1. We want to minimize the controller overhead problem by increasing the matching flow entry matching rate, so that most of the time we would be able to find relevant entries inside the flow table.
2. There will be three main cases. First when we have all flow tables in all switches are empty and analyze the traffic patters between switch and controller. Second when we have vacant spaces in flow tables and analyze the traffic. Third when we have all flow tables full and analyze the traffic.
3. An LRU caching algorithm would be required.
4. Two different types of traffics are needed i.e. Mice and Elephant for different analysis.
5. At the end, three different types of results are needed for observation. One by changing in cache size, second by changing in number of hosts and third by changing in the number of incoming packets.
6. I have base paper for experimentation format.
7. MININET tool for experimentation/implementation and PYTHON for algorithm development are required.
8. Delivery time is 6-8 weeks maximum.