Dynamic Markov Model for Resource Contention Prediction in an Edge Computing Architecture
AI & Cyber Forum: An International Journal · 2025
DOI: 10.51470/ai.2025.4.2.18 ↗Kurz gefasst
Fachartikel in der Zeitschrift AI & Cyber Forum: An International Journal, 2025. Verfasst von Madhukumar K und 1 weiteren Autor:innen.
Abstract (Originalsprache)
Edge computing seeks to minimize processing and storage demands on the cloud while delivering services with decreased latency. Devices near the network’s periphery include smartphones, tablets, routers, surveillance footage, presses, and other IoT gadgets that generate copious amounts of data. Virtual machines (VMs) located on edge servers close to the edge devices handle the data. In order to anticipate power congestion at edge computers, this research presented two models: the Dynamic Markov Chain with Regression and Markov (DMRCP) and the Hybrid Cascading of Reconstruction and Markov (CRMRCP). In order to evaluate performance, the research examines the amount of virtual machine migrations and compares the DMRCP method with initial-order and second-degree Markov models. We compare the number of virtual machine deployments using the CRMRCP method to that using the third-order Poisson model. Based on the findings, the DMRCP requires fewer virtual machine migrations compared to the first-order equations Markov algorithm (52.9%) and the second-degree Population model (21.1%). By combining ridge regression with the second-degree Markov model, the number of virtual machine migrations in CRMRCP is reduced by 81.8%.
Zitieren
APA
K, M., & Sidana, P. (2025). Dynamic Markov Model for Resource Contention Prediction in an Edge Computing Architecture. AI & Cyber Forum: An International Journal. https://doi.org/10.51470/ai.2025.4.2.18
Vancouver
K M, Sidana P. Dynamic Markov Model for Resource Contention Prediction in an Edge Computing Architecture. AI & Cyber Forum: An International Journal. 2025. doi:10.51470/ai.2025.4.2.18
BibTeX
@article{K2025,
author = {K, Madhukumar and Sidana, Prerna},
title = {Dynamic Markov Model for Resource Contention Prediction in an Edge Computing Architecture},
journal = {AI & Cyber Forum: An International Journal},
year = {2025},
doi = {10.51470/ai.2025.4.2.18}
}
EUVIMED