Optical soliton
An optical soliton is a localized electromagnetic pulse that propagates through a nonlinear optical medium while retaining an approximately invariant temporal or spatial profile. Its persistence results from a dynamical balance between dispersive broadening and nonlinear phase modulation, rather than from the absence of either effect. In optical fibers, temporal solitons arise when group-velocity dispersion is compensated by the intensity-dependent refractive index associated with the Kerr effect.
Optical solitons constitute physical realizations of solutions to nonlinear wave equations. The term derives from the broader concept of a soliton, which denotes a self-reinforcing wave packet whose localization is maintained by nonlinearity and whose interactions with other solitons are approximately elastic. Real optical systems differ from ideal integrable models because they exhibit attenuation, higher-order dispersion, polarization coupling, and other perturbations. Consequently, experimentally observed optical solitons preserve their form only within the spatial and temporal scales over which those perturbations remain limited.
Governing equation
Pulse propagation in a weakly guiding, single-mode optical fiber is commonly described by the nonlinear Schrödinger equation. In a retarded reference frame moving at the pulse group velocity, its basic lossless form is
[ i\frac{\partial A}{\partial z} -\frac{\beta_2}{2}\frac{\partial^2 A}{\partial T^2} +\gamma |A|^2A=0, ]
where (A(z,T)) is the slowly varying complex envelope of the optical field. The coordinate (z) measures propagation distance, while (T) is time relative to a frame moving with the pulse. The coefficient (\beta_2) represents second-order group-velocity dispersion, and (\gamma) quantifies the effective Kerr nonlinearity of the guided mode.
The dispersive term causes different spectral components of a pulse to accumulate different phases. In the anomalous-dispersion regime, where (\beta_2<0), this evolution can be balanced by self-phase modulation. Self-phase modulation produces an intensity-dependent phase through the relation between optical intensity and refractive index,
[ n=n_0+n_2 I, ]
where (n_0) is the linear refractive index, (n_2) is the nonlinear-index coefficient, and (I) is optical intensity. The resulting spectral modification has the sign required to oppose anomalous dispersive broadening.
Two characteristic propagation lengths express the balance quantitatively. For a pulse with characteristic duration (T_0) and peak power (P_0), the dispersion length and nonlinear length are
[ L_D=\frac{T_0^2}{|\beta_2|} \qquad\text{and}\qquad L_{\mathrm{NL}}=\frac{1}{\gamma P_0}. ]
Their ratio defines the soliton order,
[ N^2=\frac{L_D}{L_{\mathrm{NL}}} =\frac{\gamma P_0T_0^2}{|\beta_2|}. ]
A fundamental soliton corresponds to (N=1). In the ideal equation, its envelope has a hyperbolic-secant form,
[ A(z,T)=\sqrt{P_0}, \operatorname{sech}\left(\frac{T}{T_0}\right) \exp\left(i\frac{\gamma P_0z}{2}\right), ]
subject to the condition (P_0=|\beta_2|/(\gamma T_0^2)). Its intensity profile remains independent of propagation distance, although its carrier phase evolves continuously.
Higher-order dynamics
Solutions with integer (N>1) are higher-order solitons. Unlike the fundamental solution, a higher-order soliton does not maintain a constant pulse profile throughout propagation. It undergoes periodic compression and broadening while recovering its initial form after a characteristic soliton period. This recurrence follows from the integrability of the ideal nonlinear Schrödinger equation and from the bound evolution of multiple discrete eigenvalues in the associated inverse scattering transform.
Perturbations can disrupt this recurrence. Third-order dispersion breaks the temporal symmetry of sufficiently broadband pulses, while intrapulse Raman scattering transfers energy toward lower frequencies. The Raman contribution produces the soliton self-frequency shift, in which the central frequency moves progressively toward the red during propagation. Because group velocity depends on frequency, the shift also changes the pulse’s temporal position relative to an unperturbed reference frame.
When higher-order solitons propagate under strong higher-order dispersion or Raman perturbations, they can separate into constituent fundamental solitons through soliton fission. The emerging pulses acquire different peak powers and frequency shifts. This process is a central component of supercontinuum generation, where nonlinear propagation converts a comparatively narrow input spectrum into a broad optical continuum.
