Artificial Intelligence based Network Intrusion Detection with Hyper-Parameter Optimization Tuning on the Realistic Cyber Dataset CSE-CIC-IDS2018 using Cloud Computing

Kanimozhi, V.; Jacob, T. Prem

doi:10.1109/iccsp.2019.8698029

Cited by 106 publications

(31 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The intrusion prediction performance of proposed attention-based RCNN is compared with different state-of-art methods such as ANN, 19 PSO, 22 HGGWO, 23 and DBN-IDS. 24 When compared with all other existing algorithms, the proposed work delivers betters and optimal performance outputs in terms of specificity, accuracy, and sensitivity.…”

Section: Comparison Analysis Using State-of-art Methodsmentioning

confidence: 99%

“…Kanimozhi and Jacob 19 used artificial neural network (ANN), which plays a vital role to detect a malicious attack in the cloud system. Botnet attacks are detected by ANN classifier in financial and banking service-related attacks.…”

Section: Review Of Related Workmentioning

confidence: 99%

“…So, the best fittest reproduction R * and pollination are ensured. The flower constancy of FPA rule 1 is numerically explained in Equation (19).…”

Section: Flower Pollination Algorithm (Fpa)mentioning

confidence: 99%

See 2 more Smart Citations

Integration of recurrent convolutional neural network and optimal encryption scheme for intrusion detection with secure data storage in the cloud

Prabhakaran

Kulandasamy

2020

Computational Intelligence

View full text Add to dashboard Cite

Data communication security is growing day after day with the proliferation of cloud computing. It is primarily because of the few security constraints and challenges occurring in the cloud environment during data transmission. Existing research has shown that the intrusion detection system (IDS) centered on the cloud is more complicated. In this article, we address the above issues by proposing an attention-based recurrent convolutional neural network (RCNN). This proposed RCNN is used to detect whether the text data are intrusion or nonintrusion. The nonintrusion text information is then used for further processing and encrypted using a two-way encryption scheme. We introduce the elliptical curve cryptography (ECC) approach to increase the security-level performance of nonintrusion data. Moreover, the integration of ECC with the modified flower pollination algorithm (MFP-ECC) creates the two-way encryption scheme, and it is used to produce an optimal private key. The encrypted data are then stored in a cloud environment by steganography and the data with the sensitive information are replaced by some other text, thus providing security to the data at rest. The proposed MFP-ECC approach shows maximum breaking time results and can also withstand different classical 344

show abstract

Section: Comparison Analysis Using State-of-art Methodsmentioning

confidence: 99%

Section: Review Of Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Integration of recurrent convolutional neural network and optimal encryption scheme for intrusion detection with secure data storage in the cloud

Prabhakaran

Kulandasamy

2020

Computational Intelligence

View full text Add to dashboard Cite

show abstract

“…Kanimozhi et al [88] proposed Artificial Neural Networks by using an up-to-date cybersecurity dataset (CSE-CIC-IDS2018). The proposed approach provided an outstanding performance of Accuracy and average area under the ROC (Receiver Operator Characteristic) curve, and the average False Positive rate.…”

Section: Cse-cic-ids2018 Datasetmentioning

confidence: 99%

A Detailed Analysis of Benchmark Datasets for Network Intrusion Detection System

Ghurab

Gaphari

Alshami

et al. 2021

AJRCoS

View full text Add to dashboard Cite

The enormous increase in the use of the Internet in daily life has provided an opportunity for the intruder attempt to compromise the security principles of availability, confidentiality, and integrity. As a result, organizations are working to increase the level of security by using attack detection techniques such as Network Intrusion Detection System (NIDS), which monitors and analyzes network flow and attacks detection. There are a lot of researches proposed to develop the NIDS and depend on the dataset for the evaluation. Datasets allow evaluating the ability in detecting intrusion behavior. This paper introduces a detailed analysis of benchmark and recent datasets for NIDS. Specifically, we describe eight well-known datasets that include: KDD99, NSL-KDD, KYOTO 2006+, ISCX2012, UNSW-NB 15, CIDDS-001, CICIDS2017, and CSE-CIC-IDS2018. For each dataset, we provide a detailed analysis of its instances, features, classes, and the nature of the features. The main objective of this paper is to offer overviews of the datasets are available for the NIDS and what each dataset is comprised of. Furthermore, some recommendations were made to use network-based datasets.

show abstract

“…As a new computing technology, artificial intelligence algorithm is rising. And digitization, networking and informatization are also increasingly integrated into people's life, which fundamentally change the living state of human beings [1]. In the network, big data, Internet of things, and the support of technologies such as artificial intelligence develop rapidly, as far as possible to meet the diverse needs of people, enhance and improve people's living conditions.…”

Section: Introductionmentioning

confidence: 99%

Energy-Efficient Joint Resource Allocation Algorithms for MEC-Enabled Emotional Computing in Urban Communities

Yang

Che

et al. 2019

IEEE Access

View full text Add to dashboard Cite

This paper considers a mobile edge computing (MEC) system, where the MEC server first collects data from emotion sensors and then computes the emotion of each user. We give the formula of the emotional prediction accuracy. In order to improve the energy efficiency of the system, we propose resources allocation algorithms. We aim to minimize the total energy consumption of the MEC server and sensors by jointly optimizing the computing resources allocation and the data transmitting time. The formulated problem is a non-convex problem, which is very difficult to solve in general. However, we transform it into convex problems and apply convex optimization techniques to address it. The optimal solution is given in closed form. Simulation results show that the total energy consumption of our system can be effectively reduced by the proposed scheme compared with the benchmark.

show abstract

Artificial Intelligence based Network Intrusion Detection with Hyper-Parameter Optimization Tuning on the Realistic Cyber Dataset CSE-CIC-IDS2018 using Cloud Computing

Cited by 106 publications

References 12 publications

Integration of recurrent convolutional neural network and optimal encryption scheme for intrusion detection with secure data storage in the cloud

Integration of recurrent convolutional neural network and optimal encryption scheme for intrusion detection with secure data storage in the cloud

A Detailed Analysis of Benchmark Datasets for Network Intrusion Detection System

Energy-Efficient Joint Resource Allocation Algorithms for MEC-Enabled Emotional Computing in Urban Communities

Contact Info

Product

Resources

About