IoT security using Machine Learning and SDN

There are three modes in which the script can be run

1. CONTROLLER MODE

• In this mode, the controller listens to incoming features from switch and predicts whether the traffic is malicious.
• If malicious traffic pattern is found, the source of that traffic is blocked by installing appropriate flow table in switch using Floodlight SDn controller

2. SWITCH MODE

• In this mode, the script extracts features from traffic flowing through switch and sends it to the controller.
• Sampling interval for feature extraction can be set in constant.py file

3. TRAIN MODE

• In train mode, features are extracted from all the pcaps mentioned in constant.py file and saved.
• Using these features, various classifiers are trained and saved.