VCROA-DNFN: A BIG DATA APPROACH IN MAPREDUCE FRAMEWORK FOR DDOS ATTACK DETECTION USING OPTIMIZED DEEP NEURO FUZZY NETWORK
Keywords:
Distributed Denial of Service (DDoS), MapReduce, Intrusion Detection System (IDS), Deep Neuro Fuzzy Network (DNFN), Velocity Contour based Remora Optimization Algorithm (VC-ROA).Abstract
Recently, Distributed Denial of Service (DDoS) attack has become a crucial warning to network privacy and defending against DDoS attack is a hot research area among industry and academia. Various methodologies have been developed so far to identify the attacks and to mitigate the consequences they result. But still now, a plethora of techniques cannot provide effectiveidentification accuracy owing to the small number of false alarms. Such problems can be encountered easily through deep learning advancements as they are accurate and efficacious algorithm to discriminate genuine and threaten data. Here, an optimized deep learning approach named Velocity Contour based Remora Optimization (VCROA)-Deep Neuro Fuzzy network (DNFN) is introduced for DDoS attack identification in MapReduce framework, a big data approach. Here, the optimal features are effectively selected utilizing VCROA. Furthermore, the DDoS attack is identified accurately with the utilization of DNFN and the weights are finely adjusted based on VCROA, which is a consolidation of velocity contour based concept into Remora Optimization Algorithm (ROA). The proposed VCROA-DNFN for DDoS attack detection has gained maximum precision, recall and F-measure of 91.10%, 93.70%, and 91.70%.