The Onsager-Machlup functional associated with additive fractional noise

Abstract: We consider the solution of a stochastic differential equation with additive multidimensional fractional noise. In the case $\frac14<H<\frac12$, we compute the Onsager-Machlup functional (with respect to the driving fractional Brownian motion) for the supremum norm and the Hölder norms with exponent $α\in \left(0,H-\frac14\right)$ for any element of the Cameron-Martin space $\mathcal H_H$, extending a previous result of Moret and Nualart. In the more general case $H<\frac12$ and $α\in \left(0,H\right)$, we formulate a condition on $h\in\mathcal H_H$ under which the computation of the Onsager-Machlup functional $J\left(h\right)$ follows.

• The paper extends Onsager-Machlup functional analysis to SDEs driven by fractional Brownian noise, establishing existence under the supremum and Hölder norms.
• It rigorously derives explicit expressions for the functional using advanced analytical methods within the Hurst index (1/4, 1/2) regime.
• It offers a robust mathematical framework with implications for filtering theory, statistical mechanics, and financial mathematics.

Summary of "The Onsager-Machlup functional associated with additive fractional noise" (1705.08976)

Introduction and Background

This paper explores the Onsager-Machlup functional in stochastic differential equations (SDEs) driven by multidimensional fractional Brownian motion (fBm) with a Hurst index $H$ in the range $\left(\frac{1}{4}, \frac{1}{2}\right)$. The Onsager-Machlup functional is traditionally used to characterize the likelihood of a stochastic process's path; it is essential in many applications, including physics and quantitative finance. Historically, this functional has been extensively studied for classical Brownian motion. However, given the complexity and non-Markovian nature of fBm, characterizing this functional in that context presents significant mathematical challenges. The paper builds on foundational work by Moret and Nualart, extending previous results to more general norms such as the supremum norm and specific Hölder norms.

Mathematical Framework

The paper provides a rigorous analytical framework for the computation of the Onsager-Machlup functional within the specified range of the Hurst parameter. The authors focus on the fBm-defined substantive driving noise process and explore the corresponding SDE:

$X_t = \int_0^t b(X_s) \, ds + B_t,$

where $B_t$ represents the fBm. The central objective is to determine the Onsager-Machlup functional $J(h)$ for paths $h$ in the Cameron-Martin space $\mathcal{H}_H$ associated with the fBm, characterized via a condition under which this computation holds.

Main Results

The paper presents significant advancements in the analytical computation of the Onsager-Machlup functional for multidimensional fBm. In particular, the authors prove the existence of this functional for various norms and provide explicit expressions within the regime $H \in (\frac{1}{4}, \frac{1}{2})$. They extend the results for norms both in the supremum and Hölder cases, highlighting conditions under which $J(h)$ can be obtained. Notably, the functional is expressed as:

$$J(h) = -\frac{1}{2}\left[{\left(h - \int_0^{\cdot} b(h_s) \, ds\right)}^2_H + \int_0^1 \nabla \cdot b(h_s) \, ds\right],$$

where bold terms denote critical computational elements leading to superior understanding and prediction capability in such stochastic models.

Implications and Future Directions

The work has substantial theoretical implications for potential future developments in the modeling of complex stochastic systems influenced by fractional noise. The results further intensify the theoretical foundation in areas such as filtering theory, statistical mechanics, and financial mathematics, where understanding the behavior of solutions to SDEs driven by fBm is crucial. The generalization of norm types and detailed conditions outlined set the foundation for subsequent research, particularly in exploring theoretical influence patterns across different domains of stochastic processes.

Future research could explore extending these principles into broader applications, examining higher-dimensional systems, or considering other forms of noise and perturbations within the SDE framework. Moreover, further work might explore the computational methods to solve these SDEs numerically, leveraging machine learning or advanced simulation methods.

Conclusion

This paper substantially contributes to the ongoing exploration of Onsager-Machlup functionals in fractional stochastic systems. By extending previous findings and providing robust mathematical treatments of the functional under a range of conditions and norms, the authors enhance the understanding and potential application of these principles in complex, real-world scenarios characterized by fractional stochastic dynamics.