| Balancing Quality of Service (QoS) and security in wireless networks will become
increasingly important as more devices become connected and networks, especially 5G and
Beyond, grow increasingly more complex. We address a research gap such that security
and QoS are often considered separately without regard to their relation. By conducting a
thorough statistical analysis, we can provide a confidence interval that quantifies security’s
effect on QoS. Our statistical analysis aims to pave the way for further development of
intelligent and QoS-aware security. Additionally, we present a lightweight hierarchical
model for anomaly detection that can significantly reduce computational burden while still
maintaining high anomaly detection success, comparable to other state-of-the-art methods.
The proposed model uses two machine learning models. The first is a simple and lightweight
model to provide an initial inference, while the second is a more complex model. A confidence
score threshold determines which data will be sent to the more complex model to provide a
more accurate inference. The optimal threshold value is identified by employing a modified
Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS)-based method.
Keywords: Artificial Intelligence, Cybersecurity, Quality of Service, Wireless Networks |