Aspects of the present disclosure describe systems, methods and structures for optical fiber nonlinearity compensation using neural networks that advantageously employ machine learning (ML) algorithms for nonlinearity compensation (NLC) that advantageously provide a system-agnostic model independent of link parameters, and yet still achieve a similar or better performance at a lower complexity as compared with prior-art methods. Systems, methods, and structures according to aspects of the present disclosure include a data-driven model using the neural network (NN) to predict received signal nonlinearity without prior knowledge of the link parameters. Operationally, the NN is provided with intra-channel cross-phase modulation (IXPM) and intra-channel four-wave mixing (IFWM) triplets that advantageously provide a more direct pathway to underlying nonlinear interactions.

Aspects of the present disclosure describe systems, methods and structures for optical fiber nonlinearity compensation using neural networks that advantageously employ machine learning (ML) algorithms for nonlinearity compensation (NLC) that advantageously provide a system-agnostic model independent of link parameters, and yet still achieve a similar or better performance at a lower complexity as compared with prior-art methods. Systems, methods, and structures according to aspects of the present disclosure include a data-driven model using the neural network (NN) to predict received signal nonlinearity without prior knowledge of the link parameters. Operationally, the NN is provided with intra-channel cross-phase modulation (IXPM) and intra-channel four-wave mixing (IFWM) triplets that advantageously provide a more direct pathway to underlying nonlinear interactions.

Aspects of the present disclosure describe improved supercontinuum generation based upon alternating optical dispersion along a waveguide length that advantageously generates much more spectral bandwidth than possible with conventional, prior art techniques without losing coherence as well as supporting a larger range of pulse energies (i.e., for lower than conventionally allowed pulse energies or high pulse energies).

Aspects of the present disclosure describe systems, methods and structures for optical fiber nonlinearity compensation using neural networks that advantageously employ machine learning (ML) algorithms for nonlinearity compensation (NLC) that advantageously provide a system-agnostic model independent of link parameters, and yet still achieve a similar or better performance at a lower complexity as compared with prior-art methods. Systems, methods, and structures according to aspects of the present disclosure include a data-driven model using the neural network (NN) to predict received signal nonlinearity without prior knowledge of the link parameters. Operationally, the NN is provided with intra-channel cross-phase modulation (IXPM) and intra-channel four-wave mixing (IFWM) triplets that advantageously provide a more direct pathway to underlying nonlinear interactions.

Aspects of the present disclosure describe systems, methods and structures for optical fiber nonlinearity compensation using neural networks that advantageously employ machine learning (ML) algorithms for nonlinearity compensation (NLC) that advantageously provide a system-agnostic model independent of link parameters, and yet still achieve a similar or better performance at a lower complexity as compared with prior-art methods. Systems, methods, and structures according to aspects of the present disclosure include a data-driven model using the neural network (NN) to predict received signal nonlinearity without prior knowledge of the link parameters. Operationally, the NN is provided with intra-channel cross-phase modulation (IXPM) and intra-channel four-wave mixing (IFWM) triplets that advantageously provide a more direct pathway to underlying nonlinear interactions.

Methods and systems for optical signal transmission, particularly with carrier-less amplitude and phase (CAP) modulation and direct detection, are disclosed. In one exemplary aspect, a method of optical signal transmission is disclosed. The method includes receiving information bits at an input interface; mapping the information bits to a plurality of modulation symbols; separating in-phase (I) and quadrature (Q) components of the plurality of modulation symbols such that the I and Q components form a Hilbert pair in a resulting signal; pre-dispersing the resulting signal with an inverse of a phase delay of an expected chromatic dispersion to obtain a pre-dispersed signal; converting the pre-dispersed signal from digital domain to analog domain using a digital to analog conversion circuit; performing modulation of an output of the digital to analog conversion circuit to generate an output signal; and transmitting, over an optical transmission medium, the output signal from the modulation.

A method (10) of non-linear propagation impairment equalization, the method comprising the steps of: a. receiving (12) communications traffic carried by an optical communications signal transmitted over an optical communications link; b. generating (14) a time dependent filter representation of a nonlinear time-variant impulse response of the inverse of the optical communications link; and c. applying (16) the time dependent filter representation to the received communications traffic to form non-linear propagation impairment equalized communications traffic. An optical communications link nonlinear propagation impairment equalizer and optical communications signal receiver apparatus are also provided.

A method (10) of non-linear propagation impairment equalization, the method comprising the steps of: a. receiving (12) communications traffic carried by an optical communications signal transmitted over an optical communications link; b. generating (14) a time dependent filter representation of a nonlinear time-variant impulse response of the inverse of the optical communications link; and c. applying (16) the time dependent filter representation to the received communications traffic to form non-linear propagation impairment equalized communications traffic. An optical communications link nonlinear propagation impairment equalizer and optical communications signal receiver apparatus are also provided.

Methods and systems are provided for wavelength shift elimination during spectral inversion in optical networks. The method includes receiving an input optical signal, and generating a combined optical signal by combining, by Bragg scattering, the input optical signal having an input wavelength with a first pump signal having a first wavelength. The method further includes converting the combined optical signal into an output optical signal, by phase-conjugation, using a second pump signal having a second wavelength. The output optical signal has the same wavelength as the input optical signal.

A communication apparatus for correcting a situation of a spectrum inverted signal includes a channel estimation module and an equalization module. The channel estimation module determines a channel estimation parameter, and receives at least one frame signal to generate a convolution restored frame signal corresponding to the frame signal. The equalization module includes a first computation circuit and a second computation circuit. The first computation circuit receives the channel estimation parameter and the convolution restored frame signal to generate a transformation channel estimation parameter and a transformed convolution restored frame signal. The second computation circuit receives the transformed channel estimation parameter and the transformed convolution restored frame signal to generate an original frame signal corresponding to the frame signal. The first computation circuit further feeds back a transient original frame signal to the channel estimation module to update the channel estimation parameter.