Flow-based anomaly detection in high-speed links using modified GSA-optimized neural network

Sheikhan, Mansour; Jadidi, Zahra

doi:10.1007/s00521-012-1263-0

Cited by 52 publications

(29 citation statements)

References 84 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…K-Map is useful for distinction of the data but it is not supported for large volume of data. Support Vector Machine (SVM) [19][20][21][22][23] [ [25][26][27][28][29][30][31][32]: is beneficial for classification and reformation of data, but at the same time SVM only cannot powerfully in recognizing about new data. Given a set of trained-data-set which require more computational time for big-data [4,12].…”

Section: B Related Workmentioning

confidence: 99%

Fuzzy Controlled Network Intrusion Detection System (FC-NIDS)

Kumar*,

Kumar

2019

IJITEE

View full text Add to dashboard Cite

 Abstract: Intrusion Detection System (IDS) is the nearly all imperative constituent of computer network security. IDSs are designed to comprehend intrusion attempts in incoming network traffic shrewdly. It deals with big volume of data containing immaterial and outmoded features, which lead to delay in training as well as testing procedures. Therefore, to minimize the false alarm and computation complexity, the features selection technique for intrusion detection has been implemented. In this paper PCA (Principal Component Analysis) and Fuzzy Inference System (FIS) have been used on kdd99 dataset to develop FC-NIDS model. PCA is used to select the attacked features to minimize the computational work, while FIS is used to develop a fuzzy inference system for accuracy in prophecy using MATLAB. The results of the experiment are tested on UCI data sets as a standard bench-mark. It has been found efficient for true prediction of intrusion as well as to reduce the false alarm rate. The proposed fuzzy logic controller IDS (FC-NIDS), is passable to covenant with signature and anomaly based attacks to get enhanced intrusion detection, decreases false alarm and to optimize complexity.

show abstract

Section: B Related Workmentioning

confidence: 99%

Fuzzy Controlled Network Intrusion Detection System (FC-NIDS)

Kumar*,

Kumar

2019

IJITEE

View full text Add to dashboard Cite

show abstract

“…Classification-based anomaly detection methods assume that a classifier that can distinguish between normal and anomalous classes, can be learnt in the given feature space. According to the different ways of training the classifier, classification-based anomaly detection approaches consist of rule-based approaches [20,30,35], neural network-based approaches [31,32], Bayesian networkbased approaches [5,10,14,23], and support vector machine-based approaches [19,33]. -Nearest neighbor-based.…”

Section: Related Workmentioning

confidence: 99%

An anomaly detection approach for multiple monitoring data series based on latent correlation probabilistic model

et al. 2015

View full text Add to dashboard Cite

Condition monitoring systems are widely used to monitor the working condition of equipment, generating a vast amount and variety of monitoring data in the process. The main task of surveillance focuses on detecting anomalies in these routinely collected monitoring data, intended to help detect possible faults in the equipment. However, with the rapid increase in the volume of monitoring data, it is a nontrivial task to scan all the monitoring data to detect anomalies. In this paper, we propose an approach called latent correlation-based anomaly detection (LCAD) that efficiently and effectively detects potential anomalies from a large number of correlative isomerous monitoring data Jianwei Ding series. Instead of focusing on one or more isomorphic monitoring data series, LCAD identifies anomalies by modeling the latent correlation among multiple correlative isomerous monitoring data series, using a probabilistic distribution model called the latent correlation probabilistic model, which helps to detect anomalies according to their relations with the model. Experimental results on real-world data sets show that when dealing with a large number of correlative isomerous monitoring data series, LCAD yields better performances than existing anomaly detection approaches.

show abstract

“…The mass with minimum MSE will be the best. Training is finished if it achieves an acceptable error or maximum number of iterations [22].…”

Section: Proposed Flow Anomaly Detection Systemmentioning

confidence: 99%

“…The results show that GSA outperforms PSO and gradient based algorithms. Also, a Modified GSA (MGSA) is employed in [22] to optimize the interconnection weights of a neural network anomaly detector. The results confirm that MGSAbased method can monitor flow traffic in high speed networks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Metaheuristic algorithms based Flow Anomaly Detector

Jadidi

Muthukkumarasamy

Sithirasenan

2013

2013 19th Asia-Pacific Conference on Communications (APCC)

Self Cite

View full text Add to dashboard Cite

Increasing throughput of modern high-speed networks needs accurate real-time Intrusion Detection System (IDS). A traditional packet-based Network IDS (NIDS) is timeintensive as it inspects all packets. A flow-based anomaly detector addresses scalability issues by monitoring only packet headers.This method is capable of detecting unknown attacks in high speed networks. An Artificial Neural Network (ANN) is employed in this research to detect anomalies in flow-based traffic. Metaheuristic optimization algorithms have the potential to achieve global optimal solution. In this paper, two metaheuristic algorithms, Cuckoo and PSOGSA, are examined to optimize the interconnection weights of a Multi-Layer Perceptron (MLP) neural network. This optimized MLP is evaluated with two different flow-based data sets. We then compare the performance of these algorithms. The results show that Cuckoo and PSOGSA algorithms enable high accuracy in classifying benign and malicious flows. However, the Cuckoo has lower training time.

show abstract

Flow-based anomaly detection in high-speed links using modified GSA-optimized neural network

Cited by 52 publications

References 84 publications

Fuzzy Controlled Network Intrusion Detection System (FC-NIDS)

Fuzzy Controlled Network Intrusion Detection System (FC-NIDS)

An anomaly detection approach for multiple monitoring data series based on latent correlation probabilistic model

Metaheuristic algorithms based Flow Anomaly Detector

Contact Info

Product

Resources

About