Intellectual Property Office Home University of Cincinnati
News
Technologies
Issued US Patents
UC Technology Startups
Forms
Our Mission
IP Overview
Patents
Copyright
Licensing
Conflicts
Policies and Law
Annual Reports
Staff and Location
Contact Us
Links
UC Directory
IPO Home

Case Number 103027 - Efficient Network Intrusion Detection Scheme Using Dimensionality Reduction Technique for Wired and Wireless Systems

Description:  Intrusion detection is an important aspect of network security. Many current detection systems have trouble providing real-time network intrusion detection, particularly where very high dimensional data is involved. The intrusion detection system of the present invention provides real-time network intrusion detection by projecting the high dimensional dataset to a lower dimensional space using the random projection technique, then performing intrusion detection in the lower dimensional space using a support vector machine (SVM) classifier.
Detecting intrusions in the projected lower dimension reduces the complexity of the underlying algorithms, which makes it more suitable for real time detection. Moreover, lower dimensional data can be stored and transmitted more efficiently than its higher dimensional data, thereby saving system resources.




For more information please contact Geoffrey Pinski at 513-558-5696 or pinskig@ucmail.uc.edu



Back to Technologies
 
Printer Friendly Version of Efficient Network Intrusion Detection Scheme Using Dimensionality Reduction Technique for Wired and Wireless Systems - 103027









Copyright Information © Copyright 2004-2012 university of Cincinnati, Ohio. All rights reserved.