Traveling Waves in the McKean-Vlasov Equation under Sakaguchi-Kuramoto Interaction with Phase Frustration

1. Problem Statement and Motivation

Context:
The paper addresses the dynamics of large networks of weakly coupled oscillators, a fundamental problem in statistical physics and neurobiology. While the classical Kuramoto model (with purely sinusoidal coupling, $\Gamma(\theta) = \sin \theta$) successfully describes the transition from incoherence to synchronization, it fails to capture phenomena observed in large-scale neural networks, such as traveling waves and multi-stable states.

The Gap:
Realistic neural networks exhibit time delays (due to axonal conduction and synaptic transmission) which introduce phase frustration. The Sakaguchi–Kuramoto model extends the classical model by incorporating a phase shift parameter $\alpha$ in the interaction potential: $W(\theta) = -\cos(\theta - \alpha)$.

• Challenge: Unlike the symmetric Kuramoto model ($\alpha=0$), the Sakaguchi–Kuramoto interaction is asymmetric for $\alpha \neq 0$. This breaks the symmetry required by standard bifurcation analysis tools (like Crandall–Rabinowitz theory) and renders the system non-equilibrium.
• Objective: To provide a rigorous mathematical proof of the existence and uniqueness of a global phase transition from an incoherent state to a non-equilibrium traveling wave state in the mean-field limit (McKean–Vlasov equation) for the asymmetric Sakaguchi–Kuramoto model.

2. Mathematical Framework

The Model:
The study focuses on the continuum limit of $N \to \in$ weakly coupled oscillators with additive noise. The system is described by the McKean–Vlasov equation (a nonlinear Fokker–Planck PDE) for the probability density function (PDF) $\rho(t, \theta)$:
$$\partial_t \rho = D \partial_\theta^2 \rho + \mu \partial_\theta (\rho \partial_\theta (W * \rho))$$
where $W(\theta) = -\cos(\theta - \alpha)$ and $\mu$ is the coupling strength.

Steady States:

• Equilibrium ($\alpha=0$): The system admits stationary solutions (von Mises distributions).
• Traveling Waves ($\alpha \neq 0$): The system admits non-equilibrium steady states of the form $\rho(t, \theta) = \rho(\theta - kt)$, where $k$ is the wave speed. The paper seeks to characterize these solutions rigorously.

Reduction to Self-Consistency:
By assuming a traveling wave ansatz and utilizing the specific one-mode interaction, the infinite-dimensional PDE reduces to a finite-dimensional mean-field self-consistency condition. This condition relates the macroscopic order parameters (concentration $r$ and skewness/frustration $\beta$) to the coupling strength $\mu$.

3. Key Contributions and Methodology

The paper introduces novel mathematical tools and distribution families to solve the asymmetric problem.

A. The Exponentially Modified von Mises (EMvM) Distribution

The author defines a new family of probability distributions, the Exponentially Modified von Mises (EMvM), which serves as the exact solution for the traveling wave states.

• Structure: It is a skew extension of the standard von Mises distribution.
• Construction: It arises from the exponential filtering of the Fourier spectrum of the von Mises kernel.
• Parameters:
  • $\theta_0$: Location parameter.
  • $r$: Concentration parameter (analogous to inverse temperature).
  • $\beta$: Skewness parameter (related to the wave speed $k = \tan \beta$).
• Significance: This distribution preserves the "fragile Bessel moment hierarchy," ensuring the family is globally identifiable (unique mapping between parameters and moments), a property lacking in other skew circular distributions.

B. Asymmetrically Extended Modified Bessel Functions

To handle the asymmetry, the author defines asymmetrically extended modified Bessel functions, denoted $I_n(k, r)$.

• These functions generalize the standard modified Bessel functions $I_n(r)$ to account for the decay rate $k$ (related to the skewness).
• They allow the derivation of the characteristic function and the mean resultant vector for the EMvM distribution.

C. Geometric Function Theory and Analytic Combinatorics

The core of the proof relies on establishing the global invertibility of the "normalized mean resultant map."

• The Map: The map takes natural parameters $(r, \beta)$ to effective mean-field parameters $(R, \alpha)$.
• Technique: The author transforms the problem into isogonal coordinates (constant phase angle lines).
• Monotonicity: The proof demonstrates that the map is radially strictly monotone along these isogones. This is achieved by analyzing ratios of generalized power series.
• Key Lemmas:
  • Biernacki–Krzyż Lemma: Used to prove monotonicity of ratios of power series.
  • Karlin's Variation-Diminishing Theorem: Applied to totally positive (TP2) kernels.
  • Log-concavity: The proof hinges on showing that the coefficient sequences of the power series expansions are strictly log-concave, a property derived using contour integrals and Fourier analysis of the Poisson kernel.

4. Main Results

Theorem 1: Global Phase Transition (Theorem 5.2)
The system undergoes a continuous, global supercritical pitchfork bifurcation at a critical coupling strength:
$$\mu_c = 2 \sec \alpha$$

• For $\mu \le \mu_c$: The unique solution is the incoherent state (uniform distribution $\rho = 1/2\pi$).
• For $\mu > \mu_c$: A unique non-trivial traveling wave solution emerges continuously from the incoherent state.
• Global Nature: The branch of non-trivial solutions exists for all $\mu > \mu_c$ and does not terminate or bifurcate again. The solution curve can be parametrized globally by the concentration parameter $r$.

Theorem 2: Existence and Uniqueness of Traveling Waves (Theorem 5.3)
For the McKean–Vlasov equation with Sakaguchi–Kuramoto interaction:

• If $\mu \le \mu_c$, the incoherent state is the unique solution.
• If $\mu > \mu_c$, there exists a unique coherent traveling wave solution of the form:
  $$\rho(t, \theta) = \rho_{\text{EMvM}}(\theta - kt; \theta_0, k, r)$$
  where $k$ (wave speed) and $r$ (concentration) are uniquely determined by the self-consistency equations.
• The wave speed is given by $k = \tan \beta$, where $\beta$ is the skewness parameter.

Limiting Cases:

• As $\alpha \to 0$, the result recovers the classical Kuramoto phase transition at $\mu_c = 2$ with stationary von Mises distributions.
• As $|\alpha| \to \pi/2$, the critical coupling $\mu_c \to \infty$, implying synchronization is impossible for maximal frustration.

5. Significance and Impact

1. Rigorous Foundation for Traveling Waves: This is the first rigorous mathematical treatment establishing the global existence and uniqueness of traveling wave solutions for the asymmetric Sakaguchi–Kuramoto model. Previous work relied on formal reductions (Ott–Antonsen) which lacked rigorous global guarantees.
2. New Statistical Distribution: The introduction of the EMvM distribution provides a new tool for circular statistics, offering a natural, identifiable model for skewed circular data, analogous to the ex-Gaussian in linear statistics.
3. Neurobiological Relevance: The results provide a theoretical basis for understanding traveling waves in cortical circuits, which are ubiquitous in brain activity but difficult to model due to time delays and spatial extent. The paper validates the use of mean-field frustration parameters to capture these complex dynamics without explicit spatial embedding.
4. Methodological Advancement: The combination of geometric function theory, analytic combinatorics (log-concavity of coefficients), and totally positive kernels offers a powerful new toolkit for analyzing non-equilibrium phase transitions in asymmetric systems where standard symmetry-based bifurcation methods fail.

In summary, the paper bridges the gap between the idealized Kuramoto model and realistic, delay-induced neural dynamics by proving that phase frustration leads to a robust, unique, and globally stable traveling wave regime, characterized by a novel skew distribution family.