When you launch a connected-device project, the choice of radio protocol is one of the most structuring decisions in your specifications. It shapes battery life, coverage, total cost of ownership and even the long-term commercial viability of the product. No protocol is universal: it all depends on the use case, the environment and economic constraints. This guide gives you the method to choose between LoRaWAN, NB-IoT, LTE-M, Zigbee, Bluetooth LE, Matter, UWB and the others.
Why the choice of protocol is decisive
A radio protocol is more than a frequency or a data rate. It defines how your connected devices communicate with the rest of the system: the volume of data exchanged with each transmission, energy consumption, network topology, security level, latency and, above all, the economic model of your infrastructure.
A poor decision at the scoping stage can be very costly: insufficient battery life (and therefore tripled maintenance costs), range unsuited to the actual deployment environment, dependence on an operator whose pricing changes over time, or an inability to scale the fleet. On a product manufactured in the thousands, the gap between the right and the wrong protocol quickly amounts to hundreds of thousands of euros.
At Altyor, we systematically address the choice of radio protocol upstream, during the consulting phase, before any hardware development begins. It is less expensive, more reliable, and avoids discovering the mistake after manufacturing scale-up.
The 7 criteria to weigh up before choosing
Before comparing acronyms, ask yourself the right questions. Here is the framework we apply to every IoT project we support.
- Required range. From a metre (BLE, UWB) to several kilometres (LoRaWAN, NB-IoT, Sigfox, LTE-M). This is filter number one: it eliminates half the candidates straight away.
- Energy consumption. How much energy the protocol uses with each transmission, but also during sleep. A protocol that is frugal in transmission yet talkative in synchronisation can drain a battery as quickly as a poorly configured Wi-Fi connection.
- Data volume and frequency. A few bytes per day (temperature sensor) or several MB per hour (camera, audio, firmware)?
- Network topology. Star (LPWAN, cellular), mesh (Zigbee, Thread, Z-Wave), point-to-point (BLE, UWB). Topology determines resilience and ease of deployment.
- Security. AES-128 encryption or higher, device attestation, secure OTA updates, cybersecurity certification (Cyber Resilience Act from 2027). On a device sold at scale, this criterion becomes essential.
- Total cost of ownership. Radio module price, operator subscription, gateway infrastructure, maintenance, service life. Compare costs over 5 years, not just at deployment.
- Regulation. ISM bands at 868 MHz in Europe, 2.4 GHz worldwide, duty cycle imposed by ETSI, country-specific constraints. The target market dictates which protocols are compatible.
The three main families
IoT radio protocols generally fall into three broad categories, depending on the range required:
Long range – LPWAN
Designed to cover long distances with very low power consumption. Ideal for outdoor deployments, urban or agricultural sensor networks.
- Range: 1 to 50 km
- Power: very low
- Throughput: low
- Typical protocols: LoRaWAN, Sigfox, NB-IoT, LTE-M
Short range – WPAN
Protocols for confined environments (home, building). Often organised as a mesh network to extend coverage. Higher throughput.
- Range: 10 to 100 m
- Power: low
- Throughput: medium
- Typical protocols: Zigbee, Z-Wave, Thread, Matter, Bluetooth LE
Very short range – Wi-Fi / BLE
High-bandwidth protocols, derived from consumer computing. Useful when data throughput is the priority or the infrastructure is already in place.
- Range: 1 to 50 m
- Power: variable
- Throughput: high
- Typical protocols: Wi-Fi, Ultra Wide Band (UWB), BLE
Comparison of the main protocols
| Protocol | Range | Power | Throughput | Topology | Band | Security | Cost | Best for |
| LoRaWAN | 2 – 15 km | Very low | 0.3 – 50 kbps | Long-range star | 868 MHz | AES-128 | € | Smart city, agriculture, logistics |
| Sigfox | 10 – 50 km | Very low | 100 – 600 bps | Star (operator) | 868 MHz | AES-128 | € | Trackers, alerts, simple monitoring |
| NB-IoT | 1 – 10 km | Low | ≈ 200 kbps | Cellular | Licensed | 3GPP / SIM | €€ | Meters, metering, indoor |
| LTE-M | 1 – 10 km | Low | ≈ 1 Mbps | Cellular | Licensed | 3GPP / SIM | €€ | Mobile asset tracking, voice, large OTA |
| Zigbee | 10 – 100 m | Low | 250 kbps | Mesh | 2.4 GHz | AES-128 | € | Home automation, smart building |
| Z-Wave | 30 – 100 m | Low | 9.6 – 100 kbps | Mesh | 868 MHz | AES-128 | € | Connected home, retrofit |
| Thread / Matter | 10 – 100 m | Low | 250 kbps | IP mesh | 2.4 GHz | AES-128 + DTLS | € | Interoperable smart home |
| Bluetooth LE | 1 – 50 m | Very low | 1 – 2 Mbps | Star / mesh | 2.4 GHz | AES-128 | € | Wearables, healthcare, retail |
| Wi-Fi | 10 – 50 m | High | 100 Mbps – 1 Gbps | Star (AP) | 2.4/5/6 GHz | WPA3 | € | Cameras, mains-powered devices |
| UWB | 10 – 30 m | Moderate | ≈ 27 Mbps | Point-to-point | 6 – 8.5 GHz | Native cryptography | €€ | Centimetre-precise indoor positioning |
💡 Key takeaway: The right protocol is not the most powerful — it is the most appropriate. A temperature sensor in an open field running on an AA battery does not need Wi-Fi. Performance must always be weighed against the actual need.
