What do the roaming behavior of an M2M SIM card, the battery life of an IoT device, and road traffic have in common? More than you might think at first glance.
Just as a car consumes fuel when it has to keep starting up again and again in traffic, an IoT device also consumes energy when it searches for a suitable cellular network and repeatedly attempts to connect.
But it is precisely this energy efficiency that matters in battery-powered IoT applications—especially when devices are expected to function reliably for many years and replacing the battery is time-consuming or costly.
If you want to maximize battery life as much as possible, most people first look at the obvious factors: How powerful is the battery? How often does the device transmit data? How energy-efficient is the built-in hardware? All of these factors are, of course, important. However, one factor is often overlooked:
How does the device actually connect to a cellular network?
Before an IoT device can transmit data at all, it must first establish a stable connection. Depending on how freely it can choose between available cellular networks (key terms: steered and non-steered roaming), this process has a greater impact on battery life than one might think.
Spoiler alert: If a device has to try multiple times to connect to a preferred network, it consumes energy with each attempt. This happens with managed roaming, because the SIM card tells the device which network to connect to, regardless of how good or bad the reception is at that location. If, on the other hand, a device can directly select the network with the best available connection—as is the case with unmanaged roaming—it goes online faster, requires fewer connection attempts, and consequently consumes less energy.
That doesn't mean that non-steered roaming improves battery life every time. However, it does help prevent unnecessary power consumption under difficult network conditions. Why is that?
Steered vs. Non-Steered Roaming: Why Some Devices Always Take the Same Route
To understand how IoT devices can connect to the cellular network, let’s stick with our traffic analogy for a moment:
Anyone driving from Hamburg to Munich will likely select a route on their GPS and hit the road. If traffic conditions change along the way, the route is recalculated. A traffic jam here, a road closure there—and the GPS automatically suggests a better route.
The freedom to independently choose the best route to the destination aptly describes the approach of unmanaged roaming: Without external instructions, the IoT device can connect to the best local cellular network to achieve its goal—stable data transmission.
We all know from personal experience that cell phone reception can be strong, weak, or even non-existent depending on your location. Since some IoT devices require a (very strong) connection at all times, we offer our M2M SIM cards, which allow these devices to access multiple networks within a single country (national roaming). And to ensure they’re always connected to the strongest local network, we offer non-steered roaming Non-Steered Roaming).
However, not all IoT devices enjoy this freedom.
With what is known as “steered roaming,” the SIM provider specifies which mobile network the device should prioritize. The device therefore always attempts to connect to that network first. Even if another network at the same location offers a better connection. To return to our navigation system example: Whether due to a roadblock or traffic jam, the route is not changed—no matter how long it ultimately takes to reach the destination. In the worst-case scenario, these connection attempts continue until the preferred mobile network is no longer accessible at all. For this reason, steered roaming not suitable for all IoT applications.
Side note: When is managed non-steered roaming appropriate, and when is non-steered roaming appropriate?
In addition to real-time data transmission, there are other factors that should be considered when choosing a roaming plan. These include, among others:
Not every application has the same requirements. Some require maximum availability, while others require maximum uptime. Devices that transmit critical real-time data—such as a patient’s vital signs from an ambulance—face an increased risk of connection problems with controlled roaming and, in the worst-case scenario, could endanger human lives. In contrast, for a smart meter that transmits a reading only once a day, brief delays are usually inconsequential.
The situation only becomes critical when the battery runs out much sooner than expected. And this is exactly when it’s worth taking a look at a device’s roaming behavior.
Why choosing a network costs energy
Let’s go back to the example of driving for a moment. In fact, the connection between roaming behavior and battery life can be explained surprisingly well using the analogy of a daily commute.
Let's imagine we drive to work every morning. There are two routes:
- Route A, the officially recommended route, and
- Route B, which, however, is less crowded on some days and therefore faster.
If you were forced to take Route A every morning—regardless of how much traffic is on the road—you would spend more time on the road, use more fuel, and get stuck in traffic more often. The same applies to IoT SIM cards and devices that are forced to always connect primarily to a specific cellular network. They consume energy by repeatedly attempting to connect to a network that isn’t actually good enough for the desired data transmission.
But why is that?
Most people assume that an IoT device mainly consumes power when it transmits data. In fact, it uses quite a bit of energy even before a single byte is sent. One reason for this is that the device, prior to data transmission,
- Search for available mobile networks,
- establish a connection,
- log in to the network and
- must maintain the connection.
