Diffusion limit for the stationary distribution of a history-dependent two-level M/M/ 1 1 queue

Masahiro Kobayashi, Masakiyo Miyazawa, Yutaka Sakuma · 2026

Read Diffusion limit for the stationary distribution of a history-dependent two-level M/M/ 1 1 queue by Masahiro Kobayashi, Masakiyo Miyazawa, Yutaka Sakuma (2026) 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-2601-02705.txt.

Description

Masahiro Kobayashi, Masakiyo Miyazawa, and Yutaka Sakuma extend a classic queueing model in a direction motivated by capacity control. In the multi level queue they build on, arrival and service rates depend on which level the current queue length falls in. And earlier work proved such queues look like reflected Brownian motion under heavy traffic, with a matching diffusion limit for the steady state distribution. The new question is what changes when control also depends on the recent history of the queue, not just its present length.

As a first step they specialize to a two level queue with Poisson arrivals and exponential service. The dynamics are extended so rates depend on history as well as current length. Under a stability condition they derive the diffusion limit of the stationary distribution for this history dependent case, showing how memory in the control rule reshapes the long run queue.

Catalog record

CREATORMasahiro Kobayashi, Masakiyo Miyazawa, Yutaka Sakuma
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE56 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-02705

Browse the full library · About the Tower of Babel