Performance Evaluation of Monolithic and Microservice Architectures for Natural Language Processing in Command and Control Applications

The success of Command and Control (C2) operations relies heavily on clear communication among participants, often under stressful and time-sensitive conditions that demand accurate decisionmaking. Examples of such operations include military engagements, reconnaissance missions, search and rescue efforts, and peacekeeping activities. The design of systems to support C2 is therefore challenging, as they must support structured communication and interaction means among participants. To address this, we developed a method called Method to Support Semantic Interoperability in Command and Control (MAISC<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) to support C2, in which we have applied Natural Language Processing (NLP) and semantic techniques to pinpoint specific elements in sentences, thereby enhancing the participants’ understanding of the C2 communications. Microservices Architecture (MSA) are known to offer potential benefits over MA, particularly in terms of scalability, independence, and maintenance. Furthermore, some NLP applications have already been developed using MSA, confirming these benefits. This paper presents a version of our system developed using MSA and compares it to its Monolithic Architecture (MA) counterpart. Our objective is to evaluate whether the benefits of adopting MSA, observed in other domains, also apply to C2 support systems based on NLP and semantic techniques, as exemplified by the MAISC<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> method. Our results show that MA outperforms MSA in many scenarios; however, as the application load increases, MSA shows increasingly better performance. In particular, although MA had shorter processing and delivery times under normal conditions, MSA delivered better processing performance when handling significantly higher levels of concurrent activity.