You can think of it like a laptop that’s turned on in the morning. While opening an email itself only takes a few seconds, booting up the operating system, logging in, and establishing an internet connection beforehand also take time and energy.
The same is true for IoT devices. The actual transmission of a measured value often takes only a fraction of the total communication time. That’s why, for battery-powered sensors that transmit only a few bytes per day, establishing a connection can actually consume more energy than the actual data transmission. If a device has to search for a suitable network multiple times due to controlled roaming or repeatedly attempt to connect, this energy consumption increases further.
Or, to stick with our traffic example: Someone who gets to work every morning without detours or traffic jams uses less fuel than someone who regularly gets stuck in traffic or has to make several attempts to reach their destination.
Technical Deep Dive: Where Energy Consumption in Mobile Communications Actually Occurs
To wrap things up, let’s take a slightly more technical look at the background.
The fact that non-steered roaming can save non-steered roaming is not just because a device connects to the network more quickly. There are also several technical factors that affect energy consumption.
1) Lower transmission power
Mobile phone modules adjust their transmission power based on signal quality. If the signal is weak, the device must transmit at a higher power level to be detected by the mobile network. If the signal is strong, significantly less power is required on the device side.
You can think of it like a conversation: someone standing directly across from another person speaks more quietly than someone who wants to be heard across a noisy room.
2) Fewer retransmissions during data transmission
Poor wireless conditions often result in data packets having to be retransmitted—much like a conversation in which a sentence isn’t understood and has to be repeated. Perhaps even several times.
These retransmissions also consume energy. A stable connection therefore not only ensures higher availability but also reduces unnecessary power consumption.
3) More time in power-saving mode
Many modern IoT devices spend most of their time in sleep mode and wake up only briefly to transmit data. LPWAN (Low Power Wide Area Network) technologies such as NB-IoT and LTE-M, for example, were developed specifically for this use case. With the help of power-saving features such as Power Saving Mode (PSM) and extended Discontinuous Reception (eDRX), devices can deactivate their radio modules for long periods of time, thereby achieving battery life of several years.
However, these mechanisms only work optimally if communication with the cellular network is reliable. If a device has to search for a suitable network multiple times after waking up, compensate for dropped connections, or resend data repeatedly, some of the energy-saving potential is lost. The faster the communication, the faster the device can return to its power-saving sleep mode.
A real-world example
Let’s take a smart meter in the basement of an apartment building. Three cellular networks are available there:
- Network A has a weak signal
- Network B with a strong signal
- Network C with a very strong signal
With steered roaming, the device always tries to use Network A first, because it is at the top of the priority list. The result:
- It takes longer to establish a connection because the signal is weak.
- It takes several attempts before the connection is established.
- The device itself must transmit at a higher power level (and consume more energy).
- Connection drops are occurring more frequently.
With non-steered roaming, the device would select Network C directly.
- The connection is established faster, remains more stable, and requires less energy.
- The device itself needs to use less energy for data transmission and returns to sleep mode more quickly.
A single connection attempt makes little difference, but over several years of operation, these small savings add up. This effect is particularly relevant in situations where devices are expected to operate for many years without maintenance:
- Smart meters in basements
- Sensors in Building Automation
- Smart City applications
- Environmental and water level sensors
- Industrial IoT devices in hard-to-reach locations
Every connection attempt that is avoided helps extend battery life and thus reduce maintenance costs.
Conclusion: The shortest route saves energy
The battery life of an IoT device depends on many factors. In addition to battery size, hardware, wireless technology, and power-saving features, the device’s behavior when selecting a network also plays a role.
There is no one-size-fits-all right or wrong answer. Both steered and non-steered roaming their place, depending on the application, technical environment, and project requirements.
steered roaming be useful when network connections provide a continuous power supply to devices or when certain commercial and regulatory requirements must be met.
When it comes to battery-powered IoT devices, however, which are designed to operate autonomously for many years or move through changing network environments, non-steered roaming offers non-steered roaming advantages. The ability to directly select the best available network helps avoid unnecessary connection attempts and reduce energy consumption.
One thing is particularly important to keep in mind: The choice of mobile network affects not only a device’s connection quality but also its energy consumption. Or, to return one last time to the example of road traffic:
People who reach their destination without detours, traffic jams, or constant stops tend to be more efficient on the road.
