Overview

Skyfilter mitigation engine with machine learning and behavior analysis, does not require any pre-configuration to start the protection; when attacked, Skyfilter Anti-DDoS automatically detects the malicious traffic and starts mitigation up to 2Tbps Distributed Denial of Service (DDoS) attacks remain a top threat to IT security and have evolved in almost every way to do what they do best: shut down access to your online services. Skyfilter uses vital technologies -inorder to protect from malicious ddos attack- such as machine learning, packet analysis, anomaly detection, behavioral detection. Structurally ,machine learning and packet analysis are performed at network adapter level.
Functional detection and anomaly detection techniques are applied at kernel level. Dedicated drivers using the Intel chipset cards are designed to use machine learning algorithms for maximum detection and mitigation thresholds. This inturn creates a more stable and secure environment due to the fact that standard kernel operating on the FreeBsd operating system. When an attack is detected in the system, algorithms, very powerful tools, are activated to isolate harmful packets and does not block the IP addresses unlike our competitors technologies, this reduces false positive events significantly.
In the event of a possible attack the system uses various markers on the incoming packets (Packet checksum, Packet identification number, IP header, Fragment offset, Packet size, TTL, ToS, TCP sequence number, Acknowledgment Number, UDP Header, TCP Header, Dns Qcount, Dns Qname), so if the incoming request is a confirmed of malicious intent, the system can respond against said harmful packages more efficiently and effectively. Unlike the usual approach of the industry, the system does not sign the attack at the time of the attack, instead the system uses temporary rule signatures that are automatically created to isolate and overcome the issue. The system uses innovative Skyfilter algorithms in tandem with industry leading Naive Bayes and Support Vector Machine learning algorithms which can make various analyzes and packet classification. With these leading technologies we created a cutting edge automatic intrusion detection and blocking. The technology we have developed can be integrated into cloud systems and is ready to provide global protection.