Cellular vs Satellite vs Radio Telemetry

Back to Blog

How the Three Compare

Cellular is usually the first choice where coverage exists: cheap hardware, low latency (tens of milliseconds on LTE), ample bandwidth, and a recurring per-SIM data cost. Its weakness is coverage - no tower, no link - and dependence on a carrier. Licensed radio is an operator-owned network with no airtime cost and strong control, ideal for gathering systems and pipelines, but it carries less bandwidth and means owning and maintaining the infrastructure.

Satellite reaches anywhere the sky is visible, which makes it the backhaul of last resort for remote basins, offshore, and wilderness pipeline segments. Geostationary satellite adds roughly half a second of latency and data is costly; newer low-Earth-orbit services cut latency to tens of milliseconds and improve economics, making satellite viable at more sites and as a failover behind cellular.

Choosing and Mixing

The right answer is rarely one technology for a whole fleet. A common pattern is licensed radio as the field backbone tying local sites to a hub, then cellular for the hub's trip to the cloud, with satellite as backup where cellular is unreliable. Sites that must not lose data often carry dual paths - cellular primary, satellite failover - with automatic switchover in the gateway. The decision comes down to what coverage is actually available at each location, the latency and bandwidth the application needs, and the total cost over the asset's life.

Whatever the transport, the payload is the same: the RTU or flow computer's data over a protocol like Modbus, DNP3, IEC 60870, or MQTT. A cloud SCADA such as Merobix reads that data once it arrives, treating cellular, satellite, and radio as interchangeable transport underneath, so operators can pick the best link per site without changing the monitoring platform.

Ownership, Cost Structure, and Failure Profile

A compact side-by-side helps when presenting the choice to management:

Factor	Cellular	Satellite	Licensed radio
Coverage limit	Carrier tower footprint	Sky visibility at the antenna	The radio path you engineer yourself
Network owner	Carrier	Satellite operator	You
Recurring cost	Per-SIM data plans	Airtime or capacity plans	License fees and your own maintenance
Typical outage causes	Tower faults, congestion, technology sunsets	Weather fade, antenna obstruction	Path obstruction, interference, repeater power
Who fixes an outage	The carrier, on their schedule	The satellite operator, plus your dish	Your crew, on your schedule

The ownership row explains most long-run satisfaction and frustration with each option. Cellular and satellite outages are opened as tickets and waited out; radio outages are truck rolls you control. Which of those is worse depends on the operation - a carrier restores a tower without your involvement, but you cannot expedite them, while your own radio network fails only as often as your maintenance program allows.

Failure Modes Matter More Than Headline Speeds

The three transports fail differently, and designing for the failure profile matters more than comparing best-case throughput. Cellular fails through carrier-side events: tower outages, congestion during regional events, and, on a longer horizon, generation sunsets that eventually strand older modems - fleet planning should track modem generations the way it tracks firmware. Satellite fails through the sky and the site: rain fade on some bands, snow or debris on the dish, and anything that grows or gets built into the antenna's view. Licensed radio fails through the path and the infrastructure: vegetation growth, corroded connectors, and power problems at repeater sites.

Monitoring should therefore watch the link, not just the process data. Trending signal quality per site, alarming on sustained degradation, and testing backup paths on a schedule turns communications from a background assumption into a maintained asset. Where dual paths exist, the switchover logic deserves engineering attention too - a failover hold-down timer prevents a marginal primary link from flapping the site between transports.

A Worked Site Selection, Step by Step

For a single new site, the selection walks in order:

1. Survey actual coverage at the site with test hardware - carrier maps and predicted footprints are starting points, not answers.
2. Estimate the data volume from the tag count, poll rate, and protocol overhead; whether you poll or use report by exception changes the number substantially.
3. Check the power side: transmit hardware appears in the site's power budget worksheet, and at a solar site the radio or modem is often among the larger loads.
4. Price the whole life: hardware, installation, recurring charges, and the eventual modem or radio refresh.
5. Decide what an outage costs at this particular site; that answer determines whether a second path is justified.

The data-volume arithmetic is worth doing symbolically before signing anything: bytes per poll, times polls per day, times sites, gives a monthly volume to compare against plan tiers - and that comparison frequently changes the poll rate rather than the transport. Sites with generous plans get polled fast; satellite sites on tight capacity get event-driven reporting and a slower background poll. The transport decision and the polling strategy are one decision wearing two hats.

Frequently Asked Questions