DDoS Attacks and Defense Mechanisms Using Machine Learning Techniques for SDN

Distributed Denial Of Service Attacks

DDoS is considered as one of the most serious attacks nowadays. A DDoS is a cyber-attack that attempts to block the online services by overwhelming requests for a time period (Gupta et al., 2009). The target system is forced to slow down, crash or shut down by the flooding of a large number of requests. DDoS attacker keeps the system busy for a certain period of time by forcing the system to serve illegitimate requests consequently denying the services to legitimate customers. One of the most important security principles of the CIA (Confidentiality, Integrity, and Availability) model is availability, which is compromised by DDoS attacks. According to a report (Goodin, 2018), a US based service provider was targeted by a 1.7Tbps attack on 5 March 2018. The attacker spoofed its victim’s address and sent a number of packets with ping at a memcached server. The server responded by firing back as much as 50,000 times the data it received. This flooding of traffic was enough to exhaust the server and to deny the services for its legitimate users.