LoRaWAN is the one of the most popular connectivity choices for implementing IoT projects which incorporate verticals such as smart lighting, smart parking, waste management, smart metering and many more. Among its numerous advantages are extended coverage (a few kilometres in rural areas), high reliability, low battery consumption, low cost of implementation and ease of expandability. However, one or more of these advantages can be lost if the network isn’t planned and designed carefully, through a structured methodology which combines theoretical models and on-site verification.
Guaranteeing LoRaWAN coverage during the planning phase is essential. IoT sensors might communicate only a few times a day or week, so every transmission matters and the network should not allow packets to be lost. A well designed network is free from network congestion due to repeated join efforts or retransmissions. In addition, when the nodes operate under stress-free transmission parameters their battery lasts longer. Finally, it’s also important that the planning methodology produces an optimal number of required LoRaWAN gateways, keeping the cost of investment low.
Intracom Telecom's multi-year expertise both in IoT projects and in RF network design has led to a definition of a well-structured methodology for planning LoRaWAN deployments. The methodology is a hybrid approach combining theoretical modeling of signal propagation through 3D simulations of terrain and building landscapes in addition to actual on-site measurement verifications.
Intracom Telecom's methodology begins with a broad set of candidate points of installation of LoRaWAN gateways. The points are evaluated with respect to their installation feasibility in addition to the results of simulation, based on selected models of electromagnetic waves propagation. Important evaluation factors of the candidate installation points comprise the existence of mounting masts and uninterruptible power supply for the gateways, the surrounding area terrain and the existence of obstacles that may impede the propagation of the electromagnetic waves as well as the coverage and overlap revealed from the theoretical simulation.
The simulation is based on multiple propagation models and parameters, so that the coverage performance of each point is assessed under simulated conditions which are worse than the ideal in order to boost the reliability of the network under design. The theoretical coverage simulations take into account not only the terrain but also objects, such as trees and buildings in the area. Such an approach leads to an optimal set of installation points that combine strong reception power within the covered area in addition to adequate redundancy for increased reliability.
The above theoretical results are then assessed with on-site measurements. This is an essential step since theoretical models do not always fully depict reality. More specifically, sensors that in theory may seem to have weak or no reception signal may in reality have good coverage and vice versa. One reason that this may happen are the peculiarities of the verticals under study. For example, a parking sensor will be covered by a car which in reality will reduce the strength of its received signal, something that the theoretical model does not take into account. Another reason that calls for on-site verification is the fact that the theoretical models may not include recent changes in urban landscape, such as the erection of new buildings which will impede the signal propagation. Intracom Telecom can carry out detailed on-site coverage verification by using portable equipment. The results of the theoretical models versus the on-site measurement are compared and analyzed using Intracom Telecom’s in-house developed software tool.
By overlapping the results of the theoretical analysis and on-site measurements, the network designer can identify where the two differ. Subsequently the designer is able to tune the theoretical model with parameters that can be derived from the real measurements and re-run the simulation to get more accurate results. In addition, decisions related to sensor locations with poor coverage can be taken based on the overlapped information. The designer can choose among relocating a LoRaWAN gateway, investing in a new LoRaWAN gateway or even considering an alternative connectivity technology (e.g. NB-IoT) for a small number of locations. In any case, the designer has both the information and the tools to justify the finalized network topology.
With the above methodology, LoRaWAN network designers experience an unprecedented number of benefits. First, optimal sensor placement is proven and results in extended battery life. Also, gateway positioning adjustment can be easily simulated and justified mostly from the comfort of their office. And even in cases of "no coverage", the results of the above methodology will lead them to the optimal approach. Finally, the whole process can lead to designing a network that accommodates expansion easily.
It is worth mentioning that the above methodology can be also applied in existing deployments. Combining the theoretical analysis and on-site verification, poor coverage areas can be identified and modifications will be proposed.
Also, the battery lifetime and replacement costs can be predicted and thus, IoT sensors for a new vertical will be efficiently introduced in the existing network. Finally, having a comprehensive image of the network coverage, the costs of NB-IoT "patches" in areas of lack of adequate coverage are calculated.