On the distribution of the telegraph meander and its properties

Basic Information

• Paper ID: 2504.11387
• Title: On the distribution of the telegraph meander and its properties
• Authors: A. Pedicone, E. Orsingher (Sapienza University of Rome)
• Classification: math.PR (Probability Theory)
• Publication Date: April 2024 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2504.11387

Abstract

This paper investigates the distribution of the telegraph meander, a random function obtained by conditioning a telegraph process to remain above the zero level. Based on the reflection principle for finite-velocity random motions, the distribution law of the telegraph meander can be expressed as the spatial derivative of the distribution law of the telegraph process with initial negative velocity. Consequently, the authors derive the characteristic function, moments of all orders, and the hyperbolic partial differential equation governing the distribution law of the telegraph meander. Furthermore, they prove that the Brownian meander is the weak limit of the telegraph meander.

Research Background and Motivation

Problem Background

1. Importance of Telegraph Processes: The telegraph process is the simplest stochastic evolution process, introduced by Goldstein to describe the motion of particles traveling at constant velocity and randomly changing direction at exponentially distributed time intervals.
2. Application Requirements: The finite velocity of motion and the intensity of direction changes in telegraph processes make them suitable for describing real-world phenomena in geology, finance, and physics.
3. Significance of Conditioned Processes: In many practical applications, it is necessary to study stochastic processes under specific constraints, such as asset prices remaining above certain barriers in financial option pricing.

Research Motivation

1. Theoretical Refinement: Although telegraph processes themselves have been extensively studied, systematic theoretical analysis of their conditioned versions (telegraph meanders) remains incomplete.
2. Methodological Innovation: The recently developed reflection principle for finite-velocity random motions by Cinque and others provides new mathematical tools for studying telegraph meanders.
3. Limit Theory: Establishing connections between telegraph meanders and Brownian meanders provides theoretical foundations for understanding the asymptotic behavior of finite-velocity processes.

Core Contributions

1. Derivation of Finite-Dimensional Distributions: Using the reflection principle and Markov properties, the complete distribution expression of the telegraph meander is obtained.
2. Establishment of Differential Representation: It is proven that the distribution law of the telegraph meander can be expressed as the spatial derivative of the distribution law of the telegraph process with initial negative velocity.
3. Derivation of Characteristic Functions and Moments: Explicit expressions for the characteristic function and moments of all orders of the telegraph meander are computed using complex analysis methods.
4. Discovery of Governing Hyperbolic Equations: It is proven that the distribution law of the telegraph meander satisfies a modified telegraph equation (hyperbolic partial differential equation).
5. Proof of Weak Convergence Theorem: Under Kac's conditions, the telegraph meander converges weakly to the Brownian meander.

Methodology Details

Problem Definition

The telegraph meander $T^+ = (T^+(s), s \in [0,t])$ is defined as the telegraph process $T(s)$ conditioned on $\min_{0 \leq s \leq t} T(s) \geq 0$, that is: $P\{T^+(t_1) \in A_1, \ldots, T^+(t_n) \in A_n\} = P\{T(t_1) \in A_1, \ldots, T(t_n) \in A_n | \min_{0 \leq s \leq t} T(s) \geq 0\}$

Core Methods

1. Application of the Reflection Principle

Utilizing the reflection principle for finite-velocity random motions proven by Cinque: $P\{T(t) \in dx, \max_{0 \leq s \leq t} T(s) \geq y | V(0) = c\} = P\{T(t) \in 2y - dx | V(0) = -c\}$ for $x \in [2y - ct, y)$ and $y \in [0, ct)$.

2. Construction of Joint Distributions

Through the Markov property of the telegraph process $(T, V)$, finite-dimensional distributions of the telegraph meander are constructed: $P\{T^+(t_1) \in dx_1, \ldots, T^+(t_n) \in dx_n\} = \sum_{v_1,\ldots,v_n \in \{-c,c\}} \prod_{k=1}^n f_{v_{k-1}}(x_k - x_{k-1}, v_k, -x_{k-1}, t_k - t_{k-1}) \frac{m_{v_n}(-x_n, t-t_n)}{m_c(0,t)}$

3. Establishment of Differential Representation

It is proven that the distribution of the telegraph meander at terminal time can be expressed as: $q_{\lambda,c}(x,t) = \lim_{\varepsilon \downarrow 0} P\{T(t) \in dx | \min_{0 \leq s \leq t} T(s) \geq -\varepsilon, V(0) = -c\} = -\frac{\partial}{\partial x} p_{\lambda,c}^-(x,t) / p_{\lambda,c}^-(0,t)$

Technical Innovations

1. Clever Application of the Reflection Principle: The classical reflection principle for Brownian motion is generalized to the finite-velocity case, providing a powerful tool for studying telegraph meanders.
2. Discovery of Differential Relations: A differential relationship is established between the distribution laws of conditioned and original processes, analogous to the Brownian meander case.
3. Derivation of Hyperbolic Equations: Through differential relations and the telegraph equation, a modified hyperbolic equation satisfied by the telegraph meander is derived, revealing its intrinsic mathematical structure.

Main Theoretical Results

Explicit Expression of Distribution

The distribution of the telegraph meander at time $t$ is: $q_{\lambda,c}(x,t) = \frac{1_{[0,ct)}(x)}{I_0(\lambda t) + I_1(\lambda t)} \left\{ \frac{\lambda x I_0(\frac{\lambda}{c}\sqrt{c^2t^2 - x^2})}{c(ct + x)} + \frac{I_1(\frac{\lambda}{c}\sqrt{c^2t^2 - x^2})}{\sqrt{c^2t^2 - x^2}} \left(\frac{\lambda x}{c} + \frac{ct - x}{ct + x}\right) \right\}$

Characteristic Function

For $|\gamma| < \lambda/c$: $E[e^{i\gamma T^+(t)}] = 1 + \frac{c\gamma(c\gamma + i\lambda)\sinh(t\sqrt{\lambda^2 - c^2\gamma^2}) + ic\gamma\int_0^t \sinh(s\sqrt{\lambda^2 - c^2\gamma^2})I_0(\lambda(t-s))ds}{\lambda\sqrt{\lambda^2 - c^2\gamma^2}(I_0(\lambda t) + I_1(\lambda t))}$$+ \frac{ic\gamma(\cosh(t\sqrt{\lambda^2 - c^2\gamma^2}) - I_0(\lambda t) + ic\gamma\int_0^t \cosh(s\sqrt{\lambda^2 - c^2\gamma^2})I_0(\lambda(t-s))ds)}{\lambda(I_0(\lambda t) + I_1(\lambda t))}$

Governing Hyperbolic Equation

The distribution law of the telegraph meander satisfies: $\frac{\partial^2 u}{\partial t^2} - c^2 \frac{\partial^2 u}{\partial x^2} + 2\lambda \frac{\partial u}{\partial t} - \frac{2a}{t} \frac{\partial u}{\partial t} = \left(\frac{\lambda}{t} - \frac{2a}{t^2}\right)u$ where $a(t) = \frac{I_1(\lambda t)}{I_0(\lambda t) + I_1(\lambda t)}$.

Weak Convergence Theorem

Under Kac's conditions ($\lambda \to +\infty, c \to +\infty$ and $c^2/\lambda \to 1$): $T_\alpha^+ \Rightarrow B^+ \text{ on } C^+[0,t]$ where $B^+$ is the Brownian meander.

Conditional Distribution Analysis

Conditional Distribution Given Number of Jumps

When the number of Poisson events $N(t) = n$ is given: