LoRaWAN, NB-IoT and 4G on Egyptian Farmland: An Honest Comparison

The question "which network?" is usually settled by a slogan. It is in fact an engineering decision with three constraints: how much data has to move, how much power is available at the device, and what cellular coverage actually exists on that site. There is no single right answer for every farm, and a supplier who offers one has not visited the site.

LoRaWAN is a low-power wide-area network operating in Egypt on the 868 MHz band, with a practical range of several kilometres per gateway across flat open land. Its decisive agricultural advantage is that there is no monthly subscription per device, and a field unit with a small solar panel runs for years. Its drawback is real and should be stated: payload capacity is very small - tens of bytes per message, with transmission governed by duty-cycle limits - so it carries no images, no video and no per-second sampling. It also needs a gateway you own, which is infrastructure you have to maintain.

NB-IoT removes the gateway problem: the network is the operator's and the device attaches directly. Its penetration into buildings and well shafts is excellent because that is what it was designed for, and power draw is low, though higher than LoRaWAN. It has two drawbacks. A monthly subscription per SIM compounds quickly at a hundred devices, and you depend entirely on the operator's coverage at one specific location. NB-IoT coverage on distant reclaimed land does not necessarily match the voice coverage map, and the only way to know is a field measurement before purchase.

4G is the correct choice when the data is heavy: a camera, a controller that needs a permanent bidirectional link, or a single point that aggregates dozens of readings and uploads them in one batch. The price is power consumption an order of magnitude higher - a 4G device runs on a small battery for days, not years - plus a monthly data plan. That is entirely acceptable for one mains-powered device and entirely unacceptable for a hundred sensors.

The arrangement we arrive at on most large farms is hybrid rather than pure: field sensors on LoRaWAN reporting to one gateway, and the gateway backhauling the aggregate over 4G. One data plan instead of a hundred SIMs, and the field devices stay low-power. NB-IoT then has a natural place for scattered points far from any gateway - an isolated well, a measurement point on a canal - where a single device cannot justify a gateway.

Before we recommend any of the three we run a site survey: a temporary gateway and a test node carried between the proposed sensor positions to measure real signal strength, plus operator coverage measured at the same points. One day on site prevents a mistake whose correction is a full reinstallation.