Project Detail |
Information theoretical approach for wavelength division multiplexing transmission for all users
Optical fibre systems play a critical role in global internet traffic, relying on wavelength division multiplexing (WDM) for transmission. However, non-linear effects are induced by high signal intensities. Information theory explores the fundamental limits of WDM transmission, focusing on the attainable rates for individual users. The MSCA-funded NIT-NL project proposes a multi-user information theoretical approach for the WDM channel, concurrently investigating achievable rates for all users, diverging from the single-user viewpoint. The project aims to utilise optical network resources through a realistic characterisation of the capacity region, asymptotic analysis and derivation of optimality conditions. The outcomes will benefit the scientific and information theory communities, the European telecommunications industry and standardisation initiatives.
Optical fiber systems carry nearly all of the global Internet traffic and are critical infrastructure for modern
society. Long-haul optical networks are often based on the notion of wavelength division multiplexing (WDM). At
high signal intensities, WDM systems are impaired by the inherent nonlinear effects involved in fiber
propagation. In the literature, the fundamental limits of WDM transmission are investigated using information
theory. The prevalent viewpoint only focuses on achievable rates for individual users under different behavioral
assumptions for the interfering users. NIT-NL deviates significantly from this single-user viewpoint and proposes
a multi-user information theoretic approach for the WDM channel that investigates the region of simultaneously
achievable rates for all the different users. NIT-NL aims to exploit this holistic perspective and realize a more
efficient utilization of the optical network resources. This will be achieved via a) characterization of closed form
inner and outer bounds to the capacity region based on realistic channel models, b) asymptotic analyses in the
limit of high transmit power to infer the nature of optimal strategies in the highly nonlinear regime, and c)
derivation of optimality conditions for simple and practical schemes such as treating interference as noise to
achieve the sum-capacity of the given model. NIT-NL will have a direct impact on the scientific community
engaged in the long-standing quest of dealing with nonlinear impairments in optical channels, and the multi-user
information theory community that will welcome its enrichment on nonlinear channel models. It will also benefit
the European telecommunications industry and standardization initiatives. |