22 March 2024 in IEEE Access
The Industrial Internet of Things (IIoT) comprises a variety of systems, smart devices, and an extensive range of communication protocols. Hence, these systems face susceptibility to privacy and security challenges, making them prime targets for malicious attacks that can result in harm to the overall system. Privacy breach issues are a notable concern within the realm of IIoT. Various intrusion detection systems based on machine learning (ML) and deep learning (DL) have been introduced to detect malicious activities within these networks and identify attacks. The existing ML and DL-based models face challenges when confronted with highly imbalanced training. Repetitive data in network datasets inflates model performance, as the model has encountered much of the test set data during training. Moreover, these models decrease performance when confronted with datasets that include repetitions of similar data across various classes, where only the class labels are different. To overcome the challenges inherent in existing systems, this paper presents a self-attention-based deep convolutional neural network (SA-DCNN) model designed for monitoring the IIoT networks and detecting malicious activities. Additionally, a two-step cleaning method has been implemented to eliminate redundancy within the training data, considering both intra-class and cross-class samples. The performance of the SA-DCNN model is assessed using IoTID20 and Edge-IIoTset datasets. Furthermore, the proposed study is demonstrated through a comprehensive comparison with other ML and DL models, as well as against relevant studies, showcasing the superior performance and efficacy of the proposed model.
29 August 2024 in IEEE Access
Protection against unwanted intrusions is crucial for preserving the integrity and security of connected devices in the context of Internet of Things (IoT) networks. The growing number of IoT devices has made several industries more vulnerable to cyberattacks and security breaches, including smart homes, industrial automation, and healthcare. In response to this pressing dilemma, the goal of this project is to create a novel method for intrusion detection in Internet of Things systems utilizing Denoising Autoencoder (DAE) models. Traditional intrusion detection methods often prove inadequate in Internet of Things scenarios due to resource restrictions, dynamic network topologies, and a diversity of communication protocols. By utilizing DAEs’ unsupervised learning and feature extraction skills, our suggested approach creates a system that can identify and stop intrusion attempts in real-time. The evaluation of the study additionally makes use of the NSL-KDD and CICIDS 2017 datasets. DAE integration yields an unequaled accuracy of 99.991% when the CICIDS 2017 dataset is used, and an accuracy of 99.4% when the NSL-KDD dataset is used. The CICIDS 2017 dataset analysis reveals several notable performance measures, including an accuracy of 1.0, a precision of 0.995, and an F1-score of 0.998. Analyses of the NSL-KDD dataset also produce outstanding results, with an F1-score of 0.989, recall of 0.991, accuracy of 0.994, and precision of 0.984. The results also show how well the suggested DAE-based intrusion detection method works to stop unauthorized users from accessing IoT devices, which lowers the risk of issues with system integrity, privacy, and security. By strengthening resilience against evolving cyber threats in the networked Internet of Things landscape, this research enhances cybersecurity strategies tailored to address the unique challenges encountered by IoT ecosystems.
27 January 2025 in Scientific Reports
In the present digital scenario, the explosion of Internet of Things (IoT) devices makes massive volumes of high-dimensional data, presenting significant data and privacy security challenges. As IoT networks enlarge, certifying sensitive data privacy while still employing data analytics authority is vital. In the period of big data, statistical learning has seen fast progressions in methodological practical and innovation applications. Privacy-preserving machine learning (ML) training in the development of aggregation permits a demander to firmly train ML techniques with the delicate data of IoT collected from IoT devices. The current solution is primarily server-assisted and fails to address collusion attacks among servers or data owners. Additionally, it needs to adequately account for the complex dynamics of the IoT environment. In a large-sized big data environment, privacy protection challenges are additionally enlarged. The data dimensional can have vague meaningful patterns, making it challenging to certify that privacy-preserving models do not destroy the efficacy and accuracy of statistical methods. This manuscript presents a Privacy-Preserving Statistical Learning with an Optimization Algorithm for a High-Dimensional Big Data Environment (PPSLOA-HDBDE) approach. The primary purpose of the PPSLOA-HDBDE approach is to utilize advanced optimization and ensemble techniques to ensure data confidentiality while maintaining analytical efficacy. In the primary stage, the linear scaling normalization (LSN) method scales the input data. Besides, the sand cat swarm optimizer (SCSO)-based feature selection (FS) process is employed to decrease the high dimensionality problem. Moreover, the recognition of intrusion detection takes place by using an ensemble of temporal convolutional network (TCN), multi-layer auto-encoder (MAE), and extreme gradient boosting (XGBoost) models. Lastly, the hyperparameter tuning of the three models is accomplished by utilizing an improved marine predator algorithm (IMPA) method. An extensive range of experimentations is performed to improve the PPSLOA-HDBDE technique’s performance, and the outcomes are examined under distinct measures. The performance validation of the PPSLOA-HDBDE technique illustrated a superior accuracy value of 99.49% over existing models.
