The enormous growth in internet usage has led to the development of different malicious software posing serious threats to computer security. The various computational activities carried out over the network have huge chances to be tampered and manipulated and this necessitates the emergence of efficient intrusion detection systems. The network attacks are also dynamic in nature, something which increases the importance of developing appropriate models for classification and predictions. Machine learning (ML) and deep learning algorithms have been prevalent choices in the analysis of intrusion detection systems (IDS) datasets. The issues pertaining to quality and quality of data and the handling of high dimensional data is managed by the use of nature inspired algorithms. The present study uses a NSL-KDD and KDD Cup 99 dataset collected from the Kaggle repository. The dataset was cleansed using the min-max normalization technique and passed through the 1-N encoding method for achieving homogeneity. A spider monkey optimization (SMO) algorithm was used for dimensionality reduction and the reduced dataset was fed into a deep neural network (DNN). The SMO based DNN model generated classification results with 99.4% and 92% accuracy, 99.5%and 92.7% of precision, 99.5% and 92.8% of recall and 99.6%and 92.7% of F1-score, utilizing minimal training time. The model was further compared with principal component analysis (PCA)-based DNN and the classical DNN models, wherein the results justified the advantage of implementing the proposed model over other approaches.

The evolution in the attack scenarios has been such that finding efficient and optimal Network Intrusion Detection Systems (NIDS) with frequent updates has become a big challenge. NIDS implementation using machine learning (ML) techniques and updated intrusion datasets is one of the solutions for effective modeling of NIDS. This article presents a brief description of publicly available labeled intrusion datasets and ML techniques. Later a brief explanation of the literary works is given in which machine learning techniques are applied for NIDS implementation in different networking scenarios, such as traditional networks, cloud networks, Ad-Hoc, WSNs, and IoT networks. Hence, this article brings together publicly available intrusion datasets and machine learning techniques utilized in recent intrusion detection systems to reveal present-day challenges and future directions. This article also explains problems associated with NIDS. This will help researchers to enhance the existing NIDS models as well as to develop new effective models.

The internet of things (IoT) network is prone to several cyber-attacks, due to many obligations of IoT devices. Minority attack detection in the IoT network is a challenge. This paper proposed three multiclassifier models to address this challenge at different IoT layers. Random forest (RF) classifier is the main component in proposed models. RF hyperparameters are tuned with grid-search cross-validation (GSCV) to build an initial GS model that achieves a 100% normal traffic detection rate. It can efficiently separate normal and anomalous traffic at the IoT network layer. GS is extended with a feature importance (FI)-based feature reduction technique and the synthetic minority oversampling technique (SMOTE) successively to realise GSFI and GSFI_SMOTE models that achieve better minority attack detection rates and applicable to the resource-limited fog/edge infrastructure, and the critical IoT infrastructure, respectively. GSFI_SMOTE outperforms the existing methods. The UNSW-NB15 is used as a use case for experimenting with proposed models.

Distributed Denial-of-Service (DDoS) attacks are among the most destructive events in the cyber realm, causing significant disruptions to the operations of organizations, communities, and individuals. These severe incidents occur when an attacker exhausts all resources of the targeted system, preventing legitimate users from accessing it. The primary challenges in recognizing these attacks are the memory constraints and scalability issues of current attack detection systems. To detect such large-scale attacks, a thorough analysis of multidimensional volumetric network traffic is conducted. However, the presence of substandard features within this traffic diminishes the effectiveness of detection. This paper introduces a feature selection pipeline that combines a memory-efficient D-optimal subsampling method with the Jaya optimization algorithm, designed to extract important features from the CIC-DDoS2019 dataset. The attack detection model employs this feature selection to create an automated, scalable classifier that uses the H2O Automated Machine Learning (AutoML) platform to identify large-scale DDoS attacks. This classifier effectively differentiates between eight distinct types of attacks and one benign traffic type. The model reports remarkable metrics, achieving a precision of 89.9%, a recall of 90.0%, an F1 score of 89.9%, and an accuracy of 89.9% in the cross-validation data. Furthermore, it surpasses the abilities of contemporary models.