2020 International Wireless Communications and Mobile Computing (IWCMC) 2020
DOI: 10.1109/iwcmc48107.2020.9148369
|View full text |Cite
|
Sign up to set email alerts
|

Dynamically Split the Traffic in Software Defined Network Based on Deep Reinforcement Learning

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
2
2

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(3 citation statements)
references
References 11 publications
0
3
0
Order By: Relevance
“…A load balancing policy was defined to characterize the distribution of the traffic over the best n paths. Some other works have targeted at controlling the traffic split ratio to multiple paths on top of the SDN controller (see, e.g., [23] and [24]). However, routing large flows over several paths might create packet out of order problems.…”
Section: Related Work and Contributionsmentioning
confidence: 99%
“…A load balancing policy was defined to characterize the distribution of the traffic over the best n paths. Some other works have targeted at controlling the traffic split ratio to multiple paths on top of the SDN controller (see, e.g., [23] and [24]). However, routing large flows over several paths might create packet out of order problems.…”
Section: Related Work and Contributionsmentioning
confidence: 99%
“…The authors in [22] adapt the DDPG algorithm to find the optimal scheduling scheme for flows. The authors in [23,24] present a QoS optimization algorithm based on DDPG that ultimately improves the load-balancing degree and throughput rate to ensure delay and packet-loss rate. The authors in [25] propose an intelligent traffic-sampling system based on DDPG that can maintain the load balance of multiple traffic analyzers.…”
Section: Related Researchmentioning
confidence: 99%
“…Reinforcement learning opens a new way for solving complex network problems [15]. Some researchers have used traditional algorithms of reinforcement learning such as deep Q-learning network (DQN) [16][17][18][19], proximal policy optimization (PPO) [20], deep deterministic policy gradient (DDPG) [21][22][23][24][25][26][27][28], and twin delayed deep deterministic policy gradient (TD3) [29][30][31]. The DQN algorithm uses Q-tables to store value functions, but it leads to excessive memory overhead and sizeable computational complexity when the network size increases.…”
Section: Introductionmentioning
confidence: 99%