20 November 2023 in IEEE Access
The Internet of Things (IoT) based Wireless Sensor Networks (WSNs) contain interconnected autonomous sensor nodes (SN), which wirelessly communicate with each other and the wider internet structure. Intrusion detection to secure IoT-based WSNs is critical for identifying and responding to great security attacks and threats that can cooperate with the integrity, availability, and privacy of the network and its data. Machine learning (ML) algorithms are deployed for detecting difficult patterns and subtle anomalies in IoT data. Artificial intelligence (AI) driven methods are learned and adapted from novel data for improving detection accuracy over time. In this article, we introduce a Red Kite Optimization Algorithm with an Average Ensemble Model for Intrusion Detection (RKOA-AEID) technique for Secure IoT-based WSN. The purpose of the RKOA-AEID methodology is to accomplish security solutions for IoT-assisted WSNs. To accomplish this, the RKOA-AEID technique performs pre-processing to scale the input data using min-max normalization. In addition, the RKOA-AEID technique performs an RKOA-based feature selection approach to elect an optimum set of features. For intrusion detection, an average ensemble learning model is used. Finally, the Lévy-fight chaotic whale optimization Algorithm (LCWOA) can be executed for the optimum hyperparameter chosen for the ensemble models. The performance evaluation of the RKOA-AEID algorithm can be tested on the benchmark WSN-DS dataset. The extensive experimental outcomes stated the higher outcome of the RKOA-AEID algorithm with other approaches with an improved accuracy of 98.94%.
22 September 2023 in IEEE Access
5G networks offer high-speed, low-latency communication for various applications. As 5G networks introduce new capabilities and support a wide range of services, they also become more vulnerable to different kinds of cyberattacks, particularly Distributed Denial of Service (DDoS) attacks. Effective DDoS attack classification in 5G networks is a critical aspect of ensuring the security, availability, and performance of these advanced communication infrastructures. In recent days, machine learning (ML) and deep learning (DL) models can be employed for an accurate DDoS attack detection process. In this aspect, this study designs a Modified Equilibrium Optimization Algorithm with Deep Learning based DDoS Attack Classification (MEOADL-ADC) method in 5G networks. The goal of the MEOADL-ADC technique is the automated classification of DDoS attacks in the 5G network. The MEOADL-ADC technique follows a three-stage process such as feature selection, classification, and hyperparameter tuning. Primarily, the MEOADL-ADC technique employs MEOA based feature selection approach. Next, the MEOADL-ADC technique utilizes the long short-term memory (LSTM) model for the classification of DDoS attacks. Finally, the tunicate swarm algorithm (TSA) is exploited to adjust the hyperparameter of the LSTM model. The design of MEOA-based feature selection and TSA-based hyperparameter tuning shows the novelty of the work. The experimental outcome of the MEOADL-ADC method is tested on a benchmark dataset, and the outcomes indicate the betterment of the MEOADL-ADC algorithm over the current methods with maximum accuracy of 97.60%.
13 March 2025 in IEEE Access
The rising adoption of Consumer Internet of Things (CIoT) devices brings heightened concerns about security vulnerabilities and privacy breaches. Given their connection to daily routine and data collection abilities, it is vital to protect individual data against unauthorized entry and possible leaks proactively. Data protection’s maximum communication and storage charge will gradually challenge the typical ecology of integrated over-the-cloud learning and handling for IoT environments. Federated learning (FL) is developed as the most reasonable method for this challenge. In FL, training data-driven machine learning (ML) approaches is a form of association among several users with no need for data carried to a fundamental point; therefore, improving communication and storage charge offers a great degree of consumer-level secrecy. FL is an advanced ML that performs appropriately by inherently prioritizing confidentiality, circumventing requirements for centralized data collection, and supporting the defence. FL improves security in IoT by enabling decentralized model training without centralizing sensitive data, preserving privacy, and reducing communication costs. Integrating FL into attack detection models ensures zero-trust security, minimizing vulnerabilities and enhancing protection against cyber threats while keeping data on-device. This study develops an enhanced artificial gorilla troop optimizer with an FL-based attack detection and classification (EGTO-FLADC) approach on the CIoT environment. The EGTO-FLADC approach aims to improve security in the CIoT environment via FL and the attack detection process. Primarily, the EGTO-FLADC technique utilizes the FL model for collaborative training of combined models with many devices with no data-sharing process. In addition, the EGTO-FLADC technique uses a min-max scalar for data normalization. For attack detection, the EGTO-FLADC technique employs the temporal convolution networks with a gated recurrent unit (TCN-GRU) model and its hyperparameters are chosen by the EGTO model. The EGTO technique involves the integration of the conventional GTO model with the marine predator’s algorithm (MPA) model. The performance evaluation of the EGTO-FLADC approach is examined under the EdgeIIoTset dataset. The experimental validation of the EGTO-FLADC approach portrayed a superior accuracy value of 97.11% over existing models.