Cluster Computing, Год журнала: 2024, Номер 28(2)
Опубликована: Ноя. 26, 2024
Язык: Английский
Cluster Computing, Год журнала: 2024, Номер 28(2)
Опубликована: Ноя. 26, 2024
Язык: Английский
Expert Systems with Applications, Год журнала: 2024, Номер 249, С. 123808 - 123808
Опубликована: Март 23, 2024
Язык: Английский
Процитировано
38PLoS ONE, Год журнала: 2025, Номер 20(1), С. e0317713 - e0317713
Опубликована: Янв. 16, 2025
The network intrusion detection system (NIDS) plays a critical role in maintaining security. However, traditional NIDS relies on large volume of samples for training, which exhibits insufficient adaptability rapidly changing environments and complex attack methods, especially when facing novel rare attacks. As strategies evolve, there is often lack sufficient to train models, making it difficult methods respond quickly effectively new threats. Although existing few-shot systems have begun address sample scarcity, these fail capture long-range dependencies within the environment due limited observational scope. To overcome challenges, this paper proposes elevated method based self-attention mechanisms iterative refinement. This approach leverages advantages extract key features from traffic dependencies. Additionally, introduction positional encoding ensures temporal sequence preserved during processing, enhancing model's ability dynamics. By combining multiple update meta-learning, model initially trained general foundation training phase, followed by fine-tuning with data testing significantly reducing dependency while improving prediction accuracy. Experimental results indicate that achieved rates 99.90% 98.23% CICIDS2017 CICIDS2018 datasets, respectively, using only 10 samples.
Язык: Английский
Процитировано
1PLoS ONE, Год журнала: 2025, Номер 20(2), С. e0312752 - e0312752
Опубликована: Фев. 6, 2025
Autonomous transportation systems have the potential to greatly impact way we travel. A vital aspect of these is their connectivity, facilitated by intelligent transport applications. However, safety ensured vehicular network can be easily compromised malicious traffic with exponential growth IoT devices. One identification in Vehicular networks. We proposed a hybrid approach uses automated feature engineering via correlation-based selection (CFS) and principal component analysis (PCA)-based dimensionality reduction reduce matrix size before series dense layers are used for classification. The intended use CFS PCA machine learning pipeline serves two folds benefit, first that resultant contains attributes most useful recognizing traffic, second after PCA, has smaller which turn means number weights need trained (connections required layers) resulting model size. Furthermore, show post-training weight quantization further Results demonstrate effectiveness improves classification f1score from 96.48% 98.43%. It also reduces 28.09 KB 20.34 thus optimizing terms both performance Post-training optimizes 9 KB. experimental results using CICIDS2017 dataset performs well not only but yields models low parameter count Thus, low-complexity intrusion detection VANET scenario.
Язык: Английский
Процитировано
1IEEE Access, Год журнала: 2024, Номер 12, С. 63584 - 63597
Опубликована: Янв. 1, 2024
The Internet of Things (IoT) represents a swiftly expanding sector that is pivotal in driving the innovation today's smart services. However, inherent resource-constrained nature IoT nodes poses significant challenges embedding advanced algorithms for cybersecurity, leading to an escalation cyberattacks against these nodes. Contemporary research Intrusion Detection Systems (IDS) predominantly focuses on enhancing IDS performance through sophisticated algorithms, often overlooking their practical applicability. This paper introduces Deep-IDS, innovative and practically deployable Deep Learning (DL)-based IDS. It employs Long-Short-Term-Memory (LSTM) network comprising 64 LSTM units trained CIC-IDS2017 dataset. Its streamlined architecture renders Deep-IDS ideal candidate edge-server deployment, acting as guardian between Denial Service (DoS), Distributed (DDoS), Brute Force (BRF), Man-in-the-Middle (MITM), Replay (RP) Attacks. A distinctive aspect this trade-off analysis intrusion detection rate false alarm rate, facilitating real-time Deep-IDS. system demonstrates exemplary 96.8% overall classification accuracy 97.67%. Furthermore, achieves precision, recall, F1-scores 97.67%, 98.17%, 97.91%, respectively. On average, requires 1.49 seconds identify mitigate attempts, effectively blocking malicious traffic sources. remarkable efficacy, swift response time, design, novel defense strategy not only secure but also interconnected sub-networks, thereby positioning IoT-enhanced computer networks.
Язык: Английский
Процитировано
8Computers, materials & continua/Computers, materials & continua (Print), Год журнала: 2024, Номер 79(3), С. 4319 - 4347
Опубликована: Янв. 1, 2024
Intrusion detection systems (IDS) are essential in the field of cybersecurity because they protect networks from a wide range online threats.The goal this research is to meet urgent need for small-footprint, highlyadaptable Network Detection Systems (NIDS) that can identify anomalies.The NSL-KDD dataset used study; it sizable collection comprising 43 variables with label's "attack" and "level."It proposes novel approach intrusion based on combination channel attention convolutional neural (CNN).Furthermore, makes easier conduct thorough assessment suggested strategy.Furthermore, maintaining operating efficiency while improving accuracy primary work.Moreover, typical NIDS examines both risky behavior using variety techniques.On dataset, our CNN-based achieves an astounding 99.728% rate when paired attention.Compared previous approaches such as ensemble learning, CNN, RBM (Boltzmann machine), ANN, hybrid auto-encoders MCNN, adaptive algorithms, solution significantly improves performance.Moreover, results highlight effectiveness method precision, signifying noteworthy advancement field.Subsequent efforts will focus strengthening expanding order counteract growing cyberthreats adjust changing network circumstances.
Язык: Английский
Процитировано
7Egyptian Informatics Journal, Год журнала: 2025, Номер 30, С. 100669 - 100669
Опубликована: Март 23, 2025
Язык: Английский
Процитировано
1Knowledge-Based Systems, Год журнала: 2025, Номер unknown, С. 113351 - 113351
Опубликована: Март 1, 2025
Язык: Английский
Процитировано
1Internet of Things, Год журнала: 2024, Номер 25, С. 101102 - 101102
Опубликована: Фев. 1, 2024
Язык: Английский
Процитировано
6Applied Soft Computing, Год журнала: 2023, Номер 151, С. 111125 - 111125
Опубликована: Дек. 5, 2023
Язык: Английский
Процитировано
13Cluster Computing, Год журнала: 2023, Номер 27(3), С. 3639 - 3655
Опубликована: Ноя. 2, 2023
Язык: Английский
Процитировано
12