Computer Science Independent Study Defense

04:00 PM - 05:00 PM
Jun Zheng

Eric Binnendyk will present his independent study defense. 

His talk is tilted "An intelligent token bucket-based queue management strategy for QoS of named-data networking"


Named data networking (NDN) is a representative architecture for information-centric networking (ICN), which has been shown as a promising solution for supporting the communication needs of a variety of smart systems such as smart grids, smart cities, and smart homes. Compared with traditional host-centric IP-based networking, NDN is based on the content-centric communication model where the data objects are retrieved by names instead of delivered to a destination address.
One of the major challenges faced by NDN is how to provide differential quality of service (QoS) to network traffics with different priority levels. In this report, we introduce a new intelligent queuing strategy for QoS based on a token bucket strategy. The token bucket rates for queues with different priority levels are adaptively adjusted with a piecewise linear loss function according to the traffic condition. Simulations are performed by using the ndnSIM simulator. The results show that the proposed strategy has better performance than the strategies based on fixed token bucket rates.

Zoom info: