Bloom filter based traffic classification

OData support
Supervisor:
Dr. Simon Csaba
Department of Telecommunications and Media Informatics

My MSc Thesis discusses the design and implementation of a new flow based load balancer solution to handle the ever growing Internet traffic in an optimal way, ensuring that the needed quality is kept. This work is based on the SDN (Software-defined Networking) approach, and it is using the OpenFlow switch v1.3 implementation as a base, on top of which the RRBF (Round-robin Bloom filter) solution is integrated. The system is able to identify IP flows, label them with fingerprints and store them in Bloom filters in a round-robin way, after which send them out on the least loaded ports. The OpenFlow based RRBF has several input parameters, like the number of bloom filters, the number of expected flows, the false positive probability, the time period of shifting the bloom filters, the flow rate ratio and the number of packets for the flow transient eliminator.

Furthermore, my thesis provides an insight to the different parts of the solution, such as an introduction to load balancers, theory of Bloom filters, specification of OpenFlow.

And finally, my thesis shows the test results, discusses the open issues and lists the future plans.

Downloads

Please sign in to download the files of this thesis.