Diffusive-to-Ballistic transition in a Persistent Random Walk

This paper investigates a persistent random walk with time-dependent velocity reversal probabilities, identifying a critical transition at $\alpha=1$ for power-law decay $p(t)\sim t^{-\alpha}$ that separates super-diffusive and ballistic regimes, a phenomenon shown to be robust across various probability forms and arbitrary spatial dimensions under isotropy.

The Gist

Imagine a drunk person walking down a straight hallway. In a standard "random walk," every time they take a step, they flip a coin: heads, they go forward; tails, they turn around and go backward. Over time, this person wanders aimlessly, and their distance from the start grows slowly, like a slow leak filling a bucket. This is diffusion.

But what if this walker has a bit of "stubbornness"? What if they tend to keep going in the same direction for a while before deciding to turn around? This is called a Persistent Random Walk.

This paper studies a specific, slightly magical version of this stubborn walker. In this version, the walker's "stubbornness" changes over time. The longer they walk, the less likely they are to flip a coin and change direction. The authors ask a simple question: How does the rate at which they lose their stubbornness change the way they move?

The Magic Rule: The Power Law

The authors set up a rule where the chance of turning around depends on how long the walker has been walking. They use a mathematical "recipe" called a power law. Think of it like a timer that counts down the probability of turning.

The key variable in this recipe is a number called $\alpha$ (alpha). This number controls how fast the walker's stubbornness fades. The paper discovers that $\alpha = 1$ is a magical tipping point, a "phase transition," where the walker's behavior changes completely.

The Three Regimes of the Walker

1. The "Super-Runner" ($\alpha < 1$)

Imagine a walker who is very stubborn. Even as time goes on, they keep flipping the coin to turn around, but they do it less and less often. However, they never stop flipping the coin entirely.

• What happens: Because they keep changing direction, but less frequently, they manage to cover ground much faster than a normal random walker. They don't just walk; they "super-diffuse."
• The Analogy: Think of a runner who keeps getting tired and slowing down, but never actually stops running. They cover more distance than a normal walker, but they are still constantly adjusting their path.

2. The "Freeze" ($\alpha > 1$)

Now, imagine a walker who is stubborn to the point of obsession. The rule says that after a certain amount of time, the chance of them turning around becomes so tiny that it effectively hits zero.

• What happens: Eventually, this walker flips the coin, gets a "keep going" result, and never turns again. They lock into a single direction and zoom off in a straight line forever.
• The Analogy: This is like a car that gets stuck in "cruise control" and refuses to brake or steer. The motion becomes ballistic (like a bullet). The paper calls this "velocity freezing."

3. The "Tipping Point" ($\alpha = 1$)

This is the most interesting part. It's the exact middle ground between the super-runner and the frozen bullet.

• What happens: Here, the walker does keep flipping the coin forever, but the timing is just right. The correlations (the memory of which way they were going) decay very slowly. Even though they keep turning, they manage to maintain a straight-line speed.
• The Surprise: You might think that if you keep turning, you can't go in a straight line. But at this exact critical point, the "memory" of their direction lasts just long enough to create ballistic motion (straight-line speed), even though they are technically still turning occasionally. It's a delicate balance where the "turning" and the "memory" cancel each other out perfectly to create a straight path.

How They Proved It

The authors didn't just guess; they did the math and ran computer simulations.

• The "Binder Cumulant": They used a statistical tool (like a thermometer for chaos) to measure the fluctuations of the walker's position. When they plotted this for different values of $\alpha$, the lines crossed perfectly at $\alpha = 1$. This crossing is the "smoking gun" that proves a real, sharp transition is happening.
• The "Survival Probability": They calculated the odds that a walker would never turn around. For the "Freeze" regime ($\alpha > 1$), there is a real, non-zero chance the walker never turns. For the other regimes, that chance is zero. This acts like a switch that flips on at the critical point.

The Big Picture

The paper shows that this isn't just about a specific mathematical formula. The transition happens whenever the "expected number of turns" either stays finite (the walker stops turning eventually) or grows forever (the walker keeps turning forever).

They also showed that this works in any number of dimensions. Whether the walker is moving on a 2D floor or a 3D room, as long as they can turn in any direction equally (isotropy), this "tipping point" at $\alpha = 1$ remains the same.

Summary in One Sentence

The paper reveals that if a "stubborn" walker changes their mind less often over time, there is a precise mathematical tipping point where their movement shifts from a chaotic, wandering drift to a straight-line, bullet-like sprint, driven by the subtle balance between how often they turn and how long they remember their direction.

