During the early phases of software system development, error detection can be challenging due to the complexity of both the requirements and the operating environments. This paper advocates for the utilization of formal modelling and verification throughout the first phases of systems development to promptly detect and correct errors. The formalism employed throughout is Event‐B, which is backed by the Rodin toolset. To conquer requirements complexity, the frameworks of set theory and first‐order logic are employed, which provide the necessary tools for formalizing and analysing the properties and behaviours associated with Event‐B. Also, we detail the way in which modelling may be used to achieve abstraction, as well as the way in which refinement can be used to manage complexity through layering. Furthermore, we emphasize the significance of model validation and verification in improving the precision of formal models and requirements in IoT communication systems. The model is exemplified using a Content‐Based Publish Subscribe System (CBPS), with a special emphasis on a fire alarm system as a motivating example.