A tracker that reports every ten minutes along the interstate and then goes dark for six hours in a gravel yard is not asking for more data. It is telling you the radio link failed, and that failure has a specific, findable cause. Global cellular Internet of Things connections are forecast to reach 5.9 billion by 2035, up from 4.3 billion in 2026, according to industry forecasts for cellular IoT, and a growing share of those devices will sit in places mobile networks were never engineered to serve. That makes IoT connectivity coverage in remote areas a design decision, not a procurement afterthought.
Here is the fact that explains most of what follows: mobile networks are engineered to maximize population coverage rather than geographic area coverage. A carrier can accurately claim near-total population coverage in a country while leaving large parts of its landmass thin, patchy or dark. Every fix below exists because of that gap.
This guide is written for the people who actually own the problem: fleet, logistics and asset-tracking operators whose devices leave the covered map every day. It works through the five real causes of poor IoT coverage in remote areas, the five fixes in the order you should try them, how to audit a provider’s coverage claim before you sign, and three worked deployment scenarios with recommended configurations. If you want the shorter overview of remote IoT connectivity first, we have written separately on reliable connectivity in remote locations; this piece is the diagnostic version of it, aimed at teams already losing devices rather than teams planning a first deployment.
Why IoT Devices Lose Coverage in Remote Areas: The Five Real Causes
Before changing anything, identify which of these five is actually happening. They look identical in a dashboard (a device that stops reporting) and they have completely different fixes. Our own step-by-step troubleshooting method walks through the diagnostic order in more depth, but for remote IoT connectivity these are the five causes worth checking first.
- No carrier presence at all. There is no macro cell within reach of the device. This is genuine geographic absence, common on rural highways, mine sites, pipelines, farmland and long stretches of the US Interstate system between towns. No SIM change fixes it, because there is nothing to attach to.
- Weak structural penetration. A signal exists outdoors but cannot get inside. Concrete, steel racking, refrigerated panels, container walls and below-grade docks all attenuate cellular signal severely. Signals can weaken by up to 60 dB in basement areas, which is enough to take a working link to nothing.
- Roaming agreement gaps. A network is present and strong, but your provider has no commercial agreement with that specific operator in that specific country, so your SIM is not permitted to attach. The device sees bars and still cannot connect.
- Band and technology mismatch. The device modem does not support the band or radio technology the only available local carrier uses. A module certified for North American LTE bands can be functionally deaf in a market that has refarmed spectrum differently, and LTE-M or NB-IoT is only available where local carriers have deployed it.
- Antenna, placement and enclosure. The device is fine, the network is fine, and the installation is throwing the link away. An internal antenna inside a metal enclosure, a coax run that is too long, or a tracker mounted low behind a steel bulkhead will cost more signal than any tariff change can recover.
Cause one and cause two are physics. Causes three, four and five are choices, which means most IoT coverage gaps in practice are commercial and engineering problems wearing a physics costume.
Fix 1: Multi-Network SIMs and How Carrier Selection Actually Works
A multi-network IoT SIM connects to more than one mobile network in each country and across countries, rather than being tied to a single operator. The SIM continuously reports available networks and signal strengths to the device modem, and when the serving signal deteriorates the connection switches to a stronger available network. For assets crossing rural zones or borders, this is the single highest-return change you can make to IoT connectivity coverage in remote areas.
The part buyers usually miss is how the switch is decided. Dynamic roaming optimization, sometimes called steering of roaming, is the mechanism a provider uses to influence which network your SIM prefers when several are available. Steering can be commercial (push traffic to the cheapest partner), technical (prefer the operator with the best measured performance), or a blend. Aggressive commercial steering is a common hidden cause of poor rural performance: the SIM keeps re-selecting a cheap partner with thin rural build instead of the operator that actually covers the road your trucks are on.
Two questions settle it. First, ask whether steering is configurable per country and per account, or fixed at the platform level. Second, ask whether the SIM will fall back to any permitted network when its preferred one is unavailable, or whether it will keep retrying the preferred network and burn battery doing it. We have compared the leading multi-network SIMs on exactly these behaviors, because the differences between providers on remote IoT connectivity show up in steering policy far more than in headline country counts.
Multi-IMSI vs Global Roaming: Which One Solves Your Gap
These get used interchangeably and they are not the same thing.
Global roaming means one network identity that is permitted to roam onto partner networks abroad. The device is always a visitor. It works well, it is simple, and it is exposed to two risks: permanent-roaming restrictions in countries such as Brazil, Turkey, India and China, and roaming rate cards that make high-volume markets expensive.
