Adaptive Software-Defined Network Control Using Kernel-Based Reinforcement Learning: An Empirical Study

Software-defined networking (SDN) requires adaptive control strategies to handle dynamic traffic conditions and heterogeneous network environments. Reinforcement learning (RL) has emerged as a promising solution, yet deep RL methods often face instability, non-stationarity, and reproducibility chall...

Full description

Bibliographic Details

Published in:	Applied sciences Vol. 15; no. 23; p. 12349
Main Authors:	Nurakhov, Yedil, Kyzyrkanov, Abzal, Otarbay, Zhenis, Lebedev, Danil
Format:	Journal Article
Language:	English
Published:	21.11.2025
ISSN:	2076-3417, 2076-3417
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!