic conditions under study, including different average connection sizes and size diversity.

Spatial group sharing for SDM optical networks with Joint Switching

Space Division Multiplexing (SDM) is a promising approach to overcome the looming fiber capacity crunch. Joint and Fractional Joint Switching architectures, which switch multiple spatial dimensions together as atomic units, have been proposed to reduce the cost of implementing SDM and cater to strongly coupled transmission media. However, they have been shown to impose significant penalties to overall network performance, at least when many relatively small connections are used in the context of uncoupled spatial dimensions. With respect to this latter type of SDM networks, in this work we tackle the performance penalty that Joint and Fractional Joint Switching architectures impose on such a system compared to the Independent Switching Architecture, by proposing two novel Routing, Space and Spectrum Allocation heuristics that effectively share jointly-switched resources, adapted to use different types of SDM Reconfigurable Optical Add/Drop Multiplexer (ROADM) architectures requiring only a small additional cost. We find that our proposals can achieve significant improvements in the performance of SDM networks carrying large number of relatively small connections, up to parity of performance with Independent Switching designs, at a lower implementation cost.

Energy saving through traffic profiling and prediction in self-optimizing optical networks

A method that automatically learns and predicts the traffic behavior to save energy by adjusting the number of active optical carriers is presented. Simulations prove it provides large savings and ensures low traffic loss probability.

SDN application-centric orchestration for multi-layer transport networks

Modern IP/Optical transport networks are seldom jointly operated and optimized, and do not cater to the usually implicit requirements of applications, which ultimately drive network traffic. In this concept paper we propose a Software Defined Networking (SDN) based Network Orchestrator to manage multi-layer transport networks while taking explicit application requirements into account. We discuss its architecture and requirements, an interface to allow applications to explicitly specify their requirements in a network-agnostic manner, and possible strategies to optimize the network taking these requirements into account.

The role of SDN in application centric IP and optical networks

Transport IP/optical networks are evolving in capacity and dynamicity configuration. This evolution gives little to no attention to the specific needs of applications, beyond raw capacity. The ACINO concept is based on facilitating applications to explicitly specify requirements for requested services in terms of high-level (technology agnostic) requirements such as maximum latency or reliability. These requirements are described using intents and certain primitives which facilitate translation to technology specific configuration within the ACINO infrastructure. To support this application centric approach, SDN must have a key role in this evolution. There are representative use cases where SDN gives an added value when considering not only the network but also the application layer.

Energy Saving Through Traffic Profiling in Self-Optimizing Optical Networks

An increasing fraction of the electrical energy produced in western countries is being consumed by Internet infrastructure; reducing its energy footprint is therefore of utmost importance for the scalability of the Internet. We address optical transport backbones and propose a novel method to reduce the energy consumed by dynamically adjusting the number of active optical carriers to support the short-term load of the network with a small and controllable margin. This is achieved in a nondisruptive manner that does not interact with routing strategies and does not rely on any specific control plane, but exploits automated traffic profiling and prediction of the well-known circadian traffic cycle. The proposed approach works with both fixed and flexible grid optical networks. We describe a method to automatically learn these patterns and multiple techniques to predict incoming traffic. Furthermore, we present an algorithm that tunes the parameters of the proposed system in order to achieve a target a posteriori probability of causing traffic losses. The behavior of the system is studied, using simulations, under a variety of conditions. Results show that the proposed prediction algorithms can significantly reduce the number of active optical carriers, even in nonoptimal scenarios, while guaranteeing low traffic losses.

Proactive restoration of slow-failures in optical networks

Current optical networks, while offering outstanding reliability, still suffer from occasional failures. A resource-efficient procedure to handle these failures in un-protected scenarios is to perform restoration, i.e., to dynamically setup a backup lightpath after the primary one stops working, which leads to traffic losses while such operation completes. In this paper we propose a technique, applicable to optical networks with centralized control, to better handle failures with slow transients. The idea is to proactively perform the backup lightpaths setup, triggered by either a fixed or an adaptive threshold. The latter is chosen so as to balance the need of offsetting the setup time with the need of preventing unnecessary setups. Simulations show that the adaptive threshold provides better performance than the fixed one, in terms of both timely restorations and unnecessary setup operations.

Spectrally-Spatially Flexible Optical Networking

In this paper we review the recent advancements of spectrally spatially flexible optical network design and planning. We discuss benefits/drawbacks of different approaches. Finally, opening areas of further research are presented.