Hybrid load balancing mechansimby using online traffic classification means (built classifier in switch and controller to classifers the traffic to (elephant flow and mice flow)) in software defined networking (SDN) via hierarchical controllers

Firstly, I want to implement online traffic classification means (built classifier in switch and controller) by (semi-supervised techniques on incremental kmeans) that classifies service types to (elephant flow and mice flow) in software defined networking (SDN). Secondly, After I will use the input types of traffic (elephant flow or mice flow) and other parameters bandwidth utilization ratio, packet loss rate, transmission latency and transmission hops to a load balancing technique, to handle high controller load in data plane in SDN, and used fives parameters inputs to a hybrid techniques ( artificial neural Network and ant colony optimization) will be use to develop Load Balancing mechanism to obtain the best route in data plane. Finally, the overall system will be implemented in hierarchical controllers Load Balancing by using Dynamic Adaptive load balancing Algorithm, in control plane. All this work will be implement in implemented in mininet simulation by using Floodlight controller. ). The effectiveness of the proposed method will be verified through available real network traffic datasets from various sources, and it will be used to benchmark the performance of the algorithm.