@article{Zakareya Mohammed_2020, title={Forensics Based SDN in Data Centers}, volume={37}, url={https://ijcjournal.org/index.php/InternationalJournalOfComputer/article/view/1475}, abstractNote={<p>Recently, most data centers have adopted for Software-Defined Network (SDN) architecture to meet the demands for scalability and cost-efficient computer networks. SDN controller separates the data plane and control plane and implements instructions instead of protocols, which improves the Quality of Services (QoS) , enhances energy efficiency and protection mechanisms . However, such centralizations present an opportunity for attackers to utilize the controller of the network and master the entire network devices, which makes it vulnerable. Recent studies efforts have attempted to address the security issue with minimal consideration to the forensics aspects. Based on this, the research will focus on the forensic issue on the SDN network of data center environments. There are diverse approaches to accurately identify the various possible threats to protect the network. For this reason, deep learning approach will used to detect DDoS attacks, which is regarded as the most proper approach for detection of threat. Therefore, the proposed network consists of mobile nodes, head controller, detection engine, domain controller, source controller, Gateway and cloud center. The first stage of the attack is analyzed as serious, where the process includes recording the traffic as criminal evidence to track the criminal, add the IP source of the packet to blacklist and block all packets from this source and eliminate all packets. The second stage not-serious, which includes blocking all packets from the source node for this session, or the non-malicious packets are transmitted using the proposed protocol. This study is evaluated in OMNET ++ environment as a simulation and showed successful results than the existing approaches.</p>}, number={1}, journal={International Journal of Computer (IJC)}, author={Zakareya Mohammed, Oribe}, year={2020}, month={Mar.}, pages={1–26} }