A promising approach for the early detection of fires in forests consists on placing ground-based sensors that communicate through a LoRaWAN network with certain gateways. LoRaWAN networks have gained significant interest recently, especially with the rise of solutions based on Internet of Things. In this work we address the problem of sensor placement in LoRaWAN networks using inte ger programming techniques, taking into account the particular characteristics of these networks. We consider a set of points of interest, a set of potential locations for sensors, and a set of gateways. The problem consists in determining (a) a subset of locations where sensors are to be installed, (b) a gateway and a spreading factor associated with each sensor, and (c) a channel associated with each gateway, in such a way that the number of sensors is subject to a global limit (due to budget constraints) and a per-gateway limit (due to hardware constraints), and minimizing a combination of (i) the number of uncovered points of interest and (ii) the network interference. We provide an integer programming formulation for this problem and we explore the capabilities of a state-of-the-art solver over this formulation applied to the “Pereyra Iraola” park in Buenos Aires, Argentina. We also present theoretical results on this formulation.