2020
DOI: 10.1109/access.2020.2996660
|View full text |Cite
|
Sign up to set email alerts
|

Flexible Software-Defined Packet Processing Using Low-Area Hardware

Abstract: Computer networks are in the Software Defined Networking (SDN) and Network Function Virtualization (NFV) era. SDN brings a whole new set of flexibility and possibilities into the network. The data plane of forwarding devices can be programmed to provide functionality for any protocol, and to perform novel network testing, diagnostics, and troubleshooting. One of the most dominant hardware architectures for implementing the programmable data plane is the Reconfigurable Match Tables (RMT) architecture. RMT's inn… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
5
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
6
3

Relationship

1
8

Authors

Journals

citations
Cited by 13 publications
(5 citation statements)
references
References 24 publications
0
5
0
Order By: Relevance
“…Refs. [21,22] also used deep pipelines to implement parsers. Through pipeline iterations, header fields of any length can be effectively analyzed.…”
Section: Related Workmentioning
confidence: 99%
“…Refs. [21,22] also used deep pipelines to implement parsers. Through pipeline iterations, header fields of any length can be effectively analyzed.…”
Section: Related Workmentioning
confidence: 99%
“…We also add an FP32/FP16 Floating Point Unit (FPU) [55] to each core, increasing the area of the cluster to 2.29mm 2 . The processing unit in programmable switches is estimated to occupy up to 140mm 2 in a 28nm FDSOI process [56]. By scaling the area, we set a 180mm 2 target area for our PsPIN unit, and we then assume we can fit ∼64 clusters and the L2 memory in our processing unit area budget.…”
Section: L2mentioning
confidence: 99%

Flare: Flexible In-Network Allreduce

De Sensi,
Di Girolamo,
Ashkboos
et al. 2021
Preprint
“…Constraint (11) ensures that computational resources required at node i to perform all VNFs from all sources and all the paths are not exceeded. Constraint (12) finds the primary (i) and backup (j) nodes performing the different VNFs for each source. Constraint (13) ensures that the capacity reserved for the backup node j is greater than or equal to the capacity required to perform primary functions at node i, to ensure reliability against single primary hotel failure.…”
Section: Variable Definitionmentioning
confidence: 99%
“…In this way, the baseband is able to control thousands of cells, increasing both network efficiency through dynamic management of resources to respond to rapid traffic changes. In addition, baseband virtualization through paradigms as Network Functions Virtualization (NFV) and Software-Defined Networking (SDN) introduce agility, flexibility, reliability and security [12]. The application of these paradigms is called network slicing, which allows the coexistence of multiple classes of service with different KPIs within the same physical network.…”
Section: Introductionmentioning
confidence: 99%