Multi-IMSI means two or more network profiles (IMSIs) stored on the same physical SIM or eSIM. An embedded applet detects the device’s location and selects the most suitable profile and the matching regional data plan, so the SIM can behave like a local card in the markets where a local identity matters. That gives you local rates and local regulatory standing without touching the hardware.
For vehicle telematics the practical rule is straightforward. If your assets stay inside one regulatory bloc and coverage is the only problem, multi-network roaming solves it. If your assets cross into markets that restrict permanent roaming, or if roaming rates are the reason a route is unprofitable, multi-IMSI data plans are the better answer for IoT connectivity coverage in remote areas, because they change your commercial standing on the network rather than just your fallback list.
Fix 2: eUICC and Remote Profile Switching for Cross-Border Assets
eUICC is a SIM architecture that allows full operator profiles to be downloaded and swapped over the air, rather than being fixed at manufacture. SGP.32 is the GSMA remote SIM provisioning specification written specifically for IoT devices, which matters because it does not assume a screen, a user or a consumer-style menu to approve a profile change.
The reason this belongs in a remote-coverage guide is arithmetic. A physical SIM swap on a truck that is 400 km from the depot costs a service visit; a profile push costs a few minutes of a platform operator’s time. Once an asset is deployed somewhere hard to reach, the ability to change carrier without touching the device is the difference between fixing a coverage problem this week and fixing it at the next scheduled maintenance.
A pattern that works well in practice is a hybrid: use a multi-IMSI profile as the first, bootstrap-style identity so the device connects the moment it powers on anywhere, then push a localized operating profile over the air once the asset settles into its real operating region. That gives you one global SKU at manufacture and optimized IoT coverage in remote areas after deployment, which is the combination most cross-border fleets are actually looking for.
One caution worth stating plainly: eUICC only helps if the destination profile exists. Ask any provider to confirm, in writing, which operator profiles they can deliver over the air in your specific countries, not how many profiles exist in their catalog overall.
Fix 3: Choosing the Right Radio Technology for Hard-to-Reach Places
Coverage is not one thing. Changing the radio technology can extend reach further than changing the carrier.
NB-IoT vs LTE-M: Performance Differences That Matter in Logistics
NB-IoT is a narrowband cellular technology using a 200 kHz channel in licensed spectrum, designed for deep indoor penetration, very low power and small payloads. LTE-M is a wider-band low-power cellular technology that supports higher throughput, voice and, critically, handover between cells while moving.
For logistics the split is clean. LTE-M suits anything that moves, because a tracker on a trailer at highway speed needs to hand over between cells without dropping the session. NB-IoT suits stationary or near-stationary assets: yard sensors, tank monitors, utility meters, cold-store probes that stay in one building. NB-IoT reaches roughly 1 km in urban settings and up to around 10 km in rural ones, and it holds up unusually well indoors thanks to the higher link budget of licensed spectrum, with field tests keeping connections at very low signal levels. Using NB-IoT for a moving asset is one of the more expensive mistakes in this category, because the coverage looks fine in the yard and collapses on the road. There is a full NB-IoT primer if you need the deeper technical detail before committing a hardware design to it, and both technologies are available for remote IoT connectivity only where local carriers have actually deployed them.
LPWAN vs Cellular for Logistics Asset Tracking
LPWAN covers low-power wide-area technologies including LoRaWAN and Sigfox that trade throughput for range and battery life, typically transmitting small packets over distances that cellular cannot match at the same power budget. LoRaWAN runs in unlicensed spectrum and reaches 2 to 5 km in urban environments and up to around 15 km in rural ones, with devices lasting years on a battery. The LoRa Alliance reported that LoRaWAN reached 125 million deployed devices globally in the LoRaWAN ecosystem update it published in February 2026, so the ecosystem is mature enough to build on where it fits.
The trade is control of the infrastructure. Cellular gives you national reach with no gateways to own; LPWAN gives you reach into places no carrier will ever build, but you own the gateways. For logistics, that usually means cellular for anything that travels between sites and LoRaWAN for dense sensing inside a site you control, such as a distribution yard, a port terminal or a farm. Our comparison of how LPWAN and NB-IoT differ covers the spectrum, payload and power trade-offs in detail, and it is the right starting point when rural IoT coverage is the constraint rather than data volume. Choosing between them is one of the few decisions that changes remote IoT connectivity at the physics level rather than the contract level.
