Back to Search
Start Over
A High-Speed, Scalable, and Programmable Traffic Manager Architecture for Flow-Based Networking
- Source :
- IEEE Access, Vol 7, Pp 2231-2243 (2019)
- Publication Year :
- 2019
- Publisher :
- IEEE, 2019.
-
Abstract
- In this paper, we present a programmable and scalable traffic manager (TM) architecture, targeting requirements of high-speed networking devices, especially in the software-defined networking context. This TM is intended to ease the deployability of new architectures through field-programmable gate array (FPGA) platforms and to make the data plane programmable and scalable. Flow-based networking allows treating traffic in terms of flows rather than as a simple aggregation of individual packets, which simplifies scheduling and bandwidth allocation for each flow. Programmability brings agility, flexibility, and rapid adaptation to changes, allowing to meet network requirements in real-time. Traffic management with fast queuing and reduced latency plays an important role to support the upcoming 5G cellular communication technology. The proposed TM architecture is coded in C++ and is synthesized with the Vivado High-Level Synthesis tool. This TM is capable of supporting links operating beyond 40 Gb/s, on the ZC706 board and XCVU440-FLGB2377–3-E FPGA device from Xilinx, while achieving 80 Gb/s and 100 Gb/s throughput, respectively. The resulting placed and routed design was tested on the ZC706 board with its embedded ARM processor controlling table updates.
- Subjects :
- General Computer Science
Computer science
020209 energy
Traffic manager (TM)
Throughput
02 engineering and technology
Scheduling (computing)
high-level synthesis (HLS)
field-programmable gate array (FPGA)
0202 electrical engineering, electronic engineering, information engineering
Forwarding plane
General Materials Science
Queueing theory
Network packet
business.industry
flow-based networking
General Engineering
Networking hardware
Bandwidth allocation
Embedded system
Scalability
020201 artificial intelligence & image processing
data plane
lcsh:Electrical engineering. Electronics. Nuclear engineering
business
lcsh:TK1-9971
Subjects
Details
- Language :
- English
- ISSN :
- 21693536
- Volume :
- 7
- Database :
- OpenAIRE
- Journal :
- IEEE Access
- Accession number :
- edsair.doi.dedup.....1d1b5fd02fbdfb36baa6ee82d02c2413