Will vodacom replace 2G with LTE-M

Thor

Thor

Honorary Master
Joined
Jun 5, 2014
Messages
44,413
Reaction score
7,522
Location
Bellville
I'm in the IoT space and I'm just wondering with the 2G- and 3G phaseout accelerating, is vodacom and other NOs busy implementing an alternative for us?

LPWANs on LTE-M will be such an incredible opportunity for IoT in South Africa.
 
EDIT: Mixing up LTE-M with NB-IoT. I am expecting NB-IoT to first roll-out.
Vodacom supports it:
1620342419742.png
Would not expect LTE-M that soon in South Africa tbh, first MTN needs to roll out NB-IoT.

South Africa has spectrum issues, NB-IoT is what will dominate here (especially as Huawei tech).
 
Last edited:
SA is going to have a huge problem trying to turn 2G off.

There are almost 100 miliion subscribers in SA but only 25% have smart phones. So 3 out of 4 devices are only using IoT, voice/sms or m2m. Even if every subscriber had 2 sim cards in their phones, half will still be non-data subscribers.

It will become another "tv digital migration" fiasco. And unlike tv tuners, it is not as simple as giving away free stb's. Millions of phone users are only using their phones for calls and sms's. There would have to be a effort made to stop the dumping of obsolete 2G devices in SA. The networks could start a campaign of buying cheap 3G devices at bulk prices and selling them at cost (or below cost, remember they want 2G gone) to people willing to bring in their active 2G device.

Not difficult to do. The network can see how a device connects. They could then send that users an sms to take in their 2G device and have it swapped out for a new device for a small charge. Linking the IMEI and sim number the provider could authenticate the offer.
 
Johnatan56 said:
I am expecting NB-IoT to first roll-out.
Click to expand...

I really hope we do not do that.

Why does SA always choose the bad projects?

Compared to Narrowband-IoT (NB-IoT), LTE-M’s data speeds are more than 10 times faster. Its latency is about 10 to 100 times lower, and it boasts far greater coverage, as it uses existing 4G LTE infrastructure.

LTE-M can also use a wider range of frequencies.

<code>
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| Technology | Maximum Uplink Speed | Maximum | Latency | Maximum Coupling | Bandwidth |
| | | Downlink Speed | | Loss (dBs) | |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| LTE-M | 1 Mbit | 1 Mbit | 10-15 ms | 156 | 1.4 MHz to 5 MHz |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| NB-IoT | 127 Kbit | 159 Kbit | 1.6-10 s | 164 | 180 KHz |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
</code>

With lower latency and higher data speeds, LTE-M supports more use cases and helps future-proof devices by allowing vendors to update them remotely to add new capabilities and security features.
 
Dairyfarmer said:
Not difficult to do. The network can see how a device connects. They could then send that users an sms to take in their 2G device and have it swapped out for a new device for a small charge. Linking the IMEI and sim number the provider could authenticate the offer.
Click to expand...

that is a brilliant suggestion tbh.

@rpm this should make a great article "Possible solutions to getting rid of 2G and what alternative should be offered"
 
Dairyfarmer said:
The networks could start a campaign of buying cheap 3G devices at bulk prices and selling them at cost (or below cost, remember they want 2G gone) to people willing to bring in their active 2G device.
Click to expand...
It would be a lot better for them to do that for LTE rather.
And don't think it will really work as it's more a case of it will be handed down as well, so you got rid of one 2G user and it will be passed to a kid/family member and 2G will continue.

Thor said:
I really hope we do not do that.

Why does SA always choose the bad projects?

Compared to Narrowband-IoT (NB-IoT), LTE-M’s data speeds are more than 10 times faster. Its latency is about 10 to 100 times lower, and it boasts far greater coverage, as it uses existing 4G LTE infrastructure.

LTE-M can also use a wider range of frequencies.