Technical Summary

Technical Summary: Diffusive-to-Ballistic Transition in a Persistent Random Walk

Problem Statement
Conventional persistent random walk (PRW) models, characterized by a constant velocity reversal probability $p$, exhibit a temporal crossover from ballistic motion at short times to diffusive behavior at long times. However, many physical, chemical, and biological systems (e.g., bacterial run-and-tumble motion, active matter) display non-stationary dynamics and memory effects that cannot be captured by Markovian models with constant parameters. The central problem addressed in this work is to determine whether a sharp dynamical transition between different diffusion regimes can arise in persistent motion when the velocity reversal probability $p(t)$ is explicitly time-dependent. Specifically, the authors investigate the conditions under which the long-time dynamics qualitatively changes based on the growth of the total number of velocity reversals.

Methodology
The authors study a one-dimensional discrete-time persistent random walk where the velocity $v(t) = \pm 1$ evolves according to a time-dependent reversal probability $p(t+1)$. The position is updated as $x(t+1) = x(t) + v(t+1)$.

1. Exact Recursion Relations: The authors derive exact closed-form expressions for the position-velocity correlation $A(t) = \langle x(t)v(t) \rangle$ and the mean squared displacement (MSD) $\langle x^2(t) \rangle$ in terms of the persistence parameter $\gamma(t) = 1 - 2p(t)$.
2. Asymptotic Analysis: They analyze the long-time behavior by examining the expected number of velocity reversals, $N(t) = \sum_{u=1}^t p(u)$. The study focuses on a power-law reversal probability $p(t) = c t^{-\alpha}$ to identify critical regimes.
3. Scaling and Statistics: The transition is characterized using:
   • Velocity autocorrelation functions.
   • Survival probabilities and persistence length distributions.
   • Finite-time scaling of displacement fluctuations ($\sigma_x^2$) and the Binder cumulant $U$.
4. Generalization: The results are extended to arbitrary spatial dimensions $d$ under the assumption of isotropic velocity space, and the findings are compared against other time-dependent protocols (exponential and linear decay).

Key Contributions and Results

• Identification of a Dynamical Transition: The study identifies a critical threshold at $\alpha = 1$ for the power-law reversal probability $p(t) \sim t^{-\alpha}$. This transition separates two distinct dynamical regimes:

  • Super-diffusive Regime ($\alpha < 1$): The expected number of reversals $N(t)$ diverges as $t^{1-\alpha}$. Velocity correlations decay as a stretched exponential, leading to super-diffusive motion where $\langle x^2(t) \rangle \sim t^{1+\alpha}$.
  • Ballistic Regime ($\alpha \ge 1$):
    • For $\alpha > 1$, $N(\infty)$ is finite. With finite probability, the velocity never flips after a certain time ("velocity freezing"), resulting in ballistic motion $\langle x^2(t) \rangle \sim t^2$. A finite fraction of trajectories, $P_\infty(\alpha) = e^{-c\zeta(\alpha)}$, remain ballistic indefinitely.
    • For the marginal case $\alpha = 1$, $N(t)$ diverges logarithmically. Despite infinite reversals, the system exhibits ballistic scaling $\langle x^2(t) \rangle \sim t^2$ driven by long-lived algebraic velocity correlations, distinct from the freezing mechanism of $\alpha > 1$.

• Characterization of the Critical Point: At $\alpha = 1$, the transition is marked by:

  • Logarithmic Finite-Time Scaling: The displacement variance scales as $\sigma_x^2(t) \sim t^2 G[(\alpha-1)\ln t]$, indicating an exponentially large characteristic timescale $t^* \sim \exp(1/|\alpha-1|)$.
  • Binder Cumulant Crossing: Numerical simulations show a sharp crossing of the Binder cumulant curves at $\alpha_c = 1$ for different observation times, confirming the transition as a genuine dynamical phase transition.
  • Broad Displacement Distribution: At criticality, the displacement distribution becomes exceptionally broad, reflecting strong fluctuations.

• Dimensional Independence: The transition persists in arbitrary spatial dimensions $d$, provided the velocity space remains isotropic. The critical exponent $\alpha_c = 1$ remains unchanged, though non-universal prefactors depend on $d$.

• Generality of the Criterion: The authors argue that the transition is not limited to pure power laws. Any reversal protocol where the expected number of reversals $N(t)$ grows without bound for one set of parameters and remains finite for another (e.g., $p(t) \sim c \log t / t^\alpha$) will exhibit a similar sharp transition. In contrast, protocols with intrinsic time scales (exponential or linear decay) show smooth crossovers rather than sharp transitions.

Significance
The paper establishes a general criterion for non-equilibrium dynamical transitions in persistent random walks: the nature of the long-time transport is determined by whether the cumulative probability of velocity reversal converges or diverges. The work demonstrates that a sharp transition between super-diffusive and ballistic regimes can occur solely due to the temporal decay of the reversal probability, without the need for external driving or spatial heterogeneity. The identification of $\alpha=1$ as a critical point separating regimes with distinct mechanisms for ballistic motion (velocity freezing vs. long-lived correlations) provides a refined understanding of anomalous transport in systems with aging and memory effects. The results are applicable to the physics of active matter, biological motion, and search strategies where non-stationary dynamics are prevalent.