.png){kind=icon}
返回
SDN-Based Application-Aware Segment Routing for Large-Scale Network
.png){kind=icon}
PRE
AI
DOI：10.1109/JSYST.2021.3123809.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
With the rapid growth of fifth-generation (5G), network operators have met with difficulties in ensuring user satisfaction due to diverse service-level agreement (SLA) requirements. Application-aware routing is able to potentially address this issue in implementing differentiated routing policies corresponding to various applications. To facilitate the application-aware routing, software-defined networking (SDN) and segment routing (SR) are potential solutions with programmability property to steer network traffic into appropriate routing paths. However, application-aware SR encounters two main problems, including application identification for encrypted traffic and path identification with human intervention. Therefore, this article proposes a new SDN-based SR mechanism that could help network operators overcome these problems. This approach implements reinforcement learning to adapt to dynamic networks in order to optimize the quality of experience (QoE) that network operators must guarantee. The proposal also considers the application class to implement corresponding routing policies for various applications to meet the stringent SLA requirements. Identifying the applications is sometimes complicated due to encrypted network traffic. Hence, our proposal implements a traffic classification approach to classify encrypted traffic into different kinds of applications. Obtained results under considered conditions illustrate that our proposal outperforms the standard SR mechanism related to QoE and decreases up to 64.39% of overhead compared to several benchmarks.
Keyword:
Routing
Quality of experience
Cryptography
Multiprotocol label switching
Quality of service
Packet loss
Monitoring
Application-aware routing
quality of experience (QoE)
software-defined networking (SDN)
segment routing
traffic classification
期刊
I
IF:
2.4
论文数:
4.5K
被引数:
387
机构
引用论文
Reinforcement Learning Based Routing in Networks: Review and Classification of Approaches.png)
IEEE ACCESS
OAAI
IEEE ACCESSIF3.6
OAAI分享收藏Datanet: Deep learning Based Encrypted Network Traffic Classification in SDN Home GatewayDatanet: 基于深度学习的SDN家庭网关加密网络流量分类IEEE ACCESSOAAI
IEEE ACCESSIF3.6
OAAI分享收藏Segment Routing: A Comprehensive Survey of Research Activities, Standardization Efforts, and Implementation Results细分市场路由: 对研究活动、标准化工作和实施结果的全面调查
PREAI
分享收藏LogEvent2vec: LogEvent-to-Vector Based Anomaly Detection for Large-Scale Logs in Internet of ThingsSENSORSOAAI
SENSORSIF3.5
OAAI分享收藏Towards sFlow and adaptive polling sampling for deep learning based DDoS detection in SDNSDN中基于深度学习的DDoS检测的sFlow和自适应轮询采样OAAI
OAAI分享收藏
查看更多内容