Occupation time statistics for non-Markovian random walks

V. Méndez 1, R. Flaquer-Galmés 1, A. Pal 2,3 · 2024

Read Occupation time statistics for non-Markovian random walks by V. Méndez 1, R. Flaquer-Galmés 1, A. Pal 2,3 (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-05247.txt.

Introduction

Vicenç Méndez, Rosa Flaquer-Galmés, and Arnab Pal study how long random walkers linger in a region in a 2024 paper. Their walkers have memory: step timing follows a broad wait law rather than fixed ticks, which can slow spread. The team also adds random resets that snap the walker back to a set point.

They solve a backward equation to get the full spread of linger times in a bounded stretch and in a half line. The half line case revisits the famed arcsine law, which the team tests across rest regimes. Shapes, means, and long run limits are worked out for both normal and anomalous spread. Computer runs matched the math well. The results feed fields from physics to search theory.

Catalog record

CREATORV. Méndez 1, R. Flaquer-Galmés 1, A. Pal 2,3
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE93 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05247

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Occupation time statistics for non-Markovian random walks V. Méndez 1, R. Flaquer-Galmés 1, A. Pal 2,3 Abstract We study the occupation time statistics for non-Markovian random walkers based on the formalism of the generalized master equation for the Continuous-Time Random Walk. We also explore the case when the random walker additionally undergoes a stochastic resetting dynamics. We derive and solve the backward Feynman-Kac equation to find the characteristic function for the occupation time in an interval and for the half occupation time in the semi-infinite domain. We analyze the behaviour of the PDFs, the moments, the limiting distributions and the ergodic properties for both occupation times when the underlying random walk is normal or anomalous. For the half occupation time, we revisit the famous arcsine law and examine its validity pertaining to various regimes of the rest period of the walker. Our results have been verified with numerical simulations exhibiting an excellent agreement. I Introduction The study of animal movement, particularly concerning foraging behavior, provides insights into ecological dynamics and animal movement patterns. In recent years, the monitoring and subsequent analysis of animal movements or trajectories connected to statistical physics and dynamical systems has gained attention as an effective field of study [ 1 ] . The most simple approach to describe the behavioral navigation of living beings is to consider the walker’s position in time as a Markovian (i.e., a memory-less) stochastic process [ 2 ] . However, animals employ various strategies to explore their environment efficiently, often influenced by memory and other cognitive mechanisms. For instance, it has been recurrently observed for many species the tendency to repeatedly return to previously visited sites as a part of their regular foraging strategies [ 3 ] . For this reason, incorporating memory within random walk models is important for improving both predictive and descriptive movement tools. This can be done, for example, by considering the resetting of the walker’s position to a given point [ 4 ] (spatial memory) or long rests between successive jumps (temporal memory) [ 5 ] . In both cases, the resulting random walk is non-Markovian. To extract statistical information from these non-Markovian models, one can analyze stochastic functionals of the walker’s trajectory. The prototypical example is the Brownian functional, an integration of the trajectory of a Brownian walker [ 6 , 7 ] . They have diverse applications in physics, mathematics, and other fields. The probability density function (PDF) of Brownian functionals satisfies the Feynman-Kac (FK) equation, which is a Schrödinger equation in imaginary time. This equation was derived in 1949 by Kac inspired by Feynman’s path integrals. More recently the FK equation has been derived for resetting Brownian walkers [ 8 , 9 , 10 , 11 , 12 ] and non-Brownian subdiffusion [ 13 , 14 ] . However, derivation for a general FK equation for any Markovian or non-Markovian random walk is still missing. Several functionals are of ecological interest. Take e.g., the occupation time or residence time, which estimates the time spent by a random walker in a given domain which has ecological relevance. If x=0 is the starting point and the animal’s nest location then the interval [-a,a] around x=0 can be regarded as the home range and the occupation time in [-a,a] can be interpreted as the time spent by the animal within such home range. It provides useful information on the animal’s behavior when foraging. One can also consider the half occupation time (the occupation time in the domain x>0 ) or the occupation time in a finite interval which can provide information of a restricted foraging space. The distribution function for the half occupation time of a Brownian walker is given by the seminal Lévy’s arcsine law [ 15 , 16 , 17 , 18 , 19 ] and holds a special place in stochastic process, probability theory and physics. Statistics of occupation times are of course not limited to Brownian walker, and it is a topic of wide investigation for example in the context of renewal processes theory [ 20 ] , active particles [ 21 , 22 ] , fractional Brownian motion [ 23 ] , experiments on blinking quantum dots [ 24 ] , colloids modulated with optical traps [ 25 ] and light resonators in a cavity [ 26 ] and weak ergodicity breaking of dynamics generated using deterministic maps [ 27 ] . On the other hand, the occupation time in an interval for a subdiffusive walker has been studied in Ref. [ 14 ] . More recently, the statistical and ergodic properties of a Brownian walker in the presence of power-law resetting have been studied [ 28 ] . Despite these studies, a unified framework analyzing the occupation time spanned in different geometries for generic random walkers is missing. The aim of this work is to fill in this gap by developing a comprehensive framework to study the occupation time in a confinement and semi-confinement for generic random walkers namely the Markovian- and non-Markovian- walkers. To this end, we derive the Feynman-Kac equation for the occupation time of generic random walkers with arbitrary rest periods. We solve for the characteristic equation, compute the moments, obtain the PDF, analyze the ergodic properties, and extend our study considering the stochastic resetting of the walker’s position. The article is organized as follows. In Section II we review the derivation of the Generalized Master equation for the Continuous Time Random Walk (CTRW). In Section III , we derive the FK equation for the characteristic function of a general stochastic functional when the random walk is described by the Generalized Master equation of the CTRW. In Sections IV and V we study the statistical properties (mean, fluctuation, and the PDF) for the occupation time in an interval and for the half occupation time. Finally, in Section VI we analyze how these statistical properties are modified by a Poissonian resetting of the walker’s position. We summarize our work with future outlook in Section VII . Some details of the calculations are moved to the Appendices for the clarity of the presentation. II Non-Markovian random walk We use the Continuous-Time Random Walk (CTRW) framework put forward by Montroll and Weiss to deal with non-Markovian random walks [ 29 ] . We refer to [ 30 ] for an extensive review of this subject. In the current set-up, we assume that the walker begins at the initial position at time t=0 waiting a random time before performing the first jump to a new position. It then waits again for a time interval before proceeding to the next jump. Jump lengths and waiting times are independent and identically distributed random variables according to the PDFs \Psi(x) and \varphi(t) , respectively. When the waiting time PDF is exponential then the random walk is Markovian, otherwise, it is non-Markovian. We first define the Laplace and Fourier transforms of the waiting time- and jump length- distributions as \varphi(s) and \Psi(k) respectively \varphi(s)=\mathcal{L}_{t\to s}\left[\varphi(t)\right]=\int_{0}^{\infty}e^{-st}\varphi(t)dt, and \displaystyle\Psi(k)=\mathcal{F}_{x\to k}[\Psi(x)]=\int_{-\infty}^{\infty}e^{ikx}\Psi(x)dx. (1) The propagator P(x,t|x_{0}) that describes the position of the walker at time t given that it had started at the location x_{0} is given by the famous Montroll-Weiss equation [ 29 ] , which in Fourier-Laplace

Read the complete text