Survey on distributed denial of service attack using deep learning approaches

Nowadays, numerous challenges in cyberspace contribute to the emergence of network security issues. The identification of irregularities in traffic data is critical to the detection of hostile activity inside a network, which is necessary for maintaining the integrity of current Cyber- Physical Systems (CPS) as well as network security. One of the most prevalent and successful types of attacks that aims to prevent or impair the service delivery of its victim(s) is the distributed denial-of-service (DDoS) attack. For non- distributed attacks, the attack detection systems identify a source node with a high volume of packet transmission. DDoS attacks are challenging to identify or stop; thus, many academics have recently concentrated on them. The detection procedure is prolonged by certain factors, including assaults with low traffic rates, losing the durations of successive anomalies, and having a high number of analysis samples. This paper offers a comprehensive taxonomy of DDoS attacks, provides an overview of high- speed network accuracy assessment parameters, and categories detection approaches. In addition, a qualitative study of the literature is conducted to examine the parameters derived from the taxonomy of irregular traffic pattern identification offered. Suggested study areas emphasize the problems and difficulties associated with DDoS attacks on networks, and help researchers discover and build the best possible solution.