<code>
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| Technology | Maximum Uplink Speed | Maximum | Latency | Maximum Coupling | Bandwidth |
| | | Downlink Speed | | Loss (dBs) | |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| LTE-M | 1 Mbit | 1 Mbit | 10-15 ms | 156 | 1.4 MHz to 5 MHz |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
| NB-IoT | 127 Kbit | 159 Kbit | 1.6-10 s | 164 | 180 KHz |
+-------------+-----------------------+-----------------+-----------+------------------+-------------------+
</code>

With lower latency and higher data speeds, LTE-M supports more use cases and helps future-proof devices by allowing vendors to update them remotely to add new capabilities and security features.
Click to expand...
That table is both right and wrong, real-world NB-IoT latency is more on the order of 100-300ms, remember NB-IoT is focused on low power consumption (5µA in sleep, 120-300mA in peak) which should allow up to 10y battery life, range (up to about 35km) and high density (55k devices per cell / 40 per household average). That 10s of latency is the max latency it can handle rather than what will be the latency, real-world was more ~300ms, jumping up when coupling loss increased (so worse signal and probably worse than LTE-M can handle).

Note again, NB-IoT will have more range and be able to handle penetration through objects better, think smart meters in a basement.

LTE-M will consume more power than NB-IoT usually, plus NB-IoT uses less frequency. For an IoT device, senidng more than a few kbps would seem quite strange tbh, if you're sending temperature for example it will be polling a kb every second/few seconds max, electricity meter every 15 minutes usually (at least that's what it is here), heart rate can be up to 220 times a minute if you're going to send something with every heart beat, but that's still probably going to be under the limit, water measurement at a dam would also only be once every few seconds/minutes/hours, location data would be every second/few seconds etc.

So not sure why you would need such a high data rate (maybe real-world home automation for voice match to server to get result quickly? Most of the time you'll still be fine).
 
I was under the opposite impression RE the power consumption, I thought NB-IoT uses more power in some use cases. When your devices need to transmit larger volumes of data, NB-IoT’s lower data speeds require your devices to stay online longer, resulting in greater power consumption. In mobile use cases, NB-IoT uses more power as well.

It's about future-proofing, NB is fine for now, won't be tomorrow, but LTE-M will be, so do it right from the start.

PS: I also wonder why SA isn't that big on Lora, the networks could have easily created their own gateways and charge micro usage on it for edge nodes.
 
Johnatan56 said:
And don't think it will really work as it's more a case of it will be handed down as well, so you got rid of one 2G user and it will be passed to a kid/family member and 2G will continue.
Click to expand...
No the person has to hand in the working 2G device to get a new 3G device. The network will have the IMEI and sim number so there shouldn't be (SA mind you) any reason the device gets back on the network. Also the networks can blacklist the IMEI number once the person has got a new device.
 
Thor said:
I was under the opposite impression RE the power consumption, I thought NB-IoT uses more power in some use cases. When your devices need to transmit larger volumes of data, NB-IoT’s lower data speeds require your devices to stay online longer, resulting in greater power consumption. In mobile use cases, NB-IoT uses more power as well.

It's about future-proofing, NB is fine for now, won't be tomorrow, but LTE-M will be, so do it right from the start.

PS: I also wonder why SA isn't that big on Lora, the networks could have easily created their own gateways and charge micro usage on it for edge nodes.
Click to expand...
Why would that be better future proofing? And again, most IoT will be minute data amounts, I understand your argument that you can race to sleep faster with LTE-M, but most of the time the difference will be negligible for most devices and for the South African context I think the lower spectrum usage and much larger coverage area of NB-IoT will be better (thinking rural).

Should actually be a case of have both, but imho NB-IoT is the generally more important of the two for right now.
Dairyfarmer said:
No the person has to hand in the working 2G device to get a new 3G device. The network will have the IMEI and sim number so there shouldn't be (SA mind you) any reason the device gets back on the network. Also the networks can blacklist the IMEI number once the person has got a new device.
Click to expand...
Cool, just quite a large cost for probably low returns for the providers, think it would still be better for them to just refarm 3G and offer cheap LTE devices as a possibility. The cost of LTE devices has reduced substantially over the last few years and is possible to get on <R65pm contracts. Past that point it's not really viable for the network providers to subsidize if they give the phone to someone who will never use it, they just need to force a sunset date in like 5 years or something.
 
Johnatan56 said:
Should actually be a case of have both, but imho NB-IoT is the generally more important of the two for right now.
Click to expand...
I get what you are saying, our rural landscape forces us to rather take the longer-range option, that and RF frequencies because the spectrum is locked.
 
