Dr. Ezhil Selvam Kalaimannan

The high scale of cyber-attacks has revealed the shortcomings of conventional intrusion detection systems (IDS), which has prompted the development of new and sophisticated methods that can model complicated traffic patterns. In this work, the emphasis is on enhancing proactive anomaly detection using graph convolutional networks (GCNs) in conjunction with classical machine learning techniques. Two models are suggested: a GCN-autoencoder (GCN-AE) to identify any anomalies with no supervision and a GCN-k-nearest neighbors (GCN-KNN) to classify intrusions with supervision. The UNSW-NB15 dataset consists of 700,000 instances and 49 features, which were pre-processed by scaling the features, encoding labels, and converting categorical variables into one-hot encodings. To reduce class bias, a balanced data set of 22,000 anomalies and 22,000 non-anomalies was subsequently generated. Graph structures were used to encode node-to-node relationships, and both models were run in Keras-TensorFlow with hyperparameter optimization. Experimental results demonstrate that GCN-AE performed moderately (Accuracy: 52.86% and Precision: 81.48% and Recall: 7.96% and F1-score: 14.51%), and GCN-KNN performed exceptionally well (Accuracy: 99.94% and Precision: 99.90% and Recall: 99.97 and F1-score: 99.93) with an AUC of 0.99 and extremely low false positives. In all, this study demonstrates that graph-based learning, particularly GCN-KNN, offers a robust and classical framework for detecting intrusions. The suggested solution is capable of detecting various threats, including DoS, port scanning, and data exfiltration, which makes it promising to implement in a multilayer network infrastructure.

The rapid increase in network traffic has created an urgent need for more effective security systems. Intrusion Detection Systems (IDS) are essential for network security but struggle with the growing volume and complexity of data, particularly in efficiently processing high-dimensional network traffic data. We introduce AEJaya+DE, a novel optimization approach that combines the Adaptive Enhanced Jaya Algorithm (AEJaya) with Differential Evolution (DE) to address these challenges. AEJaya+DE addresses the limitations of the basic Jaya algorithm, which often gets trapped in local optima due to its single learning strategy. Our method enhances the algorithm's performance through two key innovations: it improves local search capabilities using local attractors for exploitation, and it strengthens global exploration through historical population data. The algorithm incorporates dynamic parameter adjustment and adaptive probabilistic strategy selection to maintain an optimal balance between the exploration and exploitation phases. We validated AEJaya+DE's effectiveness using two standard benchmark datasets. On the UNSW-NB15 dataset, it reduced the feature set from 42 to 21 features while achieving 96.10% accuracy using the XGBoost classifier. The NSL-KDD dataset decreased features from 41 to 24 while reaching 99.93% accuracy with the CatBoost classifier. These results demonstrate AEJaya+DE's ability to select a compact yet highly informative feature set, significantly improving intrusion detection systems' accuracy and computational efficiency. The method represents a significant advancement in network security, offering a robust and efficient solution for intrusion detection across various network environments.

Quantum memory is essential for the prolonged storage and retrieval of quantum information. Nevertheless, no current studies have focused on the creation of effective quantum memory for continuous variables while accounting for the decoherence rate. This work presents an effective continuous-variable quantum key distribution method with parameter optimization utilizing the Elitist Elk Herd Random Immigrants Optimizer (2E-HRIO) technique. At the outset of transmission, the quantum device undergoes initialization and authentication via Compressed Hash-based Message Authentication Code with Encoded Post-Quantum Hash (CHMAC-EPQH). The settings are subsequently optimized from the authenticated device via 2E-HRIO, which mitigates the effects of decoherence by adaptively tuning system parameters. Subsequently, quantum bits are produced from the verified device, and pilot insertion is executed within the quantum bits. The pilot-inserted signal is thereafter subjected to pulse shaping using a Gaussian filter. The pulse-shaped signal undergoes modulation. Authenticated post-modulation, the prediction of link failure is conducted through an authenticated channel using Radial Density-Based Spatial Clustering of Applications with Noise. Subsequently, transmission occurs via a non-failure connection. The receiver performs channel equalization on the received signal with Recursive Regularized Least Mean Squares. Subsequently, a dataset for side-channel attack authentication is gathered and preprocessed, followed by feature extraction and classification using Adaptive Depthwise Separable Convolutional Neural Networks (ADS-CNNs), which enhances security against side-channel attacks. The quantum state is evaluated based on the signal received, and raw data are collected. Thereafter, a connection is established between the transmitter and receiver. Both the transmitter and receiver perform the scanning process. Thereafter, the calculation and correction of the error rate are performed based on the sifting results. Ultimately, privacy amplification and key authentication are performed using the repaired key via B-CHMAC-EPQH. The proposed system demonstrated improved resistance to decoherence and side-channel attacks, while achieving a reconciliation efficiency above 90% and increased key generation rate.

To make smart cities more sustainable, deep neural networks (DNNs) assess data from sensors, the Internet of Things (IoT), and social media in an effective manner. Swiftly assessing text, images, and time-series data, they aid administrators in understanding complex urban trends and allocating resources. DNNs employ climate predictions and energy usage patterns to enhance energy distribution, minimize trash, and boost green energy resources. This improves grid reliability and reduces greenhouse gas emissions. DNNs trained on real-time traffic data enhance congestion, transportation efficacy, and traffic flow. Intelligent scheduling techniques enhance transportation, shipment, and fuel economy. DNNs enhance smart city garbage collection routes, recycling opportunities, and on-demand services by assessing past and present data. This strategy encourages reducing trash, recycling, and maintaining a clean atmosphere. Nevertheless, there is an urgent need to address data privacy, computational injustice, and the accessibility of computer resources. Responsible and collaborative decision-making is crucial to address these issues and avoid unfairness. By prioritizing ethics and fair distribution of benefits, DNNs can improve transportation, trash management, and energy efficiency, paving the way for more accessible and inclusive digital cities.

In the context of manufacturing, an Industrialized Control and Automation System (ICAS) is a set of interrelated economic processes that allow for the automated, data‐driven tuning of various operational parameters of machinery and other mechanical systems with minimal human intervention. Safety, productivity, and equipment are all in danger from the widespread use of ICAS in areas like manufacturing, logistics, supply chain management, and transportation. Industry Internet Systems (IIS) and the IoT have played a critical role in the evolution of smart factories, which facilitate the autonomous management of resources through the sharing of data between previously isolated devices. Industry 4.0, IIS, Industrial Internet of Things, and ICAS all rely on IoT as their base, anticipating major advancements in design. The IIoT refers to a system where machines may interact with one another and make small adjustments to their environment or behavior with little to no human interaction. Embedded sensor nodes collect data, gateways analyze and send it to the cloud, and higher‐level communication is made possible via networks like Wi‐Fi and Ethernet, making up the three layers of IIoT operations. Focusing on its data analysis and problem‐solving utility within the context of Industry 4.0, this article provides a detailed review of the theoretical and technological framework underlying the corporate implementation of IIoT.

Distributed Denial of Service (DDoS) attack has been used as a weapon by many hackers to cause a network failure thereby preventing the users from using the network service. This attack is perpetrated to create financial loss, to gain money, to seek revenge, etc. Software Defined Network (SDN) has grown widely to gain the hope of many researchers and network administrators to meet the need for maintenance of the data centers. Today’s world mostly prefers SDN due to its flexibility, cost efficiency, manageability, etc. SDN has a network architecture that changes its traffic pattern every time a change is made in its central architecture. The separation of the planes in SDN ensures the security of entire network devices and their functions thereby protecting the network. So, the attackers target the controller to bring it down leading to a network failure. Thus, the single controller has many issues with factors of both performance and scalability. Multiple controllers are used to resolve this issue. This chapter presents a full assessment of the necessity for multiple controllers in SDN, as well as the benefits and drawbacks.

Network intrusion detection systems are vital for network security, tasked with identifying and thwarting malicious activities. This research compares Random Forest (RF) and Support Vector Machines (SVM) in detecting intrusions using the UNSW-NB15 dataset. RF achieves 99.64% accuracy, with 4454 True Positives (TP), 4400 True Negatives (TN), 31 False Positives (FP), and 1 False Negative (FN). In contrast, SVM attains 78.87% accuracy, with 4422 TP, 2586 TN, 63 FP, and 1815 FN. The study also considers percentage changes in TP (0.72%), FP (-50.79%), FN (-99.94%), and TN (70.15%), providing insights into model adaptability. Validating RF with the Synthetic Minority Over-sampling Technique (SMOTE) yields 99.63% accuracy, compared to SVM’s 89%, indicating RF’s robustness with imbalanced datasets. RF’s superiority partly stems from its effective use of feature importance analysis, and its robustness to noise, maximizing the utility of selected features for better predictive performance compared to SVM which does not perform well with feature selection. This provides an understanding of RF and SVM strengths and limitations in network intrusion detection and underlines the importance of selecting appropriate machine learning models for network security applications concerning a particular kind of datase

In our interconnected digital landscape, safeguarding network security is paramount. This research juxtaposes two anomaly detection methods: an Auto-encoder model using Ten-sorFlow's Keras and the K-Nearest Neighbours (KNN) algorithm. Beyond assessing model performance, this study underscores the practical relevance of these techniques in real-world security contexts. The KNN results reveal 202,325 True Positives, 4,442 True Negatives, 960 (0.045%) False Positives (Type-I error), and 2,274 (1.08%) False Negatives (Type-II error), while the Auto-encoder model achieves 130,260 True Positives, 1,791 True Negatives, 5,208 (3.7%) False Positive (Type-I error), and 2,742 (1.96%) False Negatives (Type-II error). Crucially, this research emphasizes that timely anomaly detection is the linchpin in thwarting potential security breaches, with anomaly prevention serving as a proactive defense strategy. By harnessing machine learning and data-driven methodologies, this work contributes to fortifying network security. These findings provide security prac-titioners with valuable insights into the pivotal role of anomaly detection in intrusion prevention. Furthermore, this study paves the way for future advancements in network security, solidifying the position of proactive anomaly detection in cybersecurity.

We propose a light-weighted authentication scheme FogAutho for securing fog-cloud systems in smart home applications. It uses a novel approach to generate a session key after authenticating both ends; Fog and IoT layers. The cloud server may revoke the session key. Our proposed approach minimizes the communication between end users, fog server, and cloud server during data authentication. FogAutho achieves security objectives such as mutual authentication for both ends, a secure session key generation, data confidentiality, and attack resistance with significantly reduced communication overhead compared to the existing authentication mechanisms in the fog-cloud system.

Modern enterprise networks are complex and present countless security challenges. Understanding the nature of the systems that exist within a network environment is a vital step in securing such environments. Therefore, operating systems on the network must be identified, tracked, and continuously monitored. In this research, we consider the problem of detecting unauthorized operating systems on an enterprise network, which could exist because of the unintentional actions of an authorized user or the unauthorized actions of internal users or external attackers. We intend to utilize an artificial neural network-based classifier [ANN], which will be developed using the PyTorch and fastai deep learning libraries. Simulated network traffic has been generated through the implementation of two separate virtual network environments, and the generated traffic was passively collected and analyzed prior to traversing the network boundary. The performance evaluation of the neural network classifier will be analyzed using the collected data in this research.