Content Manager, Onomondo
LTE-M and NB-IoT are Low Power Wide Area Networks (LPWAN) developed for IoT. These relatively new forms of connectivity come with the benefits of lower power consumption, deep penetration, smaller form factors and, maybe most importantly, reduced costs.
The decision on which technology standard to use is not a simple one. When I searched the internet for a clear answer, I quickly became lost in acronyms and found many contradictions.
This article aims to make an unbiased and correct comparison of the two standards based on industry research and Onomondo’s experience. I’ll aim to cut through the complexity and offer an easy to follow NB-IoT and LTE-M guide that makes your cellular connectivity decision easier.
- A quick overview
- Coverage / Penetration
- Global deployment and roaming
- Power consumption
- Freedom to leave
A quick overview
Device Size vs Connectivity Solution
|Peak downlink speed||26 kbit/s||1 Mbit/s|
|Peak uplink speed||66 kbit/s||1 Mbit/s|
|Bandwidth||1.4 MHz||180 kHz|
|Power classes||20 / 23 dBm||20 / 23 dBm|
|Latency||1.6 – 10 s||10 – 15 ms|
In 2018, Sierra Wireless made a helpful graphic depicting which use cases NB-IoT and LTE-M were best suited for.
The above information is an incomplete but helpful starting point if you’re trying to decide whether NB-IoT or LTE-M is best suited to your IoT project.
With that quick overview in mind, let’s dive a bit deeper. Some further insights on attributes such as coverage/penetration, globality, power consumption, mobility, and freedom to leave will help your decision.
If you ask the internet, you’ll generally hear that NB-IoT supports devices in deeper places than LTE-M. But is this really the case? Recently, researchers and industry peers have shown this presumption is not clear cut.
How is penetration measured?
To understand penetration, here are a couple of terms you need to be familiar with first: decibels, receiver sensitivity, link budget, and maximum coupling loss.
We use decibels-milliWatt (dBm) to measure cellular signal strength. It’s logarithmic, and the typical range is -50 dBm to -110 dBm. The power in dBm comes from this equation:
P(dBm) = 10 · log10( P(W) / 1mW )
P(dBm) = Power expressed in dBm
P(W) = The absolute power measured in Watts
mW = milliWatts
log10 = Log base 10
Without getting further into the math, it’s good to know that every 3 dBm increase more-or-less doubles the signal strength, and roughly every 6 dBm increase doubles the coverage distance from the cell tower. You can read more on this topic here: How to measure signal strength - signalbooster.com.
Receiver sensitivity and link budget
A radio receiver’s sensitivity level indicates its capability to extract information from a weak signal and is denoted as the lowest signal level that can be useful. In this case, a lower sensitivity number means the receiver is better and has a higher sensitivity (-73 dB is better than -70 dB). Read more here: Signal strength on Wikipedia.
The link budget shows what happens to the signal between transmitter and receiver. It represents the change between the transmitted and received power at the receiver.
Maximum Coupling Loss (MCL)
Bringing this knowledge about dBm, link budget and receiver sensitivity together, 3GPP uses MCL to evaluate coverage. In theory, it can be defined as the maximum loss in the conducted power level that a system can tolerate and still be operational (defined by a minimum acceptable received power level). A higher MCL means a more robust link between transmitter and receiver.
What is the MCL for NB-IoT and LTE-M?
3GPP's website refers to the book Cellular Internet of Things (Second Edition) From Massive Deployments to Critical 5G Applications as a source for more information on Nb-IoT and LTE-M.
According to that book, the MCLs on 4G telecom technology are:
- LTE-M: 160.7 dB
- NB-IoT: 164 dB
On 5G telecom technology, the MCLs for both NB-IoT and LTE-M (with the requirement of supporting a connection density of 1,000,000 devices per km2) must be 164 dB.
So, according to those MCLs, NB-IoT and LTE-M have the same penetration depth on 5G. On 4G, LTE-M falls behind the NB-IoT by 3.3 dB (160.7 dB), meaning that NB-IoT has a better penetration depth than LTE-M on 4G telecom networks.
What does research into LTE-M coverage show?
However, prominent companies in the telecom industry have conducted simulations based on the assumptions 3GPP has made. They researched the coverage for LTE-M and its suitability for IoT applications that need deep coverage.
They found that LTE-M has a very similar penetration ability to other LPWA technologies.
“LTE-M supports a very similar if not better coverage enhancement compared to other LPWA technologies”
Gus Vos, Chief Engineer, Technology Standards, Sierra Wireless.
Why? It turns out the assumptions underlying 3GPP’s MCL calculations varied between technologies and thus was not a fair comparison.
