Table 1.

Review of existing work in the field of intrusion detection.

Reference	Methods/Techniques	Key Features	Challenges/Improvement
[10]	A hybrid method based on GA and ANN	Better precision and recall.	No real-time data set. Accuracy can be improved by adding two-way training. Analysis of variance (ANOVA) is missing.
[11]	Ensemble model based on meta-classification	Better precision and accuracy compared to other methods.	The training and testing process time is lengthy. The new IDS challenges were not covered.
[12]	Risk analysis of RPL and OFS	Capable of dealing with high-dimensional data.	Training requires a significant amount of time.
[13]	Deep learning in IDS	DNNs perform outstandingly in terms of better precision and recall.	Only limited datasets were used.
[14]	Deep-learning approach in NIDS	Reduce false alarms and training times.	ANOVA is not implemented. Only a few datasets were used.