How to Assess LPWAN Coverage Area Before You Commit
Vendor coverage maps for unlicensed LPWAN are modeled, not measured. Assess a proposed LPWAN deployment on four criteria: gateway line of sight to the furthest planned endpoint, the link budget you have left after enclosure and mounting losses, the packet delivery rate you measure at the edge of the intended area rather than at the middle, and whether the endpoints will ever move outside the gateway footprint. If the answer to the last one is yes, you need cellular, whatever the range chart says. Gateways you own are only a fix for rural IoT coverage inside a boundary you control.
Nationwide 5G and LTE for Fleets
For fleets, 5G is a throughput and latency story, not a coverage story. Video telematics, in-cab AI and high-frequency uploads benefit from it; a location ping does not. Most 5G in commercial IoT plans today is 5G NSA, which still relies on the LTE anchor layer, so your effective rural footprint is still your LTE footprint. Specify 5G where the application needs the bandwidth, and do not let it displace LTE-M or LTE fallback in the module, because the fallback layer is what keeps devices online where 5G has not been built. For IoT coverage in remote areas, the older layer is doing the work.
Fix 4: Satellite and NTN for Genuine No-Coverage Areas
Non-terrestrial networks (NTN) extend connectivity to devices using satellites instead of ground-based cells, either through 3GPP NTN standards that let a suitably capable cellular module talk to a satellite, or through dedicated satellite IoT services. This is the fix for cause one, genuine absence of carrier presence, and it is the only fix for it.
The category has moved from niche to mainstream faster than most fleet buyers realize. By April 2026 there were 275 publicly announced operator and satellite partnerships in 101 countries and territories, according to GSA’s tracking of operator and satellite partnerships. That matters commercially, because it means satellite reach is increasingly something your terrestrial provider can arrange rather than a separate contract, separate device and separate platform.
When NTN is worth it: assets that genuinely operate beyond the cellular map for meaningful periods, where a gap in reporting has real cost. Mining and forestry equipment, long-haul routes through sparsely populated corridors, offshore and coastal assets, pipeline and environmental monitoring, and cargo where a six-hour silence is a compliance problem rather than an inconvenience. In those cases satellite is not an upgrade to remote IoT connectivity, it is the only version of it that exists.
When it is not: assets that lose signal for twenty minutes in a valley and reconnect. Store-and-forward on the device solves that at a fraction of the cost. Buffer the readings locally, timestamp them, and upload the backlog when the link returns. Most “coverage problems” brought to us as satellite questions turn out to be buffering questions.
We came to that view honestly. When our team mapped which verticals to prioritize, agriculture was the one we flagged as structurally harder, precisely because those buyers are often in the middle of nowhere and a proposition with a satellite element tells a stronger story there than a purely terrestrial one. That is worth saying out loud to any buyer scoping remote IoT connectivity: be honest about whether your assets live beyond the cellular edge or merely visit it, because the two need different architectures and very different budgets.
Fix 5: Hardware, Antennas, Placement and Enclosure
The cheapest coverage improvement available is usually physical, and it is the one most often skipped.
- Antenna choice and gain. An internal PCB antenna is fine for a device that lives in open air and poor for a device inside metal. Moving to an external antenna with a proper ground plane routinely recovers more link budget than any tariff or carrier change.
- Placement. Mount high, mount outside the metal, and keep the antenna away from parallel metal surfaces and other radios. On trailers, roof or upper-bulkhead placement beats anything mounted low behind steel.
- Cable and connectors. Every meter of thin coax and every extra connector costs signal. Short runs and the right cable type matter more at the margins, which is exactly where remote deployments operate.
- Enclosure and sealing. Sealed metal enclosures are an antenna’s worst environment. If the enclosure has to be metal for durability, the antenna has to be outside it.
- SIM durability. In vibration, dust, humidity and temperature extremes, a consumer-grade SIM is a failure point that presents as a coverage fault. Purpose-built industrial-grade SIM hardware rated from -40C to 105C is a sensible default for remote IoT connectivity in refrigerated, off-road or outdoor deployments.
What Factors Affect IoT SIM Performance in Warehouses
Warehouse dead zones are their own problem, and almost nobody writes about them. The causes are specific: steel racking creates reflection and multipath rather than simple attenuation; refrigerated and freezer rooms are effectively Faraday cages; below-grade docks and loading bays sit below the surrounding radio horizon; and dense inventory changes the RF environment week to week as stock levels move.
Diagnose with numbers, not bars. RSRP (Reference Signal Received Power) tells you how much signal reaches the device; RSRQ (Reference Signal Received Quality) tells you how usable it is once interference is accounted for. An RSRP that looks acceptable with a poor RSRQ means interference or congestion, not distance, and the fix is different. Walk the building with a logging device and record both at rack ends, freezer thresholds and dock doors, not just in the office. Indoor IoT coverage gaps are measured, never estimated.
