The ever-increasing deluge in the world’s IP data traffic volume driven by the immense growth of versatile bandwidth-hungry on-line services, such as cloud computing, on-demand HD video streams, on-line business analytics and content sharing, the development of the Internet of Things, and many others, sets higher and higher requirements on the speed and quality 244characteristics of information flows interconnecting network constituents. It is currently well understood that increasing data rates in the core fibre communication systems are quickly approaching the limits of contemporary transmission technologies [1–3]. For linear channels the bounds on the transmission rate were established in works of Nyquist [4] and Shannon [5]. However, there is now a common belief that it is the inherent nonlinear transmission effects that serve as a major limiting factor in modern fibre-optic communication systems [3, 6–8]: These nonlinear properties make optical fibre channels very different from wireless and other linear ones [9, 10]. To reduce the nonlinearity impact on the transmission characteristics, there have been proposed a great number of nonlinearity compensation methods, including digital back-propagation (DBP) [11], digital [12] and optical [13, 14] phase conjugations (spectral inversion), and phase-conjugated twin waves [15], to mention a few important advances [16, 17]. Note that in the techniques itemized above, the fibre nonlinearity is treated as a “transmission degrading factor”, and so the goal of each of the aforementioned methods is to suppress its negative effect at the expense of performing some additional processing operations. With regard to the latter aspect, the nonlinear Fourier transform (NFT) based transmission can be reckoned as a conceptually different approach: Here nonlinearity enters like a constructive part of processing and transmission, defining the features of the system architecture and its characteristics. Thereby the nonlinearity can play a rather positive role in defining the performance of the NFT-based system. At the same time, the major difficulty in NFT application lies in the fact that some quite common “linear” concepts may need to be reconsidered or appended with a new non-trivial meaning. For instance, in addition to the usual notions of frequency, spectrum, power, bandwidth, etc., one has to work with their nonlinear analogs that can be drastically different from the “ordinary” counterparts. Simultaneously, these new signal parameters can serve as well-defined and adjustable characteristics of our signal [18–20], which can be used for modulation and information transmission. The overall idea of using NFT for the communications is that inside the nonlinear Fourier (NF) domain the propagation of the so-called nonlinear spectral data (that play the role of Fourier 245spectrum for a special class of nonlinear problems) is uncoupled and linear. It is exactly this remarkable property of NF spectrum that makes the NFT-based transmission potentially free from the cross-talk and bandwidth related sources of the capacity degradation, that plague most of the conventional transmission systems [6–8], as the nonlinearity is effectively included into the NFT operations.