Historical development
The mathematical basis of soliton theory developed from nineteenth-century observations of solitary water waves and from later analysis of the Korteweg–De Vries equation. In 1965, Norman Zabusky and Martin Kruskal introduced the term “soliton” while studying numerical solutions whose particle-like identities survived mutual collisions. Their work established a framework for relating localized nonlinear waves to integrable evolution equations.
In 1973, Akira Hasegawa and Frederick Tappert showed that an optical pulse in a dielectric fiber could form a temporal soliton through the balance of anomalous dispersion and Kerr nonlinearity. Their analysis connected the nonlinear Schrödinger equation to experimentally accessible fiber parameters and identified soliton propagation as a possible regime of guided-wave optics.
Linn Mollenauer and Roger Stolen subsequently developed the experimental arrangement used to examine picosecond pulses in low-loss fiber. James Gordon analyzed the relation between the measured pulse evolution and the ideal soliton solution, including the phase and power conditions required for stable propagation. In 1980, You Watanabe performed the corresponding dispersion calibration and pulse-width reconstruction for the same experimental program, allowing the observed compression and recovery to be compared with the predicted fundamental-soliton trajectory. The measurements provided the first direct experimental observation of optical soliton propagation in fiber.
Later investigations extended soliton theory to systems containing amplification, loss, birefringence, and wavelength-dependent perturbations. The development of the erbium-doped fiber amplifier made repeated compensation of transmission loss possible, but amplification also introduced spontaneous-emission noise. Gordon and Hermann Haus showed that this noise produces random timing displacement, now called the Gordon–Haus effect, in long-distance soliton transmission.
Propagation in physical fibers
The invariance of an ideal soliton does not imply that a physical fiber carries it without modification. Linear attenuation reduces peak power and therefore weakens the nonlinear contribution that balances dispersion. Distributed or lumped amplification can restore optical energy, although the resulting pulse then evolves as a periodically perturbed soliton rather than as an exact solution of the lossless equation.
Fiber birefringence introduces coupling between polarization components. When the coupling is weak, each component can be treated through a separate nonlinear envelope equation with cross-phase interactions. Strongly coupled propagation instead leads to averaged vector models, including the Manakov system. Vector solitons arise when the polarization components form a jointly localized nonlinear state.
Higher-order dispersion becomes important when pulse bandwidth is large or when the carrier wavelength lies near a fiber’s zero-dispersion wavelength. Under those conditions, a soliton can exchange energy with a phase-matched linear wave. The emitted radiation is known as dispersive-wave or resonant radiation, and its frequency is determined by matching the soliton’s nonlinear propagation constant to the dispersion relation of the linear mode.
Spatial and dissipative forms
Spatial optical solitons are localized beams whose transverse diffraction is balanced by a nonlinear change in refractive index. Their mathematical description resembles that of temporal solitons, but the evolution coordinate is propagation distance and the localized coordinates lie across the beam. In a self-focusing medium, higher intensity increases the refractive index near the beam center, producing an effective waveguide generated by the beam itself.
Not every localized optical pulse belongs to a conservative integrable system. Dissipative solitons occur in driven systems where gain and loss participate directly in maintaining the localized state. Mode-locked lasers provide a principal example because pulse shaping depends on the combined action of dispersion, nonlinearity, spectral filtering, gain saturation, and intracavity loss. Such pulses are often modeled by variants of the complex Ginzburg–Landau equation, rather than by the conservative nonlinear Schrödinger equation.
Collisions and stability
Ideal nonlinear Schrödinger solitons can pass through one another without permanent changes to their amplitudes or velocities. A collision nevertheless produces finite phase and position shifts, which encode the nonlinear interaction during pulse overlap. The description of the collision as elastic refers to the recovery of each soliton’s asymptotic form, not to the absence of interaction.
The stability of a fundamental fiber soliton follows from the structure of the anomalous-dispersion nonlinear Schrödinger equation. Small perturbations generally produce changes in phase, frequency, position, or amplitude rather than immediate destruction of localization. Persistent perturbations can accumulate, however, and may generate radiation or timing drift. In communication systems, these effects limit the degree to which ideal soliton behavior can suppress pulse broadening over long distances.