A LoRaWAN gateway is the bridge that receives radio packets from field sensors and forwards them over Ethernet, 4G or Wi-Fi to a network server. It listens to every device in range rather than pairing with one. Choose it on channel capacity, antenna, enclosure, backhaul and mounting point, and set the number you need by measuring on site.
Where projects usually stand when this question comes up
The radio technology has already been chosen: small payloads, battery-powered devices, a large site, and the decision has gone to LoRaWAN. We walked through how that decision is made in our LoRa vs NB-IoT vs cellular comparison. The next question is concrete: how many gateways, where, and which model?
At this point most teams look at a datasheet and its headline range figure. Yet the gateway is the component you install in the smallest numbers and the one that silences the most devices at once when it fails. If a single soil moisture sensor dies, one plot loses its readings; if the gateway goes down, the whole farm goes dark.
What a LoRaWAN gateway is and where it sits in the architecture
LoRaWAN uses a star-of-stars topology with four layers:
- End device (sensor node): takes the measurement, encrypts a small packet and broadcasts it. It has no idea which gateway will hear it.
- Gateway: receives every packet in range, adds signal strength (RSSI) and signal-to-noise ratio (SNR), and forwards it to the network server. It does not decrypt the payload.
- Network server: removes duplicates when several gateways hear the same packet, authenticates the device, manages data rates (ADR) and picks the best-placed gateway for any downlink.
- Application server: decodes the data and hands it to a dashboard, database or business system, very often over the protocol covered in our guide to MQTT in industrial IoT.
The practical consequence matters. A gateway is a deliberately "dumb" relay: if a sensor sits within range of two gateways, data keeps flowing when one of them fails. Because payloads are encrypted end to end, someone who physically seizes a gateway cannot read sensor data, although its management interface still needs protecting.
What a poorly planned gateway layout costs you
Start with realistic expectations on range. According to the LoRa Alliance's overview of LoRaWAN, LoRaWAN is an ITU-approved standard (ITU-T Y.4480) designed for range of up to 15 km in rural areas and device battery life of more than 10 years. Those figures describe open terrain and good mounting; concrete, metal cabinets and hills cut real-world range sharply.
IoT Analytics expects the number of connected IoT devices to grow 14% to 21.1 billion in 2025 and reach 39 billion by 2030. (IoT Analytics — State of IoT 2025)
Eurostat's data on IoT use in enterprises shows that in 2021, 30% of EU enterprises using IoT did so to manage energy consumption and 24% to monitor equipment condition in real time. When monitoring data stops, decisions stop with it. A badly planned gateway layout creates three separate costs:
- Blind spots: devices with weak signal drop packets, and alarm rules built on patchy data can no longer be trusted.
- Flat batteries: distant devices fall back to slower data rates (higher spreading factors), keep their transmitter on longer and drain their batteries well ahead of plan.
- Single points of failure: if each device is heard by only one gateway, a power cut or a lost 4G link silences an entire area.
LoRaWAN gateway selection criteria
The range figure on a datasheet is not a selection criterion. The table below summarises what to check when comparing quotes and before going on site.
| Criterion | Why it matters | How to check it |
|---|---|---|
| Frequency plan | Türkiye follows the European 868 MHz plan (EU868); a unit built for another region will not work | Band information on the label and declaration of conformity |
| Channel count | Sets how many packets can be received at once; eight channels may not suffice on a busy site | Compare with expected device count and reporting interval |
| Indoor or outdoor enclosure | Outdoor units must withstand dust, rain and heat | Ingress protection rating, operating temperature range |
| Antenna | Gain and mounting height are the two biggest drivers of range | External antenna support, cable length and loss |
| Backhaul | The gateway itself must reach the server | Ethernet, 4G and Wi-Fi options; failover link |
| Power | Off-grid sites need solar and battery | PoE, 12/24 V input, runtime during an outage |
| Remote management | Monitoring and updates without a site visit | Firmware updates, VPN, health reporting |
| Buffering | Packets should survive a short backhaul outage | Local store-and-forward support |
What ingress protection ratings mean for outdoor enclosures is covered in our IP ratings explainer. Indoor gateways work well for offices, warehouses and factory floors; across open land, an outdoor unit with a mast-mounted antenna is usually unavoidable.
