I’ve been in a presentation on the LoRA PHY on the 29th of December at the 33C3 that I can fully recommend if you are interested in the LoRA IoT radio standard and want to learn about “chirp” modulation. During the talk it was mentioned that AT&T is shutting down its GSM network in the US by the end of 2016. A day later, quite a number of German tech websites such as Heise, Teltarif and Golem had articles on the topic and it really made me wonder if this talk was the seed for them!?
Be that as it may, it did not come as a surprise. Already announced in 2012, as I discussed at the time they make good on their announcement and it shows some interesting long term planning and foresight on their part. I had a look if I could find more information to confirm that it’s actually happening and found a number of sources that confirm that AT&T is really shutting down their GSM network now.
- AT&T’s own words on their website
- The competition inviting 2G only customers to switch to their network
- A field logistics company that has assembled a map of where GSM will be shut down in the weeks to come, i.e. very specific information it is happening.
- A company offering SIM cards and subscriptions to replace AT&T 2G subscriptions.
Over the years, AT&T has massively deployed 3G and LTE as I witnessed several times over the last few years. Even in rural farmland areas, I could receive an LTE network from at least one network operator. In other words, most telephony/Internet customers won’t see much of a difference.
However, the move is a problem for embedded devices that are still equipped with 2G-only modems. I’m sure they are still being sold to unwitting customers in the US even today. This, together with a still inferior 3G/LTE coverage is the main reason why only few network operators are willing to switch-off GSM at this time. There is still some money to be made with it and in countries where network operators are not as spectrum starved as operators in the US there’s no need to go into this direction anytime soon.
Like AT&T, other network operators are likely to give an advance notice of several years to give companies that use GSM-only M2M radios the chance to gracefully move to alternatives. I’m sure that like in the US, not everybody will get the message and prepare themselves.
Related to this topic I also found it quite interesting that in 2016, GERAN, the group working on GSM/GPRS enhancements in 3GPP was folded into the 3G/LTE RAN meeting structure. A logical move since there were only few topics discussed in the group lately. A sign of the times.
GSM is dead, long live GSM!
I have also seen the talk from 33C3 but only as stream at home and can recommend this talk.
Regarding AT&T they had a clear strategy and already some years ago there was no way to get a AT&T provider approval for 2G only applications. So even if some people (especially in the EU) didn’t believe that there would be a shut down of the 2G service it was communicated clear and not on short term.
Interesting is what will happen in the EU not only from the aspect M2M and big existing Systems like some Tolling Systems who are running 2G only. Also the eCall which is currently based on circuit switched tech is a topic to consider. Unfortunatelly the nextGeneration eCall was specified late in 3GPP Rel13 and is about to be in the Networks probably 2018. Maybe Martin has here a more accurate idea of the rollout of Rel13 in the Networks in EU.
eCall can also be done over 3G over circuit switched. But you are right it’s now defined for IMS as well. If the time frame for eCall in IMS is the same as in circuit switched it will just take another 10 years…