Stochastic 1D search-and-capture as a G/M/c queueing model
José Giral-Barajas, Paul C Bressloff · 2025
Read Stochastic 1D search-and-capture as a G/M/c queueing model by José Giral-Barajas, Paul C Bressloff (2025) 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-2503-21883.txt.
Introduction
José Giral-Barajas and Paul Bressloff model a cell's supply line as a queue, in a 2025 math paper. Particles wander a one-dimensional track searching for a target, arrive at random times, and are consumed one by one by a limited number of servers. Queueing theory names this a G/M/c queue; the finite server count is the twist, since older work assumed endless servers. Joining queue theory with first-passage times, the pair derive when the stock at the target settles to a steady level.
They give formulas for the average stock and its spread at steady state. Adding a second competing target can, in defined situations, help the first one. With several searchers, they compute arrival patterns using renewal theory. Throughout, the finite-server queue reacts more sharply to the search setup than the endless one, a caution for modelers of delivery inside cells.
Catalog record
| CREATOR | José Giral-Barajas, Paul C Bressloff |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 97 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-21883
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Stochastic 1D search-and-capture as a G/M/c queueing model José Giral-Barajas, Paul C Bressloff Abstract We study the accumulation of resources within a target due to the interplay between continual delivery, driven by 1D stochastic search processes, and sequential consumption. The assumption of sequential consumption is key because it changes the commonly used G / M / ∞ G/M/\infty queue to a G / M / c G/M/c queue. Combining the theory of G / M / c G/M/c queues with the theory of first-passage times, we derive general conditions for the search process to ensure that the number of resources within the queue converges to a steady state and compute explicit expressions for the mean and variance of the number of resources within the queue at steady state. We then compare the performance of the G / M / c G/M/c queue with that of the G / M / ∞ G/M/\infty queue for an increasing number of servers. We extend the model to consider two competing targets and show that, under specific scenarios, an additional target is beneficial to the original target. Finally, we study the effects of multiple searchers. Using renewal theory, we numerically compute the inter-arrival time density for M M searchers in the Laplace space, which allows us to exploit the explicit expressions for the steady-state statistics of the number of resources within G / M / 1 G/M/1 and G / M / ∞ G/M/\infty queues, and compare their behaviour with different numbers of searchers. Overall, the G / M / c G/M/c queue shows a tighter dependence on the configuration of the search process than the G / M / ∞ G/M/\infty queue does. August 2025 Keywords : search processes, first-passage times, queueing theory, renewal theory Journal: J. Phys. A: Math. Gen. 1 Introduction Queues and random search strategies are prevalent in the natural world. Random search strategies have proven valuable when searching for one or multiple targets in an unknown domain. Examples include animal movement [ 1 , 2 , 3 ] , cellular transport [ 4 , 5 , 6 ] and software testing and optimisation [ 7 , 8 ] . On the other hand, any system in which customers—or resources—arrive randomly and receive service provided by a set of servers at random times can be studied as a queueing process. Queues are ubiquitous in human activities, such as telecommunications [ 9 , 10 , 11 ] and transportation [ 12 , 13 , 14 ] , and in the broader natural world, such as gene expression and regulation [ 15 , 16 , 17 ] and metabolic pathways [ 18 , 19 , 20 ] . Queueing processes [ 21 , 22 , 23 , 24 ] and stochastic search processes [ 26 , 25 ] have been extensively studied. More recently, they have been combined to develop a queueing model for the accumulation of resources in a target, owing to multiple rounds of search-and-capture processes [ 27 , 28 ] . The basic idea is to map the delivery of resources to customers arriving at a service station and the departure of customers from the station once they have received service, to the consumption—or degradation—of resources. So far, this mapping has always been made with the assumption that resources are consumed independently, rendering the number of servers in the service station infinite. This implies that the appropriate queueing system is a G/M/\infty queue, where G denotes a general inter-arrival time density, \mathcal{F} , M denotes Markovian service times and \infty specifies that the service station has an infinite number of servers and consumes resources independently. The assumption of independent consumption makes the sojourn times independent and ensures the existence of a steady state, regardless of the spatial configuration of the search process [ 28 , 31 ] . In this study, we change the consumption protocol from independent to sequential consumption so that there are a finite number of servers in the system consuming resources in the order they arrive. The appropriate queueing system becomes G/M/c , where c denotes the number of servers [ 32 , 33 ] . In this new system, when the number of resources exceeds the number of servers, waiting times before service emerge and sojourn times cease to be independent. As a motivating example, consider the transport of proteins and vesicles along the axon of a neuron. Neurons synthesise most of their necessary molecular components in the cell body or soma, including those that ensure healthy maintenance of the axon and synaptic cell-cell communication [ 34 , 35 ] . Malfunctions in axonal transport may lead to the atypical accumulation of cellular resources in the axon and the synaptic targets and trigger neurodegenerative diseases [ 36 ] . In previous work, the stochastic accumulation of resources in one or more synaptic targets due to the active transport and delivery of constitutive proteins was studied as a search-and-capture process with a G/M/\infty queue [ 37 ] . In this particular model, service in the queue was interpreted as a form of protein degradation, with no constraints on the number of proteins that could degrade over a given time interval. This is equivalent to having an infinite number of servers. On the other hand, if one is interested in the transport of neurotransmitter-rich synaptic vesicles, then a more realistic scenario is to consider the sequential utilisation or consumption of delivered vesicles during synaptic processing, which is better modelled as a G/M/c queue. Here c represents the maximum number of vesicles that can be processed simultaneously by fusing to the presynaptic plasma membrane, for example. One of the major consequences of having a finite number of servers is that we must now consider scenarios in which the spatial configuration of the search process causes the resources in the system to blow up, preventing the queue from converging to a steady state. Such a phenomenon cannot occur in the case of a G/M/\infty queue, and is thus the main novel feature of the current paper. For concreteness, we develop the theory by considering a diffusive search process inside the bounded interval [0,L] with a target at one or both ends of the interval. Even in this simple domain, the spatial configuration of the search processes strongly influences the behaviour of the queueing system, which translates into nontrivial analytical and numerical results for the steady-state resource statistics for the G/M/c . In particular, we obtain the following main results: (i) We derive steady-state existence conditions and construct critical regions in the parameter space of the search process that ensure convergence of the resulting G/M/c queue to the steady state. Furthermore, under convergence to a steady state, we compute the mean and variance of the queue length. (ii) We relate the mean waiting time before service to the Wasserstein distance and utilise it to examine the convergence of the G/M/c to G/M/\infty as c\to\infty . (iii) We extend the steady state criterion to the case of a single searcher and two targets and the case of a single target with multiple searchers. The structure of the paper is as follows. In Section 2 , we briefly review the classical