Johnatan56 said:
Cool, just quite a large cost for probably low returns for the providers, think it would still be better for them to just refarm 3G and offer cheap LTE devices as a possibility.
Click to expand...
You can buy a 3G phone for under R250. Wholesale is probable about R150. The networks could probably take a R50 knock on this and still come out smiling. They would, after all, be getting them off the network. It's not all those 50 million subscribers that have 2G handsets. And if they stopped selling 2G phones, those that die (and those cheap phones don't last more than a few years and the batteries even less), it wouldn't take long.
 
Dairyfarmer said:
You can buy a 3G phone for under R250. Wholesale is probable about R150. The networks could probably take a R50 knock on this and still come out smiling. They would, after all, be getting them off the network. It's not all those 50 million subscribers that have 2G handsets. And if they stopped selling 2G phones, those that die (and those cheap phones don't last more than a few years and the batteries even less), it wouldn't take long.
Click to expand...
You know Vodacom already tries and is pushing them onto 4G instead, e.g. https://www.takealot.com/nokia-215-4g-feature-phone-network-locked-vodacom-power-pack/PLID71836062 at R300, also offered at PEP stores: https://www.pepstores.com/product/cell/vodacom/nokia-215-sku-dhw90


What's funny is Vodacom itself is perpetuating the non-4G by basically offering these non-4G phones. They'd need to price match it to R179 and it would probably be the new phone.
1620405532585.png
But I am pretty sure you'd have a pretty quick call from the competition commission.
 
Im pretty sure its not 2g handsets that are the problem. Its 2G credit card machines and trackers.

And the reason for 2G is it is global banding and modules are dirt cheap ($2 rather than $10 for 3G) So you make one device and it works everywhere with 2G. 3G has different bands depending where you go, same with 4G etc. Then again 2G has ended in USA and Aus, so kills 2G just a little.

They should switch off 3G like they have in Europe.
And... If you're really doing low power you should be using LoRa rather than gsm/mobile IMHO :)
 
The problem here is that moving from 2G to anything else involves equipment changes which can get expensive. Realistically the solution to the the problem with regards to IOT is LoRa. The question then becomes, who pays for it? If you have one or ten sensors in the wild, it won't hurt. Once you hit 100's it will leave a substantial dent on your pocket
 
Dairyfarmer said:
Most of rural SA would have no comms then. And VC would have to put Telkom on 4G roaming rather than degrading them to 3G
Click to expand...
... I mean Instead of switching off 2G, not as well as!
(As far as I understood it, Vodacom is already allowing Telkom to roam on 4g!?)
 
Genisys said:
The problem here is that moving from 2G to anything else involves equipment changes which can get expensive. Realistically the solution to the the problem with regards to IOT is LoRa. The question then becomes, who pays for it? If you have one or ten sensors in the wild, it won't hurt. Once you hit 100's it will leave a substantial dent on your pocket
Click to expand...
Yup. Its a big question / problem in general with this stuff. Customers dont realise technology expires... At some point that $50 temperature sensor you bought 8 years ago, might need to be replaced :)

Lora modules are pretty cheap nowadays, but finding good final products that will be well supported for years and years that arent very expensive is hard.
 
Last edited:
Dairyfarmer said:
No the person has to hand in the working 2G device to get a new 3G device.
Click to expand...

Though from a European point of view the networks appear to be aiming at getting rid of 3G first, the technology is the most wasteful in terms of spectrum (for example, cell breathing). 2G is likely to be retained on a small sliver of spectrum (likely 5MHz) with the rest of the spectrum being re-farmed to 4G/5G.

The networks are in a win-win with this, 2G is still there for alarms, meters etc and those Luddite holdouts with old phones can still make calls and send texts.
 
Dave said:
Though from a European point of view the networks appear to be aiming at getting rid of 3G first, the technology is the most wasteful in terms of spectrum (for example, cell breathing). 2G is likely to be retained on a small sliver of spectrum (likely 5MHz) with the rest of the spectrum being re-farmed to 4G/5G.

The networks are in a win-win with this, 2G is still there for alarms, meters etc and those Luddite holdouts with old phones can still make calls and send texts.
Click to expand...
It really is the best approach!

Unfortunately the US and Australasia decided against this approach and just did it! Otherwise we could still have global thin IoT 2G GSM devices. Now its just EMEA GSM region, and at least thats better than siding with LTE-M, nbiot (and different bands for each)
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X