Distributed Denial of Service (DDoS) attacks represent an important challenge for public cloud as they invade the offender and completely delete Cloud service in order to serve the correct user and at the same time against the targets that cause system and service lack of access on infected devices. DDoS (Distributed Denial of Service) attacks are usually specific efforts to drain the resources of the victim or interrupt connections to networks by legitimate users. Traditional internet infrastructure is susceptible to DDoS assaults, and through leveraging their flaws to set up assault networks or Botnets, it offers an opening for an intruder to reach a wide number of infected machines. In order to identify and sustain improved detection accuracy, this work focuses on evaluating the different works and recommending a better solution to accommodate the cloud environment. A fuzzy logic for the detection and safety of DDOS attacks is proposed in this paper. The Fuzzy logic is used to dynamically select an algorithm from a collection of defined supervised learning that distinguish various DDoS variations and ultimately choose the relevant traffic algorithm.