For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Detecting Long Duration Flows without False Negatives
SangWoo LEE Seon-Ho SHIN MyungKeun YOON
IEICE TRANSACTIONS on Communications
Publication Date: 2011/05/01
Online ISSN: 1745-1345
Print ISSN: 0916-8516
Type of Manuscript: LETTER
Category: Network Management/Operation
traffic monitoring, network management, network security,
Full Text: PDF(77.7KB)>>
A new network measurement primitive was recently proposed, known as long duration flows (LDF). LDF deserves special attention for network management and security monitoring. This kind of traffic appears periodically and persistently through a long period, but its total amount of traffic is not necessarily large. This feature makes detection difficult especially when the resources of detection system are limited or the detection should cover high-speed networks. In this paper, we propose a new lightweight data structure and streaming algorithm to detect such traffic.