Most cellular network operators around the world still have a circuit switched core network today for voice and SMS services for their 2G/3G access networks. There are a few exceptions like one operator in India that launched with an LTE-only VoLTE network a few years ago. But apart from a few, that is the status quo. And it’s unlikely to change for most operators for the following reasons:
- They still want to serve their own subscribers with non-VoLTE capable devices.
- They still have a lot of GSM based M2M business that requires a circuit switched core network.
- They want to make business with incoming roaming customers. VoLTE roaming has been standardized many years ago but so far, only very few network operators and devices support it today.
Most network operators thus keep running their circuit switched core network. It’s virtualized by now, so there are no ‘extra’ trunks these days anymore for GSM or UMTS, it all runs over the same infrastructure as LTE and 5G. But someone has to keep the aging switching center software and hardware nodes running.
As there is still money to be earned with 2G/3G, I was surprised for a moment that Verizon in the US is now close to switching-off their legacy network (see here and here). But for them it makes sense because they don’t make money with any of the three use cases above:
- The US in general and Verizon in particular has a very closed cellular ecosystem so there are likely only few customers left without VoLTE capable devices.
- Their 2G/3G legacy access network is not based on GSM but on CDMA. As a consequence they have very few roaming customers.
- For the same reason I would assume that Verizon has very few if any M2M customers on their 2G/3G CDMA network.
So, in other words, they are one of the few network operators which will not be hurt by a 2G/3G switch-off. Quite the contrary. Not needing to support a virtual circuit switched infrastructure will save them quite a bit. But it will be the end of an era for Verizon. After all, circuit switched systems formed the global telecommunication backbone for nearly a century. Packet switching, after a tough uphill battle starting in the 1960’s and 1970 finally took over voice telephony to a large extent only in the last decade. A 50 year struggle if you like.
From my point of view I don’t see 2G going away as a general trend in mobile networks yet. That is, not until VoLTE-roaming has become more popular and M2M business goes to LTE NB-IoT or M1. (Virtual) circuit switching might become marginal, but it will remain in place. Shockingly, it seems most network operators have no programs in place whatsoever to encourage a change.