Short answer
LPWA means low power, wide area: a radio class that deliberately gives up data rate to gain range and battery life. Cat NB1 peaks at 26 kbit/s down, so it suits a meter or a tank sensor sending small readings, and it is the wrong choice for anything carrying images, video or firmware payloads.
Key takeaways
- LPWA is a trade, not an upgrade: reach and battery are bought by giving up throughput.
- Both LTE-M and NB-IoT were specified in 3GPP Release 13, so the category is licensed cellular, not a proprietary radio.
- Cat NB1 peaks at 26 kbit/s down and 66 kbit/s up, enough for a reading, not for a photo.
- NB-IoT occupies 180 kHz, one LTE resource block; the 200 kHz figure quoted everywhere is the GSM carrier raster.
- The trackers we list are Cat 1 and GSM/3G, not LPWA, a bandwidth-hungry device belongs on standard cellular.
01The four letters, and what each one is buying
LPWA stands for low power, wide area, and the phrase is best read as a purchase order rather than a description. Two of the words name what you are buying: low power, meaning a device that can run for years on a cell instead of months, and wide area, meaning a signal that survives a basement, a steel cabinet or a distance no ordinary handset would tolerate. What the phrase hides is the currency. You pay in throughput. A radio that reaches further and sleeps longer does so by sending less, more slowly, in narrower channels, and by tolerating delays that a phone call could never accept. So the useful question is never whether LPWA is better. It is whether the thing you are connecting produces small readings on a schedule, or large payloads on demand, because the category is excellent at the first and genuinely unsuitable for the second.
02It is licensed cellular, not a private radio
A common misreading treats LPWA as something outside the mobile world, a separate technology bought from a specialist. For the two members that matter to an industrial buyer, that is wrong. The GSMA deployment guides record that both LTE-M and NB-IoT were specified in 3GPP Release 13 and continued in Releases 14 and 15, which means they are standards-track cellular running in licensed spectrum on the same base stations as the rest of the network. The practical consequences are large. You buy a subscription from a mobile operator, not a gateway. Coverage is the operator's coverage, extended by features rather than by your own infrastructure. And critically, the operator decides which of the two it has launched, so the technology you specify narrows the list of carriers who can serve you before the first SIM is ordered.
03The numbers that define the ceiling
Two channel widths tell most of the story. The GSMA LTE-M guide puts a Cat-M1 device at 1.4 MHz of bandwidth with a 1 Mbps peak in each direction, which is a narrow slice of an LTE carrier but still a usable pipe. NB-IoT is far narrower: its occupied bandwidth is 180 kHz, exactly one LTE physical resource block. That figure is worth stating carefully, because the number quoted almost everywhere is 200 kHz, which is the GSM carrier raster NB-IoT can replace, not the bandwidth it actually occupies. At the device end, the published peak rates for Cat NB1 are 26 kbit/s down and 66 kbit/s up in multi-tone, half duplex. That is the ceiling. Any requirement above it is not a tuning problem; it is a different radio class.
04How reach is actually bought
Range in these systems is expressed as maximum coupling loss, how much signal a link can lose and still work. Qualcomm publishes NB-IoT at 164 dB and eMTC above 155.7 dB, a gap of roughly eight decibels, while an Ericsson analysis states both can operate at 164 dB, matching the 5G coverage requirement. Those figures are not free. The mechanism behind them is repetition: the same message sent again and again until the receiver can assemble it out of noise. The LTE-M guide caps this at 32 repetitions in CE Mode A and 2,048 in CE Mode B. Read that as a cost line. Each repetition consumes airtime the cell could give to somebody else, and battery the device will not get back, so a device sitting deep in a coverage hole pays for its reach twice over.
05The assets that genuinely fit
The shape of a good LPWA workload is easy to recognise: a small reading, produced on a predictable schedule, from something that cannot be conveniently powered or visited. A tank level sensor at a remote depot fits. A utility meter in a basement fits, because the coupling-loss budget is precisely what gets a signal out of it. A pressure gauge on a pipeline, a status flag on a static generator, a temperature probe in a cold store, all of them produce a handful of bytes and want a decade of battery. What they share is that nobody needs the reading in the next second, and nobody will ever ask the device for an image. That patience is the whole reason the radio can sleep, and sleeping is where the multi-year battery figures come from.
06Where LPWA is the wrong answer
Reverse every property above and you have the exclusion list. Anything carrying an image or a video stream is out by two orders of magnitude, not by a margin. Anything needing a command to land promptly is out too: reported latency for NB-IoT runs from 1.6 to 10 seconds against 10 to 15 milliseconds for Cat-M1, and a remote unlock that might take ten seconds is a different product experience. A device pulling large firmware updates over the air is out, because the download that takes minutes on Cat 1 takes a great deal longer through a 180 kHz channel. Our own catalogue reflects this honestly rather than fashionably: the FMC003 listing is 4G LTE Cat 1 with 3G and 2G fallback, and the JT701 listing is GSM and 3G. Neither is LPWA, because neither workload wants it.
