Skip to content
Sign inGet Started

LTE-M or NB-IoT for Remote Assets in Saudi Arabia: The Radio Choice Picks Your Operator

By Saeed Baatiah · Published:

Short answer

In Saudi Arabia the decision is not a specification comparison. GSMA's commercial launch register lists Mobily under LTE-M and STC under NB-IoT, so the radio you choose determines which carrier can serve the asset. Add that NB-IoT has no connected-mode handover, and a moving asset settles the question before any datasheet does.

Key takeaways

  • GSMA's launch register lists Mobily as LTE-M and STC as NB-IoT, the two Saudi carriers deployed different, non-overlapping technologies.
  • A single-radio device therefore cannot be dual-sourced across Saudi operators, which outweighs the roughly 8 dB coverage difference.
  • 3GPP requires no connected-mode mobility measurement for NB-IoT, so a moving asset must drop and re-attach at each cell edge.
  • The Teltonika model letter encodes the radio: FMM is Cat M1 (LTE-M), FMC is Cat 1, and RUT956 supports neither.
  • Zain KSA does not appear in the GSMA register; its documented NB-IoT activity is a 2017 Nokia trial, not a commercial launch.

01The Saudi operator split, and why it decides first

Most comparisons of these two technologies open with bandwidth and data rate. For an asset that will actually be deployed in the Kingdom, a different fact should be read first. GSMA maintains a register of commercial Mobile IoT launches, and in that register Mobily appears under LTE-M while STC appears under NB-IoT. The two largest Saudi carriers did not launch the same technology, and they did not launch both. That single line reorders the whole decision, because a device carries one radio: choosing LTE-M selects the operator that launched LTE-M, and choosing NB-IoT selects the other. There is no dual-sourcing a single-radio device across the two networks, and no switching carrier later without changing hardware in the field. Whatever the specification tables say about coverage or latency, this is the constraint that cannot be engineered around after the units are installed.

02What the register does and does not show about Zain

A third Saudi operator is usually assumed into these discussions, and the evidence does not support it. Zain KSA does not appear anywhere in the GSMA commercial launch register, not under LTE-M, not under NB-IoT. What is publicly documented is a trial: Nokia announced in October 2017 that it had successfully trialled NB-IoT with Zain Saudi Arabia in a live network. A trial in a live network is a real engineering milestone and it is not a commercial launch, and treating the two as equivalent is how a procurement plan acquires a carrier option that cannot be exercised. The honest statement is that two Saudi operators have documented commercial LPWA launches on different technologies, and the third has a documented trial from 2017 with no commercial launch listed. Anyone planning around Zain LPWA coverage should verify current commercial availability directly with the operator rather than infer it from the trial.

03The mobility difference is architectural, not a setting

The second decisive fact concerns movement, and it is not a tuning parameter that a good integrator can configure away. The 3GPP design for NB-IoT states that no connected-mode mobility measurement is required, and that a device only needs to perform mobility measurement during idle mode. That is the absence of handover as a design property: an NB-IoT device in an active session at a cell edge does not get moved to a better cell by the network, so it must lose the connection and re-attach. LTE-M is specified the other way. GSMA describes network-controlled handover in which the network decides when the device shall move, to which cell, and triggers the handover procedure, and recommends that connected-mode mobility be supported for all data traffic, not only voice. Qualcomm's own comparison table records the same split in three words for NB-IoT: cell reselection only.

04When that difference is irrelevant

The mobility gap is decisive for a moving asset and close to irrelevant for a fixed one, and conflating the two cases is how projects overspecify. A diesel tank at a remote pump station, a standby generator at a tower site, a water meter, or a static storage vessel does not cross cell boundaries, so it never enters the situation that handover exists to solve. For those assets the argument turns on link budget and power instead, and there NB-IoT is the stronger candidate: Qualcomm records a maximum coupling loss of 164 dB for NB-IoT against above 155.7 dB for eMTC, roughly 8 dB of additional link budget, which is exactly the margin that matters for a sensor inside a concrete enclosure or below grade. Ericsson, a co-author of the specifications, states that both technologies can operate at a coupling loss of 164 dB, matching the 5G coverage requirement, so the two are closer than the headline gap suggests once deployment configuration is considered.

