作者: Firoz Kaderali , Thorsten Kisner , Alex Essoh
DOI:
关键词:
摘要: Traffic modeling and classification is important in many areas like intrusion detection, anomalous traffic network planning or bandwidth management. A novelty the work presented this paper use of texture methods from domain digital image processing for classification. We strategies based on Co-occurence Matrices to derive statistical properties Using well known kNN-Classificator we are able distinguish different classes with a high probability.