Exploring Logic Scoring of Preference for DoS Attack Detection in Microservice Applications

Microservice architectures allow the development of highly scalable, flexible, and manageable systems. However, such architectures raise new security problems and exacerbate the challenge of monitoring applications at runtime due to their high service granularity and distributed nature. Developing effective monitoring and security strategies is thus crucial to effectively detect potential attacks. This paper explores the applicability of Logic Scoring of Preference (LSP), a multi-criteria decision-making method to compute a score based on a set of preferences, for attack detection in microservice applications. We present an extensive experimental study and define a model based on LSP and application-level metrics to characterize the impact of DoS attacks. The output of the model is a unique score used to determine whether a microservice is under a DoS attack. The results of the experimental study show precision, recall, and f1-score rates of more than 80%, indicating that LSP could effectively characterize the application under attack, opening several possibilities for future work.