ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Wire-speed Application Flow Generation in Hardware Platform for Multi-gigabit Traffic Monitoring
Cited 0 time in scopus Download 0 time Share share facebook twitter linkedin kakaostory
저자
최태상, 윤상식, 김상완, 강동원, 이준경, 이경호
발행일
200804
출처
Network Operations and Management Symposium (NOMS) 2008, pp.184-191
DOI
https://dx.doi.org/10.1109/NOMS.2008.4575133
협약과제
08MR2100, 고속회선(100Gbps,25Gbps) 비용정산용 측정 시스템 개발, 이준경
초록
Flow-based traffic monitoring has received strong attention from the traffic measurement research communities. Despite the value provided by flow measurement, its usage has limited for relatively lower speed traffic mainly due to the performance impact it introduces. Recent traffic measurement research has tried to overcome such a limitation of simple flow-based monitoring by utilizing payload inspection for applications signatures or/and by identifying target application group based on common traffic characteristics. Such improvement, however, requires much higher analysis performance, particularly when it comes to the monitoring of high-speed links like 2.5Gbps or higher. Thus, the traditional approach which consists of traffic capture in a dedicated hardware and traffic analysis on a server with database may be unable to meet such harsh requirements. Although a dedicated hardware is used for traffic measurement, the generation of application flows is normally done in the software. In this paper, we propose our novel traffic measurement methodology which pushes application flow generation functionality into the Network Processor (NP) based hardware to meet such requirements. We have embedded packet capturing capability without loss, deep packet inspection, and flow generation into the hardware. We describe major features, design concepts, implementation, and performance evaluation result. ©2008 IEEE.
키워드
Hardware-based flow record generation, Traffic measurement
KSP 제안 키워드
Application flow, Design concept, Flow measurement, Flow-based, Group based, Hardware platform, Measurement Methodology, Multi-Gigabit, Packet capturing, Performance evaluation, Performance impact