Search

Today’s traditional, hierarchical, fullelectronic data centers cannot guarantee quality of service (QoS) for latency-sensitive applications or carrier-grade multitenancy.We propose cloud burst optical-slot switching (CBOSS), an intra-data center network architecture leveraging the original optical components and custom software- defined network (SDN) control to enable traffic engineering. In particular, we demonstrate network slicing on demand with subwavelength granularity and deterministic per-flowQoSguarantees in a 3-top-of-rack(ToR)CBOSSnetwork prototype. We rely on an integrated silicon photonics (SiP) wavelength dropper, a fast-tunable laser, and custom SDN-based controller. [ABSTRACT FROM AUTHOR]

In this paper, we propose to lower the network design margins by improving the estimation of the signal-tonoise ratio (SNR) given by a quality of transmission (QoT) estimator, for new optical demands in a brownfield phase, based on a mathematical model of the physics of propagation. During the greenfield phase and the network operation, we collect and correlate information on the QoT input parameters, issued from the established initial demands and available almost for free from the network elements: amplifiers output power and the SNR at the coherent receiver side. Since we have some uncertainties on these input parameters of the QoT model, we use a machine learning algorithm to reduce them, improving the accuracy of the SNR estimation. With this learning process and for a European backbone network (28 nodes, 41 links), we could reduce the QoT inaccuracy by several dBs for new demands whatever the amount of uncertainties of the initial parameters. [ABSTRACT FROM PUBLISHER]

Optical slot switching has formerly been proposed as a flexible solution for metropolitan ring networks to carry data traffic with a subwavelength switching granularity and with good energy efficiency, which is enabled by optical transparency. In this paper, we propose for the first time, to the best of our knowledge, several architectures for the electronic side of optical slot switching nodes to increase flexibility through the addition of electronic switches, working either at client packet granularity or at slot granularity. Such electronic switches can be located at either the transmitter side or receiver side, or at both sides of a node, thereby decreasing traffic latency, at the expense of increased node cost and/or energy consumption. This paper focuses on the latency aspect. We investigate the impact of a timer that can be used to upper bound the slot insertion time on the medium. We also propose a novel client packet queuing model in an optical slot switching ring and assess and compare the latency of these node architectures analytically and with simulations. [ABSTRACT FROM PUBLISHER]

Due to the ever-growing network capacity demand, energy-efficient and scalable networking solutions are needed in all networking segments. This paper focuses on the metro segment and compares the power consumption for four different technologies: Ethernet, Reconfigurable Optical Add-Drop Multiplexers (ROADM), Optical Transport Networks (OTN), and an optical packet switching (OPS) solution called Packet Optical Add-Drop Multiplexers (“POADM”), which was recently proposed. A novel resource allocation strategy for POADM-based OPS multi-rate rings is proposed for the first time. The multi-rate POADM ring network resource allocation problem is shown to be NP-hard and solved using both an MILP formulation and a heuristic. The MILP formulation and the heuristic are compared, along with dimensioning results for Ethernet and ROADM/OTN ring networks. Results show that multi-rate POADM rings achieve savings up to 10% with respect to single-rate design and can be enabled by assigning a different rate to each wavelength. Furthermore, POADM rings can be up to 5 times more energy-efficient than Ethernet rings and 30% than ROADM rings, assuming similar network efficiencies. [ABSTRACT FROM PUBLISHER]

The performance of current optical networks is inherently limited by the speed of electronic components and, in particular, by electronic switches. A new generation of optical networks, referred to as all-optical networks, overcomes this limitation by switching data entirely optically using all-optical crossconnects (OXCs). However, all-optical networks are prone to phenomena that are unknown to current optical networks with electrical regeneration: OXCs are subject to optical leaks, called crosstalk, resulting in unwanted components being added to transmitted signals, and this crosstalk is transmitted over very long paths without any signal regeneration. In this paper, we consider the interplay between fiber nonlinearity and crosstalk signals over long distances as the source of performance degradation, measured in terms of Q factor. We present an analytical crosstalk model for all-optical networks and give expressions for the performance degradation resulting from the joint propagation of a signal [using a continuous-wave (CW) assumption and perturbation theory] and crosstalk in large networks. Analytical calculations required by this model are shown to be much less computationally intensive than simulations. Simulations are carried out to validate our analytical model and good agreement is found between the analytical model and simulations for wide ranges of parameters. [ABSTRACT FROM PUBLISHER]

In all-optical networks the physical layer impairments accumulate along a lightpath and also vary dynamically, and a number of impairment aware routing and wavelength assignment (IA-RWA) techniques have been proposed in order to mitigate the physical layer impairments and find lightpaths that meet a required quality of transmission (QoT) constraint predefined by the network operator. However, in order to compute lightpaths, IA-RWAs typically rely on analytical models, which cannot be guaranteed to be fully accurate, and hence acceptance of lightpaths with poor QoT or rejection of lightpaths with acceptable QoT may ensue. We present a novel IA-RWA algorithm that not only considers the impact of physical impairments on RWA decisions, but also, for the first time, accounts for inaccuracy of the QoT estimators. The performance of our algorithm is compared with algorithms selected from the recent literature. All algorithms are evaluated through simulations in a realistic scenario. Our proposed novel algorithm outperforms the selected algorithms in terms of blocking rate and also the amount of required resources for achieving a very low (i.e., 5×10−6) blocking rate under similar assumptions. In addition we show that accounting for QoT estimation inaccuracy changes the performance of the proposed IA-RWA substantially, and hence that the QoT estimator inaccuracy is an important design parameter in transparent optical networking. [ABSTRACT FROM PUBLISHER]

In transparent optical networks, signals propagate over all-optical lightpaths. The absence of regenerating devices that act in the electrical domain renders end-to-end monitoring difficult. Quality of transmission (QoT) metrics quantify the degradation in quality that a signal experiences as it traverses a lightpath. Hardware monitors that can directly measure QoT are expensive, which motivates the development of monitoring schemes that require fewer monitors but can still generate accurate QoT estimates. In this paper we describe a monitoring scheme that estimates the QoT of multiple lightpaths in a network. Our focus is on estimating bit-error-rates (BERs), but the methodology is also applicable for other metrics. One of the primary innovations in this monitoring framework is the establishment of “active lightpaths”—lightpaths that carry no useful data but are instead used as measurement probes. We describe a method for choosing where to establish the active lightpaths in order to maximize the information gain. We demonstrate with simulations the possibility to trade off the amount of costly hardware monitoring equipment with cheaper, temporary active lightpaths, while still achieving accurate monitoring. [ABSTRACT FROM PUBLISHER]