IoT Connectivity: How To Choose The Right Technology

IoT connectivity is the network layer that makes smart devices useful, and the right choice depends on four variables: range, power budget, data throughput and cost. Wi-Fi, BLE, Zigbee, LoRaWAN, NB-IoT, 5G and Matter each fit different applications.

Key takeaways

Definition: IoT connectivity carries sensor data, firmware updates and control commands between devices and the cloud.
4 variables: Range, power budget, data throughput and cost decide the choice.
Fit: LoRaWAN and NB-IoT win on battery life, Wi-Fi on bandwidth, and 5G on latency.
Method: A 5-step framework ends in a proof of concept with 5 to 10 nodes in the real environment.
Dev Station: Dev Station runs IoT protocol evaluation, firmware and certification for US and UK teams.

Why does IoT connectivity decide project success?

IoT connectivity decides project success because the radio link carries sensor data, firmware updates, and control commands between edge devices and the cloud. A protocol that undersells your throughput hurts the user experience, and one that over-delivers on bandwidth burns battery, inflates data fees, and shortens device lifespan.

IoT connectivity technologies span a range from a few centimeters (Bluetooth Low Energy) to tens of kilometers (LoRaWAN, NB-IoT). They trade range against data rate, power consumption against latency, and infrastructure cost against coverage. There is no single best protocol, only the best fit for the constraints of an application. Connectivity shapes hardware selection, antenna design, firmware architecture, the cloud backend, and ongoing operational cost, so treat the protocol decision as a first-class product requirement.

Which IoT connectivity options exist, and how do they differ?

Seven technologies cover most projects: Wi-Fi, BLE, Zigbee, LoRaWAN, NB-IoT, 5G with LTE-M, and Matter. They sit at different points on the range, power, and bandwidth triangle, and each has strengths, limitations, and ideal use cases.

Wi-Fi (802.11 b/g/n/ac/ax)

The default high-bandwidth choice for devices that plug into wall power or charge daily. Wi-Fi 6 and Wi-Fi 6E deliver high throughput and low latency, which suits IP cameras, smart displays, and any device streaming rich media. The trade-off is power: a typical Wi-Fi IoT module draws hundreds of milliamps while active, which rules it out for battery telemetry. Onboarding can also be friction-heavy because each device needs SSID and credential provisioning.

Best for: indoor, mains-powered, high-data devices. Range: 30 to 100 m indoor. Power: high. Throughput: up to 9.6 Gbps (Wi-Fi 6E).

Bluetooth Low Energy (BLE)

BLE is the short-range choice for wearables, beacons, and smartphone-paired devices. It draws microamps in sleep mode and wakes in milliseconds, so coin-cell sensors can run for months. Bluetooth Mesh extends BLE to multi-hop networks of hundreds of nodes. The weak point is range, typically 10 to 30 m line-of-sight, and a small payload per packet, which makes BLE unsuitable for continuous streaming or long-haul telemetry.

Best for: wearables, beacons, phone-paired sensors. Range: 10 to 30 m. Power: ultra-low. Throughput: up to 2 Mbps.

Zigbee (IEEE 802.15.4)

Zigbee builds self-healing mesh networks over low-power 2.4 GHz radios and is the workhorse of home automation. Each mains-powered device acts as a router, which extends coverage without dedicated repeaters. The downside is gateway dependence: Zigbee devices need a hub to bridge to IP, and the mesh stays healthy only while router nodes stay powered. Battery devices are end-nodes and cannot relay.

Best for: smart-home mesh automation, lighting, sensors. Range: 10 to 100 m per hop. Power: low (routers need mains). Throughput: 250 kbps.

LoRaWAN

LoRaWAN is the leading non-cellular LPWAN. It uses unlicensed sub-GHz ISM bands to reach multi-kilometer range on milliwatts of power. A single gateway can cover an entire farm or factory, and devices can run for years on AA cells thanks to duty-cycle limits and adaptive data rate. The trade-off is throughput: data rates sit between 0.3 and 50 kbps, with duty-cycle regulations in most regions limiting airtime. LoRaWAN suits periodic telemetry such as meter reads, soil moisture, and asset presence, and it does not suit real-time control.

Best for: long-range, low-power telemetry in agriculture and asset tracking. Range: 2 to 15 km. Power: ultra-low. Throughput: 0.3 to 50 kbps.