These are the numbers used for calculating the MCL for LTE-M and NB-IoT in 3GPP standards:
|Transmit Power||46 dBm||23 dBM||43 dBm||23 dBm|
|Noise Figure||9 dB||5 dB||5 dB||3 dB|
3GPP 36.888 and 3GPP 45.820 reveal the numbers are based on different transmit power, noise figure, and target throughput assumptions.
They went deep on the MCL calculations using typical LTE-M assumptions for the receiver noise figure and transmit power and found the penetration for LTE-M to be identical or better than NB-IoT.
How about real-world tests?
Jussi Ratilainen, CTO at Loopshore, is one example of someone who didn't want to rely on what others said about the two standards.
He tested the standards using their Loopshore air quality sensing devices. The results surprised him "I assumed that Cat-NB would outperform Cat-M easily, I just wanted to know how much. But results hint that it might be the other way around." (NB Cat-NB = NB-IoT and Cat-M = LTE-M)
You can read his complete analysis on medium: NB-IoT vs LTE-M in practice.
The penetration conclusion
Both NB-IoT and LTE-M are well suited for deep penetration. Research shows that LTE-M is capable of attaining the same penetration depth as NB-IoT (but at the cost of extra power consumption? You can read more on this below).
However, it’s important to keep in mind that the parameters involved in tests and simulations around the penetration capabilities could vary a lot from device to device, cell tower to cell tower, etc.
As a final note, a disadvantage that NB-IoT has here compared to LTE-M is that NB-IoT is not available wholesale. Currently, your device will only have access to a subset of base stations in a country when using NB-IoT. LTE-M already has wholesale agreements available, meaning today it’s possible to attach to multiple LTE-M networks from one device, thus increasing the chance of finding a strong enough connection.
Global deployment and roaming
NB-IoT can be deployed on both 2G (GSM) and 4G (LTE) networks, while LTE-M is solely for 4G. However, LTE-M is already compatible with the existing LTE network, while NB-IoT uses DSSS modulation, which requires specific hardware. Both are planned to be available on 5G. These factors, plus some others, impact availability around the world.
Fortunately, GSMA has a handy resource called the Mobile IoT Deployment Map. In it, you can see the global deployment of NB-IoT and LTE-M technologies.
Operators typically deployed LTE-M first in countries that had LTE coverage already (e.g. the US). It’s relatively easier to upgrade an existing LTE tower to support LTE-M than to add NB-IoT support.
However, if LTE is not supported already, it is cheaper to put up new NB-IoT infrastructure.
Another factor to consider is that companies like Huawei are pretty invested in NB-IoT intellectual property (IP) and support its rollout. You can read more about this topic here: LTE-M vs NB-IoT: An Overview of Narrowband IoT.
Deployment is something you will need to consider if you are sending a device somewhere. And if that device is not staying in the same place forever or you’re not sure where in the world it will deploy, you’ll also need to consider the nuances of roaming with LTE-M and NB-IoT.
LTE-M has the advantage of having cellular roaming agreements already in place, while NB-IoT is far behind. Therefore roaming for NB-IoT is comparatively limited.
Global deployment and roaming conclusion
The decisions here depend on where your device will be situated and whether or not it will need roaming. If your device might need to roam, LTE-M is your natural choice.
If your device will be static in, e.g. China, then NB-IoT is currently the choice that makes sense.
Nevertheless, NB-IoT and LTE-M are still in their early days, and it’s currently helpful to have 2G or 4G fallback options, for example.
Traditional cellular standards are designed to function with voice and SMS, and so are always listening for incoming calls. But most IoT use cases don’t require that devices listen, which means traditional standards (2G, 3G, 4G) use more power than NB-IoT and LTE-M.
Long battery lifetimes are essential in many modern IoT applications. Think of use cases like smart waste management, asset management, and environmental sensors here.
NB-IoT and LTE-M represent a significant improvement in power consumption compared to other cellular standards. However, comparing the two IoT connectivity solutions is tricky due to the many factors that influence power consumption and battery lifetime.
NB-IoT and LTE-M have power-saving features
Both technologies support PSM (power saving mode) and eDRX (Extended Discontinuous Reception), which extend battery life. The video below has a good overview of these two features (NB eMTC is a subset of LTE-M).
However, in practice, it’s crucial that you thoroughly test these features on your device first. While they are great in theory, using them may not be straightforward. We discussed this in a recent webinar with LightBug CEO and Founder Chris Guest. LightBug experienced difficulties in getting these power-saving features to work reliably and instead created their own “sleep” logic for their tracking devices. You can read the webinar summary here: How to extract maximum value from global IoT.
