When trying to download a large file in the city center during rush hour, 4G struggles. The network becomes congested because thousands of smartphones are using the same antennas on the same frequencies. This is exactly the type of congestion that 5G aims to resolve, with a network architecture rethought from the radio layer to the core of the infrastructure.
5G Frequency Bands: What Changes on the Ground
5G does not operate on a single frequency. It utilizes three distinct frequency ranges, each exhibiting very different behavior in real-world situations.
Low bands (below 1 GHz) cover vast areas, including inside buildings. They are mainly found in rural or suburban areas, where the density of antennas remains low. The speed is better than 4G, though not spectacular.
Mid bands, around 3.5 GHz, form the core of the deployment in France. It is on this frequency that the difference with 4G is most felt in daily use: faster downloads, more stable video streams in dense areas. The range remains decent, although it decreases compared to low bands.
Millimeter bands (beyond 24 GHz) offer very high speeds, but their range is limited to a few hundred meters. A wall, a tree, or even heavy rain can block the signal. Their use is confined to very specific sites such as stadiums or industrial areas.
To fully understand the definition and functioning of 5G, it is essential to keep this reality in mind: the network you perceive directly depends on the band your antenna is transmitting on.
Latency and 5G Core Network: Why Speed Isn’t Everything
Speed is often discussed when talking about 5G. The gain is real, but the reduction in latency changes usage more significantly. Latency is the time it takes for a data packet to make a round trip between your device and the remote server.
In 4G, this response time is around several tens of milliseconds. In contrast, 5G aims to significantly reduce this delay, opening the door to applications that cannot tolerate any lag: remote machine control, assisted surgery, connected vehicles.
Network Slicing, a Structuring Innovation
The 5G core network introduces a feature absent in previous generations: network slicing, or dividing the network into virtual slices. Each slice can be configured independently to meet a specific need.
- One slice can be optimized for maximum speed, for example, for high-resolution video streaming in a public place
- Another prioritizes ultra-low latency, intended for industrial use such as controlling robots on a production line
- A third can simultaneously connect a very large number of low-power devices, typically agricultural sensors or smart meