The global shutdown of 2G and 3G networks is often framed as a spectrum-efficiency story. Operators are retiring legacy technologies to reuse valuable spectrum for 4G and 5G, improve network capacity, reduce operational complexity, and lower the cost per bit.
All of that is true. But beneath the spectrum refarming discussion sits a much more critical dependency: voice.
For decades, mobile networks created a simple expectation. If your phone had a battery and could detect a mobile signal, there was usually a path to emergency help — even outside your home network and, in many markets, even without an active SIM.
The shutdown of 2G and 3G is starting to challenge that expectation. The solution is not a new radio technology or a headline feature. It is VoLTE.
Why does voice become complicated after 2G and 3G
2G and 3G networks were built with circuit-switched voice as a core service. Voice calls were part of the native architecture.
LTE is different. It is an all-IP network. There is no native circuit-switched voice domain in LTE. That means operators need VoLTE, which carries voice over IMS, to deliver voice services properly on 4G.
Without VoLTE, many devices rely on CSFB (circuit-switched fallback). In this model, the phone uses LTE for data but drops back to 2G or 3G when making a voice call. That works only as long as 2G or 3G is still available.
Once legacy networks are switched off, CSFB disappears with them. This makes VoLTE readiness a gating factor for 2G and 3G sunset. Shutting down legacy networks is not simply a switch-off exercise; it is about ensuring that voice works reliably on the replacement architecture.
The three pillars of VoLTE readiness
VoLTE depends on three main areas: the network, the device, and roaming.
First, the network must be ready. Operators need an IMS core, emergency calling support, proper quality of service, and mobility or continuity mechanisms. As 5G is introduced, operators also need to consider VoNR and EPS Fallback strategies.
Second, the device must support VoLTE and be correctly provisioned. This is an important point because many “4G phones” are not automatically ready for VoLTE in every network. Some devices support VoLTE technically, but require the operator’s IMS profile or carrier bundle to be enabled. Others may support normal VoLTE calls but still have limitations around emergency calling.
Third, roaming must work. This is often the most underestimated part of the VoLTE transition.
Why VoLTE roaming is not automatic
Circuit-switched roaming became mature over many years. In many cases, it simply worked because the interconnect models, roaming agreements, and operational processes were already well established.
VoLTE roaming is different. It has to be built and tested between operators. It requires technical interoperability, IMS/IPX readiness, commercial agreements, device compatibility, and proper emergency call handling in the visited country.
This is where the risk becomes real.
A traveler may arrive in a country where 2G and 3G have already been switched off. His phone may work perfectly at home using VoLTE. But if his home operator does not have a VoLTE roaming agreement with the visited network, and emergency calls over IMS are not properly supported for inbound roamers, there may be no fallback.
On the home network, that emergency call might previously have dropped to 3G. In the visited country, if 2G/3G is gone and IMS emergency handling is not working for that roamer, there may be nothing left to fall back to.
The Australia example
This is not only a theoretical risk.
In markets such as Australia, regulators found many handsets that could make normal VoLTE calls but still relied on 3G for emergency calls. These devices appeared to work under normal conditions, but they became a safety concern once 3G shutdown plans progressed.
As a result, the regulator required telcos to stop providing service to mobile phones identified as unable to access emergency services after the 3G shutdown.
The lesson is clear: normal VoLTE calling and emergency VoLTE calling are not always the same thing. A device can appear ready for the future while still depending on the legacy layer for the most important call it may ever make.
The real work behind 2G/3G sunset
The 2G and 3G sunset is not just about switching off legacy networks. It is about making sure VoLTE is ready end-to-end.
That means increasing VoLTE device penetration, ensuring IMS provisioning is enabled on older handsets, testing VoLTE roaming in real-world scenarios, and validating emergency calling for inbound roamers before the legacy network disappears.
Spectrum efficiency may be the business case, and Voice continuity is the responsibility. So Operators should not sunset the old layer until the new one can reliably carry the call that matters most.