07How to decide, in the order that matters
Work from the payload outward, because the payload is the only constraint you cannot negotiate. First write down what one report contains and how often it is sent; if the answer exceeds a few hundred bytes at intervals of minutes, stop and specify standard cellular. Second, ask whether anything needs to reach the device promptly, and whether a delay of seconds is acceptable. Third, ask how the device is powered, an externally powered unit gains far less from LPWA than a battery one, because the sleep behaviour is the main prize. Only then look at operators, and expect the choice to narrow, since a carrier that never launched your technology cannot serve you regardless of its coverage map. Deciding in that order stops a radio class from being selected before the requirement is written.
| Workload | Payload shape | Suitable radio class |
|---|---|---|
| Tank level at a remote depot | Tens of bytes, scheduled | LPWA |
| Utility meter in a basement | Tens of bytes, scheduled | LPWA |
| Static generator status flag | Small, scheduled | LPWA |
| Vehicle tracker with driver events | Continuous small records | Standard cellular (Cat 1) |
| Remote unlock command | Small, but must land promptly | Standard cellular |
| Photo verification of a seal | Images | Standard cellular |
| Large over-the-air firmware update | Megabytes on demand | Standard cellular |
NB-IoT peak data rates by device category
3GPP device category specification table (Narrowband IoT)
Figures and sources
- 3GPP release in which both LPWA members were specified: Release 13, continued in 14 and 15 (both LTE-M and NB-IoT) (Source: GSMA LTE-M and NB-IoT Deployment Guides to Basic Feature Set Requirements, v3 (June 2019), as of 2026-08-31)
- Cat-M1 device bandwidth and peak rate: 1.4 MHz bandwidth; 1 Mbps peak in DL and UL (Source: GSMA LTE-M Deployment Guide v3 (June 2019), as of 2026-08-31)
- NB-IoT occupied bandwidth, the figure commonly misquoted as 200 kHz: 180 kHz (one LTE physical resource block) (Source: GSMA NB-IoT Deployment Guide v3 (June 2019), as of 2026-08-31)
- NB-IoT peak data rates by 3GPP device category, and its duplex mode: Cat NB1: DL 26 kbit/s, UL 66 kbit/s multi-tone, half duplex; Cat NB2: DL 127, UL 159 kbit/s (Source: 3GPP device category specification table (Narrowband IoT), as of 2026-08-31)
- Maximum coupling loss of each technology, the roughly 8 dB link-budget gap: NB-IoT 164 dB; eMTC above 155.7 dB (Source: Qualcomm, Leading the LTE IoT evolution, as of 2026-08-31)
- LTE-M coverage-enhancement repetition ceilings by mode: CE Mode A up to 32 repetitions; CE Mode B up to 2,048 (Source: GSMA LTE-M Deployment Guide v3 (June 2019), as of 2026-08-31)
- Reported latency contrast between the two LPWA technologies: Cat-M1 10–15 ms; NB-IoT 1.6–10 s (Source: IoT For All, What Is the Difference Between Cat-M1 and NB-IoT?, as of 2026-08-31)
- Radio class of the two trackers in our own catalogue, neither is LPWA: FMC003: 4G LTE Cat 1 / 3G / 2G; JT701: GSM / 3G (Source: ENTEK.AI product catalogue, as of 2026-08-31)
Questions
Is LPWA a single technology?
No, it is a category. For an industrial buyer the two members that matter are LTE-M and NB-IoT, both specified in 3GPP Release 13 and continued in Releases 14 and 15. Because they are standards-track cellular in licensed spectrum, you buy them as an operator subscription rather than as your own network.
How much data can an LPWA device actually send?
For NB-IoT the published Cat NB1 peaks are 26 kbit/s down and 66 kbit/s up in multi-tone, half duplex, in an occupied bandwidth of 180 kHz. Cat-M1 is roomier at 1.4 MHz and a 1 Mbps peak. Both are ceilings, so a requirement above them means a different radio class, not a configuration change.
Why is NB-IoT bandwidth sometimes quoted as 200 kHz?
Because 200 kHz is the GSM carrier raster that NB-IoT can be deployed in place of, not the bandwidth it occupies. The GSMA NB-IoT deployment guide gives the occupied bandwidth as 180 kHz, which is exactly one LTE physical resource block. The distinction matters when you are reading a spectrum plan rather than a brochure.
Are ENTEK.AI trackers LPWA devices?
No, and deliberately so. The FMC003 listing is 4G LTE Cat 1 with 3G and 2G fallback, and the JT701 listing is GSM and 3G. Those workloads send continuous records and need commands to land promptly, which is exactly the profile LPWA is not built for.