Creating and sharing knowledge for telecommunications

Project: Digital Nonlinear Equalization in Very-High Speed Coherent Optical Transmission Systems

Acronym: DiNEq
Main Objective:
The exponentially growing traffic demand in optical core networks has exhausted the single-mode fiber capacity within the linear regime. We aim to develop efficient digital equalization strategies in order to counteract the signal distortion imposed by fiber nonlinearities, in very-high speed (>= 100 Gbit/s/ch) and ultra-dense (>= 4 bit/Hz) multi-wavelength coherent optical transmission systems. Building on our previous work and experience on this topic, we aim to optimize some of the currently proposed algorithms, mainly in terms of computational requirements, and to develop new solutions for nonlinear digital equalization in optical fiber systems. Our efforts include the extension of nonlinear equalization to polarization-multiplexing and WDM transmission. We will focus a great deal of our attention into the problem of inter-channel nonlinear equalization, developing efficient and moderate complexity techniques to partially mitigate cross-phase modulation (XPM) and four-wave mixing (FWM) distortions. We plan to develop simplified approaches to enable adaptive equalization of XPM resorting to either blind or data-aided equalizers. We also plan to explore artificial neural networks (ANN) techniques for fiber impairment mitigation. Given the hardware implementation advantages of ANNs due to their inherent parallel nature, we expect to obtain efficient real-time implementation solutions. During the last couple of years, the interest on few-mode fibers (FMF) and mode-division multiplexing (MDM) has exploded, with numerous publications in major international conferences and journals. This sudden revival of multi-mode signal propagation is closely related with the need to reduce the effect of fiber nonlinearities, which is thereby achieved by increasing the fiber core area. From the digital signal processing (DSP) point of view, the adoption of MDM requires an extra effort to implement multiple-input-multiple-output (MIMO) equalization of inter-modal transmission and coupling. Although signal impairments in FMFs are currently assumed to be mostly caused by linear processes, some pioneering works are already exploiting the impact of nonlinearities in FMFs. A timely development of nonlinear equalization techniques for MDM signal transmission is therefore of utmost importance. Within this topic there are still many problems to be tackled, ranging from development of modeling numerical tools for nonlinear signal propagation to the development of lower complexity digital equalization techniques. We propose to exploit the reported potentialities of the Volterra series transfer function, extending its scope to MDM transmission, therefore providing a powerful numerical tool for the analysis and compensation of inter-modal nonlinearities. The development and optimization of digital equalization algorithms must be accompanied by meaningful assessment through simple and accurate figures of merit. We propose to develop simple figures of merit for BER gain and required computational effort for nonlinear equalization algorithms under different propagation conditions. These figures of merit should guide the development of new equalization algorithms and the design of new very high-speed and ultra-dense transmission systems. Finally, we aim to finalize this project with an experimental test and validation stage, composed by two implementation phases: offline validation of the proposed algorithms recurring to experimental data, and hardware implementation in a field programmable gate array in order to test real-time operation. This implementation stage will benefit both from IT facilities and expect investment in coherent optical systems and from a well-established international cooperation with top-level European laboratories. Given the current absence of feasible solutions for the equalization of critical nonlinear fiber impairments, we will have the opportunity to generate crucial contributions to the state-of-the-art of coherent optical systems. In fact, solving the nonlinear limitation on fiber capacity has been identified as one of the key scientific breakthroughs to enable future Terabit optical channels.
Reference: PTDC/EEI-TEL/3283/2012
Funding: FCT/PTDC
Start Date: 01-07-2013
End Date: 01-06-2015
Team: Armando Humberto Moreira Nolasco Pinto, A. Teixeira, Fernando Pedro Pereira Guiomar, Gil Gonçalo Martins Fernandes, Jacklyn Dias Reis, José Rodrigues Ferreira da Rocha, Manuel Alberto Reis de Oliveira Violas, Nelson de Jesus Cordeiro Muga, Nuno Silva, Paulo Sergio de Brito Andre
Groups: Optical Communication Systems and Networking – Av
Local Coordinator: Armando Humberto Moreira Nolasco Pinto
Links: Internal Page
Associated Publications