Advances, Systems and Applications
Table 1 Summary of the Main Characteristics of the Proposed Methodology versus other State-of-the-Art Methodologies
From: Enhancement of an IoT hybrid intrusion detection system based on fog-to-cloud computing
The methodology | Algorithms used | Dataset used | Feature selection methodology | Fog devices used | Detection technique |
|---|---|---|---|---|---|
Proposed methodology (EHIDS) | GA and BPNN | UNSW-NB15 and ToN_IoT | Filter methods (Standard Deviation) | Raspberry Pi4 | Anomaly |
CF-OSELM-PRFF [19] | Cholesky Factorization based Online Sequential Extreme Learning Machines with Persistent Regularization | NSL-KDD | N/A | Azure Cloud Service | Anomaly |
ABA-IDS [22] | ANN | Legitimate Commands | Pearson Product-Moment Correlation Coefficient | Raspberry Pi3 model B | Anomaly |
ICNN-FCID [12] | CNN and LSTM Networks | NSL-KDD | N/A | N/A | Anomaly |
Ensemble-Based IDS [1] | logistic regression, naive Bayes, and decision trees | CICIDS 2017 | LinearSVM feature selection | N/A | Anomaly |
cyber–physical systems network intrusion detection model [25] | Random Forest, SVM and Logistic Regression | SDN-IoT, KDD-Cup-1999, UNSW-NB15, WSN-DS, and CICIDS-2017 | kernel-based principal component analysis technique | N/A | Anomaly |
BiDLSTM-based intrusion detection system [10] | bidirectional LSTM | NSL-KDD | N/A | N/A | Anomaly |
XAI-based IDS [8] | RuleFit, Local Interpretable Model-Agnostic Explanations, and SHapley Additive exPlanations | NSL-KDD and UNSW-NB15 | N/A | N/A | -- |