Optical Networks & Systems (ONS)
Call ID: TEC2012-38119
Global IP traffic has increased eightfold over the past five years, and will increase fourfold over the next five due to the accelerated proliferation of users, not only with broadband Internet access at home (e.g., VDSL or FTTH) but also with mobile broadband (smartphones and tablets). Moreover, the traffic behaviour is suffering a complete change in its pattern. For the first time, Internet video will reach the 50% of total consumer IP traffic by the end of the current year (2012) and 90% by end 2015 (end of this project). The main implication is that the Internet traffic is evolving from a relatively steady and symmetric stream profile (dominant peer-to-peer/file sharing traffic) to a more dynamic traffic with a highly asymmetric profile. Therefore, future optical transport networks will be required to provide high, flexible, elastic and adaptive bandwidth connectivity between IP routers, based on the traffic demands and network conditions. Current ITU-T DWDM wavelength grid with fixed channel spacing is particularly inefficient for large granularities (e.g., 10 Gb/s and 100 Gb/s) since a whole wavelength is assigned to a lower rate optical path (e.g.,10 Gb/s) that does not fill the entire wavelength capacity. This situation has traditionally forced to aggregate and groom low-bit-rate data flows with electrical TDM crossconnects (e.g. SONET/SDH) or, more recently, through electrical packet switching technology (e.g., MPLS), since no alternative in the optical domain (e.g. all-optical packet switching) was commercially feasible. However, a new architecture introducing elasticity and adaptation in optical networks has been recently proposed to provide flexible, highly-efficient and adaptive optical spectrum management, attaining granular grooming in the optical domain. In elastic optical networks (i.e., with variable channel spacing), the optical spectrum is partitioned into basic fixed-size spectrum slots (e.g., 6.25 GHz or 12.5 GHz). The required spectral resources are dynamically and adaptively allocated by assigning the necessary number of contiguous basic fixed-size spectrum slots according to the traffic demand and the network conditions.
The key enablers of elastic optical networks are the Optical OFDM (O-OFDM) technology and the GMPLS/PCE control plane. O-OFDM is the most promising technology for the design of bandwidth-variable transponders, since it enables software-defined optical transmission (SDOT). That is, the transponder can be adapted to multiple modulation formats or variable bandwidth occupancy by means of electronic digital signal processing. Moreover, the transmission of multiple orthogonal subcarriers provides a high spectral efficiency (subcarriers are overlapped), unique flexibility and adaptive bit-rate/bandwidth (modifying number of subcarriers and modulation formats), and sub/super-wavelength granularity. A GMPLS control plane with PCE allows the automatic provisioning and recovery of flexible connections and its bandwidth modification / re-routing in real time in elastic and adaptive optical networks. The integration of these technologies is the basis to support scalable (beyond 100 Gb/s) and large data rate granularities in an energy and cost-effective manner. However, this new network architecture arises new technical challenges that must be properly investigated at both transmission and networking level by means of not only theoretical, analytical and simulation studies but also from an experimental research to capture the whole set of interdependencies between system components.
The aim of the FARO project (Integrating control and transmission technologies for flexible, elastic and spectrum-efficient optical networks) is to investigate, from an integrated approach, O-OFDM transmission and GMPLS/PCE control technologies as the key enablers for elastic optical networks, leveraging the bandwidth/bit rate variable optical transponders and an advanced control plane, for the dynamic provisioning of adaptive connectivity services with recovery, in both single and multi-domain contexts. This macroscopic objective includes the following scientific and technological ones:
- Identification of a concrete list of functional requirements, use cases and scenarios that drive the main scientific and technological work within the FARO project.
- Design and implementation of SDOT systems based on tuneable, bandwidth and bit rate variable O-OFDM transponders
- Development of resource-efficient techniques for O-OFDM transmission enhancement, including distortionless PAPR mitigation and performance monitoring with reduced overhead.
- New techniques, algorithms and procedures advancing the state of art of multilayer/RSA algorithms with electrical/optical grooming, the subsequent network optimization, and their dissemination.
- Implementation and assessment of the applicability of a GMPLS/PCE control plane for multi-layer optical networks by means of its deployment, along with the numerical (quantitative) evaluation of its performance.
- Conception of a virtual elastic optical network resource broker and service composition manager.
Experimental validation and evaluation of the integrated elastic optical network prototype.