Understanding network resilience and disaster recovery is essential particularly for operators providing connectivity in regions of seismic activity. Emblasoft Evolver can test how your network would cope in unexpected situations.

The news regularly highlights  the impact of natural – and unnatural (human created) – disasters. Almost all of them are unpredictable in terms of severity and location. One of the most important needs in coping with the aftermath, and providing aid, is the mobile network. It enables emergency services and workers to communicate with each other over relief plans and allows the impacted region to retain essential connectivity between family and friends.

Operators must meet international guidelines for disaster resilience and recovery

Of course, such incidents will have a significant impact on mobile network operations. MNOs are obliged to follow national and international guidelines for disaster planning, resilience, and recovery. So, it’s essential to test network protocol resilience and performance on return to service in multiple scenarios.

For example, in April 2025 a huge power outage significantly affected mobile communications in Portugal and Spain. It closed travel infrastructure, schools, and businesses.

Parts of France were also affected, as was Morocco – prompting Orange Maroc to post on its X feed: “Internet traffic has been disrupted following a massive power outage in Spain and Portugal, which is affecting international connections.”

Most importantly, in some areas, access to emergency services via 112 was disrupted. 112 is a mandated service and operators need to ensure that they can restore access as quickly as possible.

Natural disasters can similarly severely impact mobile connectivity, as highlighted by the recent earthquakes in Venezuela, which occurred minutes apart. Many countries sit on seismic fault lines and shifting tectonic plates, and are therefore prone to earthquakes, which can deliver serious and often lasting, shocks. In these situations, operators need to be sure that they can keep their networks running with the necessary resilience and with effective recovery protocols.

That’s why some operators are modelling disaster scenarios, so they can explore how their network would respond to such an event – and how it would recover to restore services.

Emblasoft helped an operator in a region of seismic activity to test its disaster resilience

Emblasoft has worked with one leading operator in a country that suffers frequent earthquakes and seismic disruption. The operator is pre-empting how different events might impact network operations and disrupt services – and how to recover from that.

This operator has more than 200,000 base stations in its 5G network (gNodeBs). When an earthquake occurs, each base station in the affected area can enter a period of instability, with multiple restarts. Multiple electricity generators can be deployed, but they have limited duration, which may be insufficient for a major seismic event.

In situations like these, there are a number of things to consider. For example, instability in the RAN means that an individual UE may fail to attach to the network. At which point, it enters search mode to find an available base station instead of the one to which it was connected.

This search mode has a cost – and, if other UEs are affected in the same area, there may be many attempting to do the same thing. As a result, the volume of signalling activity increases significantly. In turn, if neighbouring cells are affected (which is likely), the volume of activity surges leads to a storm.

There is also a parallel effect with the AMF (Access and Mobility Management Function), which serves a given number of cells. The AMF manages the active context for each UE in the network – crucial information regarding identity, security, capabilities and more.

Where there is a loss of contact and no data has been exchanged, the AMF initiates paging activity to locate the UE and to restart the initial registration procedures and establish a new context.

This is a heavyweight procedure: authentication, security key computation, and profile lookup, which adds to the network burden. The situation is further compounded by the number of UEs in the area that are impacted by the shock.

Disasters have a knock-on effect throughout the network

Of course, other network entities might be affected and there may be other repercussions – so the operator wanted to model these and other potential scenarios.

The operator used Emblasoft Evolver to run trials that emulated all (or a subset) of the gNodeBs — and the UEs behind them. Evolver can also emulate the AMF, of which there are several in the live network.

Using Evolver, the operator emulated and focused on different scenarios – for example, gNodeB shutdown – in a way that emulated an earthquake scenario. Starting with those affected in the epicentre of the shock, other nodes could be shut down to emulate the spreading disruption until the whole network can be tested under this scenario.

Many other scenarios are possible. For example, paging-driven load. In this case, AMF pages were simulated by Evolver over a large population of UEs. This emulates fan-out procedures, through which the AMF sends new requests to locate UEs, based on specified tracking rules (the order in which the requests are sent, and the priority of distributing pages, based on last known location, for example).

It means that any operator can then measure how the AMF responds under either scenario — measuring, for example, registration success rates, response times, and how quickly service is restored as the network recovers.

Different scenarios can be run separately, or combined – since, in a real disaster, many issues could occur simultaneously, or within a short space of time.

But beyond the initial storm, there's a further effect worth noting: once the AMF starts processing this load, it has to call out to other core functions — the UDM for subscriber profiles, the SMF for session setup — to complete each registration. This knock-on effect means that the AMF isn't absorbing the load alone – it's just the first stop in a chain reaction across the core network.

Evolver provides vital insight into network resilience

Evolver's testing capabilities can show how the AMF would hold up under the initial load, as well as under the full ripple effect across the rest of the core. This is where passive monitoring also plays a role — but this will be covered in greater depth in our upcoming nScan paper.

Using Evolver, any operator can understand not just the impact of an earthquake on its current network configurations and operations, but also network resilience and recovery. Evolver can track what happens as services are restored, node by node.

A useful methodology here is to first establish a baseline. So, operators can run Evolver under a simulated normal load to capture a 'before' picture. Then, Evolver can run any disaster scenario – the two can then be compared. That before and after comparison turns a one-off test into a repeatable resilience benchmark, which is worth running after every major release, gNodeB change, or new slice deployment to provide the full picture.

Collecting such data means that the root cause of any issues detected can be located and pinpointed. In turn, fixes can be applied in the captive network, tested again under the same scenario conditions, and rolled out into the production environment.

As operators seek to enhance the resilience and survivability of their networks under stress conditions, the ability to emulate such scenarios – such as mass failure of the gNodeBs, registration storms and AMF behaviours – is a crucial tool in assessing readiness to survive such events. Evolver allows operators to prepare for unexpected situations in advance, particularly in regions where earthquakes, for example, are common.

We are also preparing a new paper on how Evolver and its sister product, the nScan passive monitoring solution can help operators model  different scenarios and integrate the results to provide a comprehensive framework to support disaster readiness across all generations of your mobile network.

Watch out for the paper soon or get in touch to find out how we can help you enhance mobile network resilience with comprehensive testing and monitoring. You can also download our original paper on managing network power cycles and how that relates to unexpected events, by clicking here.