Powered usage during transmission
The average transmission current of LTE-M is slightly higher than NB-IoT according to 3GPP 36.888 and 3GPP 45.820. But this potentially higher transmission current is often outweighed by the dramatically shorter transmission time that LTE-M has due to its higher uplink and downlink speeds. This is thanks to the generally higher bandwidth LTE-M operates at compared to NB-IoT.
|Transmit Power||46 dBm||23 dBM||43 dBm||23 dBm|
|Max speed||26 kbit/s||66 kbit/s||1 Mbit/s||1 Mbit/s|
However, factors such as coupling loss (coverage conditions) will significantly affect the uplink/downlink speeds. You could say that in good conditions, LTE-M will reap the rewards of faster data speeds. However, in bad conditions, NB-IoT may have faster data speeds. See the graph below for an illustration of one study on NB-IoT and LTE-M (eMTC) power consumption.
Source: Power Consumption Analysis of NB-IoT and eMTC in Challenging Smart City Environments - Joerke, Falkenberg, Wietfeld - 2018
Power consumption conclusion
Power consumption in NB-IoT and LTE-M is heavily dependant on your particular use case and the devices you use.
A straight comparison between the two technologies is complicated as many variables dictate power consumption (transmission frequency, transmission size, coupling loss).
When moving, your device will need to switch between base stations (aka cell towers) frequently. The image below is from a study by Fabien Sanglard and elegantly shows how his phone jumped between base stations while moving.
As you can see, there are a lot of handovers between base stations in a short journey. This handover process is about your device’s connection being seamlessly transferred from one base station to another.
NB-IoT isn’t capable of this kind of seamless handover. As a device moves out of cell tower range, it will increase the power (conductive Tx power) to stay connected. Once it becomes disconnected, it must register with the network again, causing a connectivity drop and increasing power consumption.
It’s simple here, use LTE-M for mobile devices.
Freedom to leave
Freedom to leave (aka freedom to operate) is not often discussed in our industry as it’s not commonly available. But it’s something that runs to the core of Onomondo (and is possible because of our network core), so it would be a shame not to look into the possibilities here for the two technologies as well.
In short, freedom to leave is about the ability to transfer a SIM to a new operator over the air (or load a new operator profile onto the SIM). If you don’t have freedom to leave and haven’t already got support for your desired operator on deployed SIMs, you’ll need to physically visit all of your devices to swap SIM cards to change operator. Visiting devices to switch SIMs is generally either very expensive or almost impossible.
Most will tell you this is not a problem with eUICC. But there is still an issue with vendor lock-in for eUICC thanks to eUICC platforms, meaning you don’t have total freedom to operate with your SIMs (you can read more about this issue here: What are the existing lock-ins with eUICC?).
Onomondo makes it possible to switch operators on regular UICC SIMs via over-the-air (OTA) updates. We can do this thanks to our network setup and our complete control over our SIMs’ IMSI, Ki and OPC keys, a vital part of switching operators.
NB-IoT and LTE-M SIM identity switching
Most operators don’t support SMS on NB-IoT. This means eUICC will not work in many networks, and Onomondo’s OTA operator updates are also not possible.
“Only some of the operators deploying NB-IoT will support SMS, thus no clear deployment recommendation can be provided at this time”
NB-IoT deployment guide, June 2019, GSMA
LTE-M supports OTA updates. Additionally, the higher bandwidths of LTE-M make the transmission of SIM profiles easier (or any software update for that matter).
Freedom to leave conclusion
If you are considering eUICC, then it’s best to go for LTE-M. eUICC most probably won’t work in your region(s) of interest on NB-IoT.
Additionally, NB-IoT is more of an operator lock-in risk as switching operators are generally not possible OTA.
Based on the above considerations, you can see that LTE-M and NB-IoT are both strong technology standards for IoT applications.
- Coverage/Penetration: LTE-M is probably as good as NB-IoT.
- Global deployment and roaming: Roaming is better with LTE-M. However, it’s important to check availability in your region(s) of interest.
- Power consumption: This is a hard one to call. It appears that LTE-M is better in good to average coverage conditions, and NB-IoT is better in poor coverage conditions.
- Mobility: LTE-M is suitable for static and mobile devices, while NB-IoT is only suitable for static.
- Freedom to leave: LTE-M supports freedom to leave and eUICC, NB-IoT generally doesn’t.
My opinion is that LTE-M is overall the more robust choice for IoT solutions. If you add that OTA firmware updates are easier with LTE-M and that it's voice ready, it’s the more flexible solution of the two and represents a solid, future-proofed connectivity choice.
I hope this article helps you on your IoT journey. Feel free to contact us via firstname.lastname@example.org if you have any questions regarding NB-IoT or LTE-M IoT deployments.