Multischeme feedforward artificial neural network architecture for DDoS attack detection

Abstract

Distributed denial of service attack classified as a structured attack to deplete server, sourced from various bot computers to form a massive data flow. Distributed denial of service (DDoS) data flows behave as regular data packet flows, so it is challenging to distinguish between the two. Data packet classification to detect DDoS attacks is one solution to prevent DDoS attacks and to maintain server resources maintained. The machine learning method especially artificial neural network (ANN), is one of the effective ways to detect the flow of data packets in a computer network. Based on the research that has carried out, it concluded that ANN with hidden layer architecture that contains neuron twice as neuron on the input layer (2n) produces a stable detection accuracy value on Quasi-Newton, Scaled-Conjugate and Resilient-Propagation training functions. Based on the studies conducted, it concluded that ANN Architecture sufficiently affected the Scaled-Conjugate and Resilient-Propagation training functions, otherwise the Quasi-Newton training function. The best detection accuracy achieved from the experiment is 99.60%, 1.000 recall, 0.988 precision, and 0.993 f-measure using the Quasi-Newton training function with 6-(12)-2 neural network architecture.

K-nearest neighbor algorithm analysis for path determination in network simulation using software defined network

“…However, the arrival of the package may not be together. This time lag is called jitter [31]. Jitter in IP networks is the variation in the latency on a packet flow between two systems when some packets take longer to travel from one system to the other.…”

Gradient Techniques To Predict Distributed Denial-Of-Service Attack

“…RNN is structured to apply the same set of weights to variable-sized input structures, create predictions, or traverse specified structures in topological order, resulting in a neural network that produces predictions [11]. What are dynamic recursive neural networks?…”

Semi-supervised approach for detecting distributed denial of service in SD-honeypot network environment

“…Previously, several researchers have already investigated the capability of DDoS for overloading traditional networks. The former concept for detecting DDoS was categorized into two approaches namely the statistic [8], [9] and artificial intelligence (AI) [10]- [14]. Several statistics approaches have been deployed e.g., the Entropy [8] which calculated the data randomness and specified the DDoS threshold by its value, Bloom-filter [9] which focused the detection phase by comparing the hash value of the incoming packet to assure the packet was not considered as SYN flood attack.…”

“…The results stated the Naïve Bayes algorithm could predict the outcomes precisely even though there was no apparent result for the classification metric. Several papers also conducted the classification based on available datasets [12]- [14] (NSL-KDD, UNB ISCX 12, and UNSW-NB15). Idhammad et al [12] proposed the Semisupervised learning for classifying the DDoS attack gained 98.23% for accuracy.…”

“…Idhammad et al [12] proposed the Semisupervised learning for classifying the DDoS attack gained 98.23% for accuracy. Mohammed et al [13] and Muhammad et al [14] utilized deep learning (DL) for detecting DDoS and achieved an accuracy at 97.82% and 99.60% respectively. The programmability feature in SDN may provide manageable structure for implementing AI to detect DDoS.…”