How many gateways do you need? A placement plan step by step
Gateway count comes from site measurement, not a formula. These seven steps connect the estimate on the desk to reality in the field:
- Map the devices. Mark each sensor's approximate location, mounting height and surroundings (basement, manhole, metal cabinet, open ground).
- Shortlist candidate sites. List high points with power, such as rooftops, water towers, lighting columns or existing radio masts, and raise landlord permission early.
- Measure with a test node. Put a temporary gateway at each candidate site and log RSSI and SNR at the hardest sensor locations.
- Design for overlap. Make sure critical devices are heard by at least two gateways; leave no area dependent on a single unit.
- Verify backhaul and power. Check 4G signal, Ethernet availability and power continuity at every site.
- Run a pilot area. Over a few weeks, track packet loss, battery drain and gateway availability.
- Document the result. Record gateway locations, antenna types and connection details in the asset register; it becomes the map for every future device.
Livestock on pasture, a remote pumping station and scattered meters each call for a different layout. We covered coverage planning for moving assets in our livestock monitoring guide, and for fixed, near-underground devices in our automatic meter reading guide.
Private network or operator network?
You do not have to own a gateway to use LoRaWAN; there are three operating models. In a private network, you install the gateways and the network server yourself and the data stays on your own infrastructure. In an operator network, you use the gateways of a LoRaWAN operator serving the area and pay a per-device service fee. A hybrid model adds your own gateways wherever operator coverage is weak.
Where no operator network exists, typically in rural areas, a private network is often the only route for agriculture, water and energy projects. That brings the network server into scope too: whether it runs in the cloud or on premises, how it is backed up and who patches it. How to store sensor readings over the long term is covered in our time-series database guide.
Security and the regulatory picture in Türkiye
LoRaWAN encrypts application data with AES-128 and uses a separate network key to check the integrity of every message, and over-the-air activation (OTAA) is safer than activation by personalisation (ABP), which hard-codes keys into the device. The weak link is usually not the radio but the gateway itself: default passwords, a management page exposed to the internet and firmware nobody updates. If the gateway connects to your corporate network, put it in its own segment and manage it over a VPN; we looked at the wider field-device attack surface in our OT security and SCADA article.
On regulation, LoRaWAN runs in unlicensed spectrum, so a separate frequency licence is generally not required for devices used in that band. Devices must still comply with the rules of Türkiye's telecoms regulator, the Information and Communication Technologies Authority (BTK), for short-range radio devices: staying within output power and duty cycle limits and being placed on the market with a declaration of conformity. Check the current BTK rules before installation. Duty cycle limits vary by sub-band and constrain how often the gateway can send downlink commands, so any design that relies on frequent commands to sensors should be questioned early.
How we approach LoRaWAN gateway projects at Digital Bridge
We do not treat the gateway as a purchase separate from the sensor, or the sensor as separate from the dashboard. When the end device, gateway layout, network server and software are planned by one team, a lost packet has one owner.
- We start with a free technical feasibility study. We capture device count, reporting interval, mounting points, power and connectivity, then share a written recommendation on gateway numbers and models. This is the first step of our IoT device design and manufacturing service.
- We survey and measure on site. We set up temporary gateways at candidate points and measure signal at the hardest sensor locations; the design is finalised on those results.
- We build for the application. In fields and greenhouses through our precision agriculture IoT work, and in meter reading through our OSOS meter reading system, we weigh LoRaWAN against GPRS/4G and NB-IoT according to local conditions. Our agriculture and livestock solutions run on the same foundation, and the irrigation side is covered in our irrigation monitoring guide.
- We bring the data into one place. Sensor readings from every site are collected on the map and dashboard of our remote monitoring platform, and the right person is notified when an alarm condition you define is met, for example when a reading crosses a threshold. Data flows on to your ERP, SCADA or own software via API.
The platform side is explained in our IoT remote monitoring platform guide, and the journey from idea to series production in our IoT device development process.
For on-site processing, see edge computing in manufacturing; more field hardware guides sit in the IoT and hardware hub.
Next step
If you are planning a LoRaWAN deployment, gather the following: approximate device count and locations, how often each device will report, candidate high points for gateways, and the power and internet situation at each. Then get in touch with us, and we will set gateway numbers and positions together by measuring on site rather than guessing.