SDN traffic anomaly detection method based on convolutional autoencoder and federated learning

Wang, ZiXuan; Wang, Pan; Sun, Zhixin

doi:10.1109/globecom48099.2022.10001438

Cited by 8 publications

(7 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different from the research [3,4,7,8] which employ the method of artificial feature engineering, we use raw traffic data as the input of SAT-NTAD. In data pre-processing phase, we clip the single packet data to a fixed length p l , and fill in zero if the packet length is insufficient.…”

Section: Traffic Data Pre-processingmentioning

confidence: 99%

“…Several studies have selected Auto-encoder (AE) for patterns of normal samples of models [2,8] . When the trained model encounters abnormal samples in the detection phase, the decoded output and input differ greatly to determine that abnormal flow occurs.…”

Section: Spatial Feature Extraction Of Packetmentioning

confidence: 99%

“… CAE: Based on the model parameters [8] , we utilize CAE to rebuild the model, which can identify anomalies by using reconstruction error.…”

Section: Comparison Approachesmentioning

confidence: 99%

“…Similarly, CNN and Long-Short Term Memory (LSTM) are used to extract flow feature and detect malicious traffic [7] . In addition, Convolutional Auto-encoder (CAE) is used to train normal flow [8] and test with normal and abnormal samples.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

An effective network traffic anomaly detection method based on deep learning for low-orbit satellite

Wang

2023

International Conference on Intelligent Systems, Communications, and Computer Networks (ISCCN 2023)

View full text Add to dashboard Cite

With the development of satellite networking technology, low-orbit satellites are gradually facing security threats caused by abnormal traffic. Since some types of attack traffic do not pass through ground stations, we are trying to migrate ground anomaly detection capabilities to satellites. In prior arts, many deep learning methods have been applied to network traffic anomaly detection. However, most of them have deep layers of neural networks and are unsuitable for satellites with limited energy and computing capacity. In this paper, we present SAT-NTAD, an effective low-orbit satellite network traffic anomaly detection method that combines shallow Convolutional Auto-encoder (CAE) and Bidirectional Simple Recurrent Unit (Bi-SRU) to extract the spatial and contextual features of traffic flows, and detects anomalies by utilizing dynamic threshold. Extensive experimental evaluations show that our method can stably maintain the high performance for abnormality detection, which outperforms state-of-the-art approaches, and effectively reduces the computational cost, which is more suitable for the low computing capacity environment of satellites.

show abstract

Section: Traffic Data Pre-processingmentioning

confidence: 99%

Section: Spatial Feature Extraction Of Packetmentioning

confidence: 99%

“… CAE: Based on the model parameters [8] , we utilize CAE to rebuild the model, which can identify anomalies by using reconstruction error.…”

Section: Comparison Approachesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An effective network traffic anomaly detection method based on deep learning for low-orbit satellite

Wang

2023

International Conference on Intelligent Systems, Communications, and Computer Networks (ISCCN 2023)

View full text Add to dashboard Cite

show abstract

“…Many research endeavors have also sought to apply deep learning algorithms to intrusion detection. However, compared to machine learning methods, deep learning demands more significant resource consumption [13]. Deep Forest (DF), introduced by Zhou Zhihua et al [14], is an innovative deep learning model.…”

mentioning

confidence: 99%

Resource-Constraint Deep Forest Based Intrusion Detection Method in Internet of Things for Consumer Electronic

Chen,

Wang,

Yang

et al. 2024

IEEE Trans. Consumer Electron.

View full text Add to dashboard Cite

Consumer Electronics (CEs) are smart devices using IoT for connectivity. They're susceptible to attacks like DoS, DDoS, and web attacks, impairing functions and enabling remote hijacking. Attackers can exploit CEs to target other systems, like vehicles. Malicious code can propagate through networks or CEs, causing vehicle failures. Existing ML/DL based IDS have high classification accuracy and robustness in traditional internet environments, but they are overly complex for performance improvement, which hinders their deployment in edge small computing environments. Furthermore, the comparison experiments of these intrusion detection algorithms with other algorithms are not sufficiently comprehensive to evaluate their performance in small computing environments. Therefore, balancing "detection performance and resource consumption" is a key issue in CE network detection. To address this issue, this paper proposes a hybrid feature selection model based on chi-square test and information gain combined Ig-Chi, which effectively reduces the feature dimensionality and improves the classification accuracy of classifiers for high-dimensional data sets. Additionally, layered intrusion detection is employed to perform intrusion detection on the data after feature selection. The experiments on four public data sets demonstrate that this method surpasses six ML/DL algorithms in terms of accuracy and resource indicators.

show abstract

Collaborative DDoS defense for SDN-based AIoT with autoencoder-enhanced federated learning

Ma,

2025

Information Fusion

View full text Add to dashboard Cite

SDN traffic anomaly detection method based on convolutional autoencoder and federated learning

Cited by 8 publications

References 13 publications

An effective network traffic anomaly detection method based on deep learning for low-orbit satellite

An effective network traffic anomaly detection method based on deep learning for low-orbit satellite

Resource-Constraint Deep Forest Based Intrusion Detection Method in Internet of Things for Consumer Electronic

Collaborative DDoS defense for SDN-based AIoT with autoencoder-enhanced federated learning

Contact Info

Product

Resources

About