NB-IoT (LTE Cat-NB1)

NB-IoT is the 3GPP-standardized cellular LPWAN, running on licensed spectrum inside existing LTE infrastructure. It penetrates basements and underground vaults that LoRaWAN cannot reach, supports long battery life with Power Saving Mode, and needs no gateway because the carrier handles coverage. The trade-offs are per-device SIM costs, recurring data fees, lower peak throughput (around 250 kbps downlink), and higher latency than LoRaWAN, which matters for time-critical reporting.

Best for: smart metering, fixed asset monitoring, deep-indoor sensors. Range: 1 to 10 km (cell-dependent). Power: ultra-low. Throughput: up to 250 kbps.

5G (and LTE-M / Cat-M1)

5G brings three capabilities no other IoT radio matches: very high peak throughput, ultra-reliable low-latency communication (URLLC), and massive machine-type connectivity for dense device populations. For Industry 4.0, autonomous vehicles, and real-time video analytics at the edge, 5G is the option that clears the latency bar. LTE-M (Cat-M1) is the lower-power cellular option, supporting mobility and voice over LTE at moderate throughput. Module cost, data plans, and power consumption keep 5G out of low-end telemetry, so reserve it for applications that need its speed or latency.

Best for: mission-critical, real-time, high-bandwidth industrial and vehicular use. Range: cell-dependent (300 m to 10 km). Power: high. Throughput: up to 10 Gbps.

Matter (over Wi-Fi / Thread)

Matter is the connectivity unifier the smart home has waited a decade for. Backed by the Connectivity Standards Alliance, Matter is an application-layer protocol that runs over Wi-Fi for high-bandwidth devices and Thread (802.15.4 mesh) for low-power ones. It guarantees interoperability across ecosystems, so an iPhone, an Echo, and a Google Nest can all control the same Matter bulb without vendor lock-in. Matter does not replace the underlying radio. It standardizes how devices discover, commission, and communicate.

Best for: cross-ecosystem smart-home devices. Range: inherits Wi-Fi or Thread. Power: varies by transport. Throughput: varies by transport.

How do the IoT connectivity technologies compare side by side?

The table condenses the technologies into the metrics that drive architecture decisions. Use it to eliminate non-starters before reading datasheets: a protocol that cannot meet your hard requirements on range, power, or throughput will not be fixed by optimization.

Technology	Range	Throughput	Power	Latency	Topology
Wi-Fi 6	30 to 100 m	Up to 9.6 Gbps	High	Low (ms)	Star
BLE 5.3	10 to 30 m	Up to 2 Mbps	Ultra-low	Low (ms)	Star / Mesh
Zigbee 3.0	10 to 100 m per hop	250 kbps	Low	Low (ms)	Mesh
LoRaWAN	2 to 15 km	0.3 to 50 kbps	Ultra-low	High (s)	Star-of-stars
NB-IoT	1 to 10 km	Up to 250 kbps	Ultra-low	High (s)	Cellular
5G (URLLC)	300 m to 10 km	Up to 10 Gbps	High	Ultra-low	Cellular
Matter	Varies	Varies	Varies	Low (ms)	Star / Mesh

Power in the table reflects the active draw of the radio, and total system consumption differs. A LoRaWAN sensor reporting once an hour can run for years on AA cells, while the same sensor on Wi-Fi would drain its battery within days. Model duty cycle, sleep current, and wake frequency against your actual reporting cadence.

Which IoT connectivity technology fits which use case?

NB-IoT fits smart metering, LoRaWAN fits agriculture, 5G fits industrial control, BLE fits wearables, Matter fits smart-home lighting, Wi-Fi fits cameras, and LTE-M fits fleet tracking. The mapping follows the trade-offs in the comparison table.

Use case	Primary choice	Why	Alternative
Smart metering (water and gas)	NB-IoT	Deep indoor penetration, long battery life, no gateway	LoRaWAN
Precision agriculture	LoRaWAN	Multi-km rural range, no cellular coverage needed	NB-IoT
Industrial robotic control	5G URLLC	Ultra-low latency, deterministic delivery	Wi-Fi 6 (TSN)
Wearable health monitor	BLE	Coin-cell friendly, native phone pairing	Zigbee
Smart-home lighting	Matter / Thread	Cross-ecosystem, low-power mesh, no vendor lock-in	Zigbee 3.0
IP security camera	Wi-Fi 6	Multi-Mbps video stream, mains powered	5G (remote sites)
Fleet asset tracking	LTE-M	Mobility, handover support, moderate throughput	NB-IoT (fixed)

How do you choose an IoT connectivity technology?