The practical fixes, in order of cost: relocate the antenna to the exterior wall or roof; switch handhelds and yard assets to LTE-M or NB-IoT where the deeper indoor link budget helps; add a permitted signal repeater or a small cell where the carrier allows it; or accept the dead zone and design the device to buffer and forward at the door.
How to Audit a Provider’s Coverage Claim Before You Sign
Country counts are marketing. What follows is what to ask for instead, and it is the same list we ask our own suppliers.
- Ask for the per-country carrier list, not the country count. “180 countries” tells you nothing about whether you can attach to the operator that covers your route. Ask which named operators you are permitted to use in each country you actually run in.
- Ask which technologies are enabled per carrier. LTE-M and NB-IoT availability varies by operator within a single country. A yes at the country level can be a no at the carrier level.
- Ask how the provider handles remote IoT connectivity specifically. Ask for a named customer running assets in a comparable environment, and what changed for them.
- Ask about permanent roaming exposure. Which of your markets restrict permanent roaming, and what is the provider’s plan when a partner enforces the restriction on your SIM estate?
- Ask how steering is configured and whether you can change it. If the answer is vague, assume commercial steering and expect rural surprises.
- Ask for eUICC profile availability by country. Not the catalog size. The specific profiles they can push to your devices in your markets.
- Ask to test before you commit. A pre-activation allowance that lets you put real devices on real routes with a small amount of free data before billing starts will teach you more in two weeks than any coverage map.
- Ask who answers when a device drops. A named account manager who is also your service manager will resolve a coverage incident faster than a ticket queue, and that difference compounds when the asset is remote.
That last point about specificity cuts both ways. The best inbound requests we receive read like engineering documents: one recent inquiry simply stated 30 SIM cards on a private APN, OpenVPN between two to three of them for a secure private connection, and an average of under 300 megabytes per SIM. A private APN, for clarity, is a dedicated closed network that isolates your devices from public internet traffic and lets you control IP addressing and access policy. A buyer who can describe their deployment in those terms gets a useful answer immediately; a buyer who asks only for a price gets a price. If you are trying to fix IoT connectivity coverage in remote areas, specify like the first buyer.
Worked Scenarios With Recommended Configurations
A North American Fleet With Cross-Border Runs Into Mexico
The coverage failure here is rarely in the US. It is at and past the border, where the roaming relationship changes and steering decides whether your tracker attaches to a carrier with real coverage on the corridor you use.
Recommended configuration: a multi-network SIM with named, permitted carriers in the United States, Canada and Mexico and steering weighted to network quality rather than partner cost. LTE-M as the primary radio technology with LTE fallback, because the assets are moving. eUICC hardware even if you do not use profile switching on day one, so a Mexico-optimized operating profile can be pushed later without a service visit. Industrial-grade SIMs, because desert heat and vibration are the operating environment. Confirm in writing which Mexican operators you are permitted to use, and test on your actual routes rather than the highways in the coverage map. Cross-border runs are where IoT coverage in remote areas is won or lost.
A Mid-Sized Regional Carrier Running About 150 Long-Haul Trucks
At this size the economics are different from a 10,000-unit rollout. You have too many vehicles to chase failures manually and too few to justify custom engineering, and every truck-roll to re-SIM a unit is a real cost against a thin margin.
Recommended configuration: multi-network SIMs on a pooled data plan, so the trucks that sit idle subsidize the ones that upload heavily and you avoid overage entirely. eUICC form factor for future carrier flexibility. A single platform view with data-limit alerts, so a device stuck retrying a network shows up as an anomaly before it becomes a bill. LTE-M primary with LTE fallback. Then pick your three worst-performing routes and instrument two trucks each with signal logging for a month before you change anything else, because at 150 vehicles the pattern of poor remote IoT connectivity is almost always concentrated in a handful of corridors rather than spread evenly. This is the profile most fleet management deployments at this scale converge on, and it keeps IoT coverage in remote areas improving without a hardware refresh.
Temperature-Monitored High-Value Pharma Cargo
Here a coverage gap is a compliance event. The cargo does not care that the truck was in a valley; the audit trail has a hole in it.
Recommended configuration: treat store-and-forward as mandatory, not optional. The device logs temperature on its own schedule regardless of connectivity and uploads a timestamped backlog when the link returns, so the chain of custody is complete even when the reporting is not continuous. Layer a multi-network SIM with quality-weighted steering on top for live alerting. Add a private APN so the data never touches the public internet, which simplifies the security conversation with a pharmaceutical client considerably. Specify industrial SIMs rated for refrigerated environments. And for lanes with known long dark stretches, evaluate NTN as an alert-only channel: you do not need satellite bandwidth for the whole telemetry stream, only for an excursion alarm that cannot wait six hours.
How the Major Multi-Network Providers Compare on Remote Coverage
Buyers in this market usually shortlist from the same set: Onomondo, emnify, KORE, Wireless Logic, 1NCE, Hologram, Eseye, Pelion, Telnyx, Transatel and Soracom, alongside providers like ours. The honest framing, and one we have said internally more than once, is that most companies in this space do the same thing and try very hard to make out they have something unique, and most of them do not. The core capability is broadly comparable. The differences that matter for IoT connectivity coverage in remote areas are narrower and more specific:
- Architecture. Some providers route traffic through a single global core with in-country breakout; others deliver a foreign roaming identity; others lead with eUICC profile management. Each has different exposure to permanent-roaming rules.
- Carrier depth per country. How many named operators you can actually use in the countries you run in, which is a very different number from the countries listed.
- Steering control. Whether you can influence network selection or only observe it.
- Support model. Whether a dropped device gets a named engineer or a ticket number.
- Commercial flexibility. Contract length, pooled versus per-SIM data, and whether platform access and activation are billed separately.
Two things we have learned from being on the receiving end of these evaluations. First, a shortlist is often won or lost on functionality details rather than price: we lost a promising scooter-rental deployment in New Zealand, which would have grown to thousands of units, on technical fit rather than commercial terms. Second, bigger does not automatically mean better coverage on your route. Our largest customer, a video telematics provider, deliberately uses a native single-network IoT SIM rather than a multi-network one, because in their specific North American footprint the direct connection gives lower latency and more predictable behavior than switching ever would. That is unusual, and it is a useful reminder that the right answer to remote IoT connectivity depends on where your devices actually are.
Frequently Asked Questions
Does a multi-network SIM guarantee coverage everywhere?
No. It guarantees access to the carriers your provider holds agreements with in that country, which is not the same as every carrier present. Ask for the per-country carrier list before you assume a route is covered.
Can I switch carriers on a device already in the field?
With an eUICC SIM and SGP.32 support, yes, over the air, provided the destination profile is available in that market. With a standard SIM, no, and the device needs a physical visit.
Is NB-IoT or LTE-M better for moving assets?
LTE-M. It supports handover between cells and higher throughput, which a moving vehicle needs. NB-IoT suits stationary assets such as meters, tank sensors and fixed cold-store probes.
What is the difference between multi-IMSI and global roaming?
Global roaming gives your device one identity that visits partner networks. Multi-IMSI gives the SIM several network identities and switches between them so it can behave like a local subscriber, which matters in markets that restrict permanent roaming.
How do I fix a warehouse dead zone without changing provider?
Measure RSRP and RSRQ at the dead spot, move the antenna outside the metal, consider LTE-M or NB-IoT for better indoor link budget, and design the device to buffer and forward if the zone cannot be fixed economically.
Do I need satellite for rural coverage?
Usually not. If assets lose signal briefly and reconnect, store-and-forward is far cheaper. Satellite and NTN are for assets that genuinely operate beyond the cellular map for extended periods.
Key Takeaways
- Diagnose before you buy. The five causes of poor IoT coverage in remote areas look identical in a dashboard and have completely different fixes.
- Multi-network first, then multi-IMSI. Roaming solves reach; multi-IMSI solves regulatory standing and local rates in restricted markets.
- Specify eUICC even if you do not use it yet. It converts a future truck roll into a profile push.
- Match the radio to the movement. LTE-M for anything mobile, NB-IoT for stationary assets, LPWAN inside sites you control.
- Treat NTN as the fix for genuine absence of coverage, and store-and-forward as the fix for temporary gaps.
- Fix the antenna before you fix the contract. Placement, cable and enclosure often cost more signal than any tariff.
- Audit on carrier lists, not country counts, and test on your real routes with real devices before you commit.
Improving coverage in hard places is rarely one decision. It is a sequence: find the cause, change the SIM behavior, change the radio if you must, change the hardware where it is throwing signal away, and reach for satellite only where the map genuinely ends. Work through it in that order and most remote coverage problems resolve without the expensive answer.
If you want a second opinion on your own deployment, Trafalgar Wireless supplies multi-network IoT SIM and eSIM data plans across 500+ networks in 180+ countries to fleet, logistics and asset-tracking operators who need better IoT connectivity coverage in remote areas, and we are happy to walk through the per-country carrier detail for your routes before you change anything.
