The traditional process of finding a vacant parking slot is often inefficient: it increases driving time, traffic congestion, fuel consumption and exhaust emissions. To address such problems, smart parking systems have been proposed to help drivers to find available parking slots faster using latest sensing and communications technologies. However, the deployment of the communications infrastructure of a smart parking is not straightforward due to multiple factors that may affect wireless propagation. Moreover, a smart parking system needs to provide not only accurate information on available spots, but also fast responses while guaranteeing the system availability even in the case of lacking connectivity. This article describes the development of a decentralized low-latency smart parking system: from its conception, design and theoretical simulation, to its empirical validation. Thus, this work first characterizes a real-world scenario and proposes a fog computing and Internet of Things (IoT) based communications architecture to provide smart parking services. Next, a thorough analysis on the wireless channel properties is carried out by means of an in-house developed deterministic 3D-Ray Launching (3D-RL) tool. The obtained results are validated through a real-world measurement campaign and then the communications architecture is implemented by using ZigBee sensor nodes. The implemented architecture also makes use of Bluetooth Low Energy beacons, an Android app, a decentralized database and fog computing gateways, whose performance is evaluated in terms of response latency and processing rate. Results show that the proposed system is able to deliver information to the drivers fast, with no need for relying on remote servers. As a consequence, the presented development methodology and communications evaluation tool can be useful for future smart parking developers, which can determine the optimal locations of the wireless transceivers during the simulation stage and then deploy a system that can provide fast responses and decentralized services.