Throughput Comparison Between 50-GHz and 375-GHz Grid Transparent Networks [Invited]

Morea, Annalisa; Renaudier, Jérémie; Zami, Thierry; Ghazisaeidi, Amirhossein; Bertran-Pardo, O.

doi:10.1364/jocn.7.00a293

Cited by 56 publications

(49 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It is assumed that linear impairments are ideally compensated at the receiver and that the only significant physical layer impairments are amplified spontaneous emission (ASE) noise from the EDFAs and nonlinear interference. The imperfections of non-ideal network components, for example crosstalk and optical filtering [29], [30], in the ROADMs along with PDL and EDFA imperfections, have not been included in this study. The ASE noise accumulates linearly with the number of spans, where the additional noise per span in the receiver matched filter bandwidth, n ASE , is given by…”

Section: A Transmission Impairment Modelmentioning

confidence: 99%

Throughput Gains From Adaptive Transceivers in Nonlinear Elastic Optical Networks

Ives

Alvarado

Savory

2017

J. Lightwave Technol.

View full text Add to dashboard Cite

Abstract-In this paper we link the throughput gains, due to transceiver adaptation, in a point-to-point transmission link to the expected gains in a mesh network. We calculate the maximum network throughput for a given topology as we vary the length scale. We show that the expected gain in network throughput due to transceiver adaptation is equivalent to the gain in a point-to-point link with a length equal to the mean length of the optical paths across the minimum network cut. We also consider upper and lower bounds on the variation of the gain in network throughput due to transceiver adaptation where integer constrained channel bandwidth assignment and quantized adaptations are considered. This bounds the variability of results that can be expected and indicates why some networks can give apparently optimistic or pessimistic results. We confirm the results of previous authors that show finer quantization steps in the adaptive control lead to an increase in throughput since the mean loss of throughput per transceiver is reduced. Finally we consider the likely network advantage of digital nonlinear mitigation and show that a significant trade off occurs between the increase in SNR for larger mitigation bandwidths and the loss of throughput when routing fewer large bandwidth superchannels.

show abstract

Section: A Transmission Impairment Modelmentioning

confidence: 99%

Throughput Gains From Adaptive Transceivers in Nonlinear Elastic Optical Networks

Ives

Alvarado

Savory

2017

J. Lightwave Technol.

View full text Add to dashboard Cite

show abstract

“…In this context, the nodes are expected to adopt a switch-and-select internal architecture with colorless and directionless A&Ds in order to prevent channel crosstalk [4]. This architecture involves the crossing of several filtering elements, even for pass-through flows, representing a challenge for the performance of flexi-WSSes.…”

Section: B Flexi-grid Elastic Optical Networkmentioning

confidence: 99%

“…Employing flexi-WSSes the optical spectrum can be used more flexibly and efficiently, being not restricted to operating within approximately 90 slots of 50 GHz. This allows, for example, a new 37.5 GHz slot size, providing an additional 33% of aggregated throughput when comparing to the legacy 50 GHz networks [4]. Moreover, the flexible frequency slot size supports the transmission of super-channels, giving a further spectral efficiency enhancement thanks to the tight spectral packing of the optical sub-channels.…”

Section: Introductionmentioning

confidence: 99%

“…Indeed, these nodes contain non-ideal optical filters and, when the optical signals traverse several of them, the resulting equivalent channel bandwidth may be significantly decreased, entailing spectral distortions and optical signal to noise (OSNR) penalty. This effect is usually referred to as filter narrowing effect, being the source of non-negligible performance degradation [4]. This penalty is more likely to be present in transparent aggregation networks (typically covering the metropolitan and regional segments) where signals are expected to pass through a high number of nodes and low symbol rate connection may force the use of the smallest available slots sizes.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the Filter Narrowing Issues in Elastic Optical Networks

Fàbrega¹,

Moreolo²,

Martin³

et al. 2016

J. Opt. Commun. Netw.

View full text Add to dashboard Cite

Abstract-This paper describes the problematic of filter narrowing effect in the context of next generation elastic optical networks. First, three possible scenarios are introduced: the transition from actual fixed-grid to a flexi-grid network; the generic full flexi-grid network; and a proposal for filterless optical network. Next, we investigate different transmission techniques and evaluate the penalty introduced by the filtering effect when considering: Nyquist WDM, SSB DD-OFDM and symbol-rate variable DP-4QAM. Also, different approaches to compensate for the filter narrowing effect are discussed. Results show that the specific needs per each scenario can be fulfilled by the aforementioned technologies and techniques, or a combination of them, when balancing performance, network reach and cost.Index Terms-Networks, optical communications, elastic optical networks, flexi-grid, WSS. I. INTRODUCTIONThe future adoption of elastic optical network (EON), mainly fostered by the advent of next technologies (e.g., media, HDTV, 5G, Internet of Things, etc.) and backed by the considerable advances of transmission techniques in terms of flexibility and capacity, is heading to undertake new challenges and goals. In fact, when adopting the flexi-grid paradigm [1], optical channels with different bandwidth occupation can coexist within the same fiber. Some of these channels, denominated as super-channels, are wider in frequency and comprise multiple sub-channels transmitted

show abstract

“…An MPLS router is co-located with every optical node and connected to it through 100 Gb/s bandwidth-variable transponders (BVT) using 37.5 GHz [25]. Note that MPLS vlinks are supported by lightpaths on the optical layer and CUVINET creates and releases them to serve the actual load.…”

Section: A Network Scenariomentioning

confidence: 99%