Monitoring particulate matter in cities with high population density or in areas with nearby industries (e.g., pollutants or toxins) is essential in these cities and is a necessary response to the increasing rates of diseases associated with air quality. In this context, and framed within the study of Wireless Sensor Networks (WSN), the idea of implementing a sensor network capable of constant monitoring of “particulate matter" arises, which, through the use of AI, can generate predictions of the behavior of this material, to generate a work plan with the corresponding authorities. This research aims to propose a distributed architecture design for a WSN, where the efficient use of sensor node resources (energy, messages, computation, among others) of the “particulate matter” of the environment of study. A distributed leader selection algorithm will send the data from WSN to Edge Computing (EC) to process the collected data and then send the processed information to cloud computing (CC).