Use cases: which protocol for which project?
Connected agricultural sensor (soil moisture, temperature)
A connected IoT sensor deployed in an open field needs to last at least 5 years on battery, with transmissions limited to a few messages per day. The maintenance cost of a sensor in the field is ten to twenty times higher than its hardware cost: battery life trumps everything.
Recommended choice: LoRaWAN. Kilometre-scale range, very low power consumption, private network deployable with a single gateway covering 100 hectares. A 30 to 50% saving on maintenance visits compared with a cellular solution.
Lighting and HVAC management in a commercial building
Dense network within a limited perimeter, near-real-time commands, infrastructure often already wired. Robustness and multi-vendor interoperability are decisive in avoiding lock-in to a proprietary ecosystem.
Recommended choice: Thread / Matter or Zigbee. Robust mesh network, low latency, and — for Matter — native interoperability between manufacturers. A major asset for buildings whose service life exceeds product cycles.
Fleet management and asset tracking
National or European mobility, regular position updates, on-board power or rechargeable battery. The choice lies between cellular coverage (national, partial indoor) and indoor accuracy (warehouses, fine-grained logistics).
Recommended choice: LTE-M for long-distance mobility, UWB for indoor tracking. LTE-M offers cellular coverage with controlled subscription costs and native roaming support. UWB becomes essential as soon as the required accuracy drops below the metre, in warehouses or industrial environments.
Health monitoring patch (medical wearable)
Proximity to a smartphone, regular synchronisation, lightness and discretion as priorities. Regulatory compliance (medical device) and cybersecurity often close the discussion.
Recommended choice: Bluetooth LE. Native integration into every smartphone, low power consumption, standardised APIs (BLE GATT, health profiles), and an ecosystem of certified modules that accelerates time to market.
Quality control camera in industrial production
Video streaming, near-real-time image analysis, mains power available. Latency and bandwidth reign supreme.
Recommended choice: Wi-Fi 6 or Wi-Fi 7. More than enough bandwidth, low latency, infrastructure often already in place. Wi-Fi 6 brings the density required in workshops saturated with connected devices; Wi-Fi 7 anticipates the spread of edge-side embedded AI.
Our approach at Altyor
The choice of radio protocol is an integral part of our upstream consulting phase. Before touching a single component, we analyse the constraints of your project: deployment environment, target battery life, transmission frequency, data volume, connectivity cost, geographic market and roadmap for the fleet.
Our teams then design electronic boards integrating market-certified radio modules (LoRa, BLE, NB-IoT, LTE-M, Zigbee, Thread, UWB, Wi-Fi) and guide the choice of the appropriate network infrastructure — whether a private LPWAN, a public IoT operator or integration with an existing cloud platform. Our manufacturing expertise ensures compliance (CE, FCC, sector-specific certifications) and the long-term viability of your solution throughout its life cycle.
Every project is unique. That is why we never offer off-the-shelf solutions: we build a bespoke approach, from scoping the need to volume delivery.
Have a connected sensor project?
Let’s discuss the radio protocol best suited to your constraints. Our engineering team supports you from design to series production.
Get a quote tailored to your project!
FAQ: IoT radio protocols
Is there a universal radio protocol for IoT?
No, and that is precisely what makes the topic strategic. Every protocol is a trade-off between range, power, throughput and cost. An industrial camera and a temperature sensor in an open field cannot share the same technology. That is why an upstream scoping exercise, grounded in your business constraints, makes the difference between a profitable project and one that runs over in operation.
What budget should I plan for the connectivity of a connected device?
Two cost lines structure the budget: the hardware cost of the radio module (€1 to €15 per unit depending on the technology) and the operating cost (operator subscription, network infrastructure, monitoring). Over five years and a fleet of 1,000 devices, the gap between a private LPWAN and an operator-managed cellular solution can amount to hundreds of thousands of euros. The right trade-off depends on the number of devices, their geographic dispersion and your roadmap.
Will my product work everywhere in the world?
Not automatically. ISM frequency bands differ by region (868 MHz in Europe, 915 MHz in North America, 920 MHz in Asia), which often imposes a hardware variant per market. Cellular protocols (NB-IoT, LTE-M) are the easiest to internationalise, thanks to roaming agreements between operators. This constraint must be anticipated at design stage: addressing it after manufacturing scale-up costs much more.
Will Matter replace Zigbee and Z-Wave?
Not immediately, but Matter reshuffles the deck. It is an IP application layer that runs on top of Thread (mesh radio) or Wi-Fi, and guarantees interoperability between the Apple, Google, Amazon and Samsung ecosystems — something neither Zigbee nor Z-Wave could deliver. For a new consumer product, Matter on Thread is currently the most future-proof choice. Zigbee remains relevant for existing fleets and certain mature industrial ecosystems.