05Bandwidth and rate, with the usual figure corrected

Both technologies were specified by 3GPP in Release 13 and continued in Releases 14 and 15, so neither is newer than the other in any meaningful sense. The low-complexity device category introduced in Release 13, Cat-M1, has a device bandwidth of 1.4 MHz and a peak rate of 1 Mbps in both downlink and uplink. NB-IoT is much narrower, and here the commonly quoted number is wrong in a way worth fixing: GSMA states that NB-IoT occupies a frequency band of 180 kHz, one LTE physical resource block, and the academic primer on the technology gives the same 180 kHz as the minimum system bandwidth for downlink and uplink. The familiar 200 kHz is the GSM carrier raster that an operator can replace with an NB-IoT carrier, not the occupied bandwidth. Peak rates follow from that width: Cat NB1 reaches 26 kbit/s downlink and 66 kbit/s uplink multi-tone, half duplex only, with Cat NB2 raising those to 127 and 159 kbit/s.

06Latency, and what a ten-second ceiling means

Latency separates the two technologies by orders of magnitude, and the useful way to read it is against what the application actually needs. Cat-M1 is reported at 10 to 15 milliseconds while NB-IoT ranges from 1.6 to 10 seconds. Ericsson states that both technologies meet the 5G connection-density requirement of one million devices per square kilometre with a service latency of at most 10 seconds, which is the clue to what NB-IoT was designed for. A ten-second service latency is entirely acceptable for a tank level reported every fifteen minutes and entirely unacceptable for a command a dispatcher expects to take effect now, such as releasing a lock or cutting a fuel pump. So the latency figure is not a quality ranking between the two radios; it is a statement about which class of application each was specified to serve, and an alert that needs acknowledgement within seconds belongs on the LTE-M side of that line.

07Power saving is available on both, with numbers

Battery life on a remote asset is usually the reason LPWA is considered at all, and the sleep mechanisms are not an NB-IoT exclusive. GSMA recommends that LTE-M deployments support idle-mode extended discontinuous reception and consider accepting the 3GPP Release 13 device-requested values, whose range runs from a minimum of 5.12 seconds to a maximum of 43.69 minutes. Power Saving Mode is defined with an extended T3412 timer whose maximum is 413 days under 3GPP TS 24.008, and GSMA recommends operators support a minimum of four hours. Those are large numbers, and they mean the practical battery question is rarely which radio sleeps deeper, it is whether the operator has enabled and provisioned the sleep parameters the device requests, since a device asking for a 4-hour timer on a network that does not grant it will not get that sleep regardless of its own capability.

08Coverage enhancement, and the cost of the deeper mode

LTE-M reaches difficult locations by repeating transmissions, and the specification defines two coverage-enhancement modes with very different ceilings: CE Mode A supports up to 32 repetitions and CE Mode B up to 2,048. That range is what allows one technology to serve both an easy outdoor site and a deeply shadowed one, but the ratio between the two ceilings is also a warning. Repetition buys link budget by spending airtime and energy on the same payload, so a device that has fallen into the deeper mode is transmitting far longer for each message than the same device on a clear link. The practical consequence is that a battery-life estimate calculated at CE Mode A does not describe a unit that ends up operating in CE Mode B, and a site survey that measures actual signal conditions is therefore not optional detail, it decides which of two very different power budgets applies.

09The model letter that ships the wrong radio

This is the failure that turns a correct technology decision into wrong hardware, and it is specific enough to check on a purchase order. In Teltonika's naming, the model letter encodes the radio class. FMM denotes Cat M1, so the wiki describes the FMM003 as an LTE Cat M1 advanced plug and play tracker. FMC denotes Cat 1, and the wiki describes the FMC003 as an LTE Cat1 tracker. The two part numbers differ by one letter and are not the same radio technology. Separately, the RUT956 industrial router is a 4G LTE Cat 4 device supporting 3G and 2G fallback, and it supports neither LTE-M nor NB-IoT, the Teltonika gateway for those is the TRB255. A project that specified LTE-M correctly and then ordered FMC parts has bought Cat 1 devices, which will attach to a normal LTE network and deliver none of the LPWA power or coverage behaviour the design assumed.

10Voice, and the asset that has a person on it

One capability separates the two technologies absolutely rather than by degree, and it is easy to overlook when the design brief says the word sensor. NB-IoT has no voice support at all. LTE-M can carry VoLTE, though GSMA is careful to note that VoLTE is not a mandatory recommendation for global LTE-M deployments and applies its guidance only where an operator does deploy it. For a tank level or a meter reading this is irrelevant. It becomes decisive the moment a person is attached to the asset: a driver who needs a duress call, a lone technician at a remote pump station, a service lift or an alarm panel with a speak-to-operator function. If any of those requirements exists now or is likely within the hardware's service life, the voice question removes NB-IoT from consideration before coverage or power is discussed, because no configuration adds a capability the technology does not define.

