The global growth of the Internet and multimedia applications with its corresponding great impact on educational, commercial and social activities, leads to a constant increase in the amount of information that needs to be transmitted, stored and retrieved. Nowadays, there is a serious disproportion in global information traffic between the high speed at which tremendous amounts of digital data travel in ultra-long haul fibre-optic links connecting different continents and the comparatively lower rates at which information is processed electronically, either at the network nodes or at the end-user's personal computer. It has been well recognized for many years that, notwithstanding tremendous developments in electronics, ultimately only photonics has the required processing speed to satisfy future demands. However, in spite of the heroic efforts of the past decades, all-optical signal processing technologies capable of resolving the bottleneck imposed by electronic components are still to come. It is likely that a solution to this problem cannot be found through an incremental development of existing technologies and truly novel breakthrough in ideas and methodologies are required to achieve real progress in all-optical switching, processing and computing. A change of paradigm requires the combined efforts of a broad spectrum of specialists from mathematics to material science. One of the fundamental concepts of nonlinear photonics is to use solitons -stable optical pulses resulting from the interplay between material nonlinearity and dispersion- as the information carriers (elementary bits) to transmit and process digital signals at very high speeds. New mathematical nonlinear theories have emerged recently generalizing the basic soliton paradigm to systems with gain and dissipation. The aim of this project is to introduce fundamentally new nonlinear approaches into the realm of digital optics and to develop the concept of dissipative solitons for all-optical digital signal processing. The present approaches used to implement all-optical data processing in communication systems are mainly drawn on conservative (based on the energy conservation, at least on average) nonlinear processes. Since their stationary solutions typically form families, meaning that for a particular system the asymptotic state depends on the input parameters, conservative systems exhibit a continuous family of output states, which is fundamentally different from what one needs for digital signal processing: a discrete number (two in a binary coding) of possible output states. A theoretical concept that might help to resolve this fundamental issue is the engineering of systems with stable attractors, a dynamical feature that can be realized in dissipative nonlinear systems when signal attenuation is balanced by gain. In such systems, the soliton family degenerates into a single isolated solution, completely determined by the preset system parameters. In this project, we propose to develop the concept of nonlinear optical devices based on dissipative solitons. Such devices will exploit the fundamental properties of dissipative solitons to achieve truly digital signal processing functions. Ultimately our research is aimed at providing fundamental information on the feasibility of using dissipative solitons in all-optical digital data processing and exploring the potential of this concept for optical computing. We concentrate our research efforts on two photonic media where the dissipative soliton concept can be implemented: active fibre systems and a novel artificial nonlinear medium (dissipative photonic crystal) formed by periodic patterns of semiconductor optical amplifiers and saturable absorbers. To foster the rapid emergence of new paradigms and nonlinear photonic devices using dissipative solitons, it is important to transfer fundamental theories of nonlinear science into the field of optical engineering, and this is the main target of this project.