MQTT is one of the major messaging protocols in the Internet of things (IoT). In this work, we investigate the expected performance of MQTT implementations in various settings. We present a model-based performance testing approach that allows a fast simulation of specific usage scenarios in order to perform a quantitative analysis of the latency. Out of automatically generated log-data, we learn the distributions of latencies and apply statistical model checking to analyse the functional and timing behaviour. The result is a novel testing and verification technique for analysing the performance of IoT protocols. Two well-known open source MQTT implementations are evaluated and compared.
CITATION STYLE
Aichernig, B. K., & Schumi, R. (2018). How fast is MQTT?: Statistical model checking and testing of IoT protocols. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11024 LNCS, pp. 36–52). Springer Verlag. https://doi.org/10.1007/978-3-319-99154-2_3
Mendeley helps you to discover research relevant for your work.