Choose in five steps: define hard constraints, map the coverage footprint, model power and duty cycle, score total cost of ownership, and build a proof of concept. Run the framework once for each device class, because a single project often needs more than one radio.

Define hard constraints first. Write down the non-negotiables before evaluating any protocol: maximum battery life, maximum latency, minimum throughput, deployment environment (indoor, outdoor, underground), and unit cost ceiling. Hard constraints eliminate many candidates immediately. If a sensor must run 10 years on a coin cell, Wi-Fi and 5G are already off the table.
Map the coverage footprint. Where do the devices physically live? Indoor basement meters demand deep penetration (NB-IoT). Rural soil sensors need unlicensed long range (LoRaWAN). Factory floor robots need low-latency coverage across a bounded area (5G private network or Wi-Fi 6). Devices that move between cells need cellular handover (LTE-M or 5G), not LoRaWAN. Carrier coverage maps and your own RF site survey should drive this step.
Model power and duty cycle together. Module datasheets quote peak current, but battery life is governed by average consumption: sleep current times idle time plus active current times transmit time. A LoRaWAN sensor reporting every 10 minutes draws far less average current than a BLE device polling every second. Build a spreadsheet that models your reporting cadence against the module power profile, then translate that into battery life.
Score total cost of ownership. Compare protocols on five-year TCO, not just module price. LoRaWAN modules are cheap but you fund the gateway and its backhaul. NB-IoT has no gateway cost but carries per-SIM activation and recurring data fees. 5G modules and data plans cost more. Wi-Fi has near-zero operational cost if you own the LAN. Include certification, integration, maintenance, and end-of-life replacement.
Build a proof of concept. Never scale on datasheets alone. Buy a developer kit for your top two candidates, deploy 5 to 10 nodes in the actual operating environment, and measure range, latency, packet loss, and battery drain. RF propagation is site-specific: concrete walls, metal cabinets, and competing 2.4 GHz traffic can cut a theoretical range sharply.

Hybrid designs are normal. Many production deployments run two radios, for example BLE for on-device commissioning via smartphone and NB-IoT for backhaul telemetry, or Wi-Fi for high-bandwidth video with a Thread fallback for low-power sensor nodes. Designing for a second radio from day one is cheaper than retrofitting one after launch.

Frequently asked questions

Five questions come up most often when teams choose IoT connectivity.

Which IoT connectivity technology has the longest range?

LoRaWAN reaches 2 to 15 km, and NB-IoT reaches 1 to 10 km depending on the cell.

Which technology suits battery-powered sensors?

LoRaWAN and NB-IoT, because both are ultra-low power and suit small, periodic data packets.

Is Matter a replacement for Wi-Fi or Thread?

No. Matter is an application-layer protocol that runs over Wi-Fi and Thread and standardizes how devices discover, commission, and communicate.

When should a product use two radios?

When one radio cannot cover both commissioning and telemetry, for example BLE for phone commissioning and NB-IoT for backhaul.

Why run a proof of concept before choosing a protocol?

RF propagation is site-specific, and concrete walls, metal cabinets, and competing 2.4 GHz traffic can change real range and battery drain sharply.

Dev Station Technology helps businesses in the United States and the United Kingdom move from protocol evaluation to deployed IoT solutions. Our engagement covers six stages:

Requirements deep-dive. We pressure-test your range, power, throughput, and cost targets against real-world physics before any hardware is specified.
Unbiased technology evaluation. We benchmark the candidate protocols in your operating environment and present trade-offs in business terms.
Proof-of-concept development. We build and instrument a PoC with your shortlisted radios and measure actual performance against your hard constraints.
Hardware and firmware engineering. We design antenna layouts, select certified modules, and write production firmware that extracts the best battery life from your chosen protocol.
Carrier and vendor sourcing. We negotiate with network operators and module vendors on data plans, lead times, and volume pricing.
Scale and certification. We take your product through regulatory certification (FCC, CE, MIC) and launch at production volume with monitoring and OTA update infrastructure.

Contact Dev Station Technology at sale@dev-station.tech or visit dev-station.tech to schedule a consultation with our IoT engineering team. Device and telemetry data stays in your own cloud tenant, in the region your policy requires. Where a client needs SOC 2, HIPAA or UK GDPR evidence, we build the technical controls those frameworks ask for and work alongside the assessor who issues the certificate. Our engineers work from Vietnam with overlap into US Eastern, US Pacific and UK GMT hours, and we invoice in USD or GBP.

Ask an AI about this