DDoS detection tool

Researchers using Machine learning as a new technique to create a Real-Time Internet of Things(IoT) DDoS detection tool to prevent the DDoS attack from IoT botnets.

IoT botnet attacks are dramatically increasing and conduct distributed denial of service (DDoS) on Internet infrastructure in recent years by various botnets families such as Mirai, HNS,Doubledoor.

Advanced IoT Botnet attacks are bypassing an IoT layered security that leads to taking complete control of the targeting network systems and attackers always find the many ways to bypass it.

Researchers using Machine learning techniques to develop a new  IoT DDoS Detection Tool to detect the suspicious DDoS traffic in real time.

Real-Time DDOS detection tool will perform based on the IoT network behavior such as regular time interval between packets.

In this case, Variety of machine Learning algorithm such as neural networks will be implemented with this tool for high accuracy DDoS detection in IoT network traffic.

This Technique will effectively work for home gateway routers or other network middleboxes to could automatically detect local IoT device sources of DDoS attacks.

Also, at the same time, it works with low-cost machine learning algorithms and traffic data
that is based on the Traffic Flow.

Also, Researchers develop a machine learning pipeline in order to collect the data feature extraction and binary classification for IoT traffic DDoS detection.

machine learning pipeline

Researchers concentrate with two backgrounds to detect IoT Based DDoS Attack

1.Network Anomaly Detection 

Anomaly detection aims to identify patterns in data that do not conform to expected behavior. In the context of this research, anomaly detection techniques may be used to discern attack traffic from regular traffic

2.Network Middlebox Limitations

Network middleboxes have limited memory and processing power, imposing constraints on the algorithmic techniques used for anomaly detection.

Researchers explained in their Research Paper, Our classifiers successfully identify attack traffic with an accuracy higher than 0.999,” the team writes. “We found that random forest, K-nearest neighbors, and neural net classifiers were particularly effective. We expect that deep learning classifiers will continue to be effective with additional data from real-world deployments.