The agenda for the seminar this week is listed below:
Title: Microboxes: Localized evolving semi-supervised learning based anomaly detection scheme for wireless multi-hop software defined networks
Abstract:
Massive machine-type-communications (mMTC) enables intelligent networking and brings added-value in various aspects of our everyday lives, such as smart home and e-health monitoring system. Software defined network (SDN) is a promising technology to facilitate management on mMTC devices. However, the capability-constrained mMTC devices and current centralized anomaly defense systems open security and scalability concerns in the future network.
This seminar will focus on solving the scalability issues in the current anomaly defense system via localization. This seminar will introduce a Localized Evolving Semi-supervised Learning Based Anomaly detection scheme (LESLA) to detect abnormal traffics without overwhelming the controller or periodically generating a huge volume of traffic in wireless SDN enabled mMTC network. Besides, LESLA considers the feasibility of executing anomaly defense algorithms mMTC devices, the self-evolving characteristics and the risk of performance degradation in the case of ‘outlier’.
For more details, please go to the link
http://networks.eecs.qmul.ac.uk/category/seminar/
Venue: Engineering Building- Bancroft 1.01
Time and Date: 2 to 3 PM, Thursday, 17/10/2018
Please keep the dates on your calendar.
Any comments, advice would be appreciated.