11A decision order that survives procurement

The sequence that follows from the evidence above inverts the usual one. Ask first whether the asset moves between cells during an active session, because that question is answered by 3GPP's design rather than by tuning and it eliminates one option outright for a moving asset. Ask second which operator actually serves the asset's location, and check the technology that operator has commercially launched rather than the technology the industry discusses. Ask third whether any command or alert needs to take effect within seconds, since a ten-second service latency is a specification, not a worst case. Only then compare link budget, and treat the roughly 8 dB difference as a tiebreaker for a fixed, deeply shadowed installation rather than as the headline. Verify the radio class on the part number before the order is placed, since one letter separates the right device from the wrong one.

12What ENTEK.AI does not claim here

Two things are deliberately absent. This article links no product, because the LTE-M parts in Teltonika's range are the FMM series and the TRB255 gateway, and our marketplace listing does not currently carry them, linking an FMC device here would contradict the section above that explains why FMC is the wrong radio class. We would rather publish the correct engineering conclusion and no product link than link a device that fits the article's topic only by name. Second, we publish no figure for LTE-M or NB-IoT coverage measured at specific Saudi sites. Coverage is local, changes as operators add capacity, and a number we measured at one location on one date would be read as a general claim. The GSMA register tells you which technology each operator launched; only the operator can tell you whether it reaches your site.

LTE-M versus NB-IoT on the properties that change a decision
PropertyLTE-M (Cat-M1)NB-IoT (Cat-NB1/NB2)Which asset it decides
Saudi operator (GSMA register)MobilySTCEvery asset, the radio picks the carrier
Connected-mode handoverNetwork-controlled handoverNot required by 3GPP; idle-mode onlyAnything that moves between cells
Occupied bandwidth1.4 MHz180 kHz (one LTE PRB)Spectrum planning, not the buyer
Peak rate1 Mbps DL and ULNB1: 26/66 kbit/s; NB2: 127/159 kbit/sPayload size and firmware updates
Latency10–15 ms1.6–10 sAny command needing effect within seconds
Maximum coupling lossabove 155.7 dB164 dBFixed, deeply shadowed installations
Voice (VoLTE)Deployable, not mandatoryNo voice supportAssets with a driver or duress call
Teltonika part prefixFMM (e.g. FMM003); TRB255 gatewayTRB255 gatewayThe purchase order itself

Peak data rate by device category

Cat-M1 downlink1,000 kbit/s
Cat-M1 uplink1,000 kbit/s
Cat NB1 downlink26 kbit/s
Cat NB1 uplink (multi-tone)66 kbit/s
Cat NB2 downlink127 kbit/s
Cat NB2 uplink159 kbit/s

GSMA LTE-M Deployment Guide v3 (1 Mbps) and the 3GPP device category table for NB-IoT

Figures and sources

Questions

Which is better for Saudi Arabia, LTE-M or NB-IoT?

Neither in general, because the choice selects an operator. GSMA's commercial launch register lists Mobily under LTE-M and STC under NB-IoT, so the right question is which carrier covers the asset's location. A moving asset then needs LTE-M regardless, since NB-IoT has no connected-mode handover.

Can I use NB-IoT on a truck or a moving container?

Not reliably during an active session. The 3GPP design requires no connected-mode mobility measurement for NB-IoT, so the device cannot be handed to a better cell mid-session and must drop the connection and re-attach at each cell edge. LTE-M supports network-controlled handover instead.

Is a Teltonika FMC920 or FMC003 an LTE-M device?

No. In Teltonika's naming the model letter encodes the radio: FMM is Cat M1, which is LTE-M, and FMC is Cat 1. The wiki describes the FMC003 as an LTE Cat1 tracker and the FMM003 as Cat M1. The RUT956 router is Cat 4 and supports neither LTE-M nor NB-IoT.

Is NB-IoT bandwidth 200 kHz?

The occupied bandwidth is 180 kHz, one LTE physical resource block, per GSMA and the academic primer on the technology. The 200 kHz figure is the GSM carrier raster an operator can replace with an NB-IoT carrier, which is why the two numbers are so often confused.

Does Zain KSA offer NB-IoT or LTE-M?

Zain KSA does not appear in the GSMA commercial launch register for either technology. The publicly documented activity is a 2017 Nokia trial of NB-IoT in a live network, which is not a commercial launch. Verify current availability with the operator rather than inferring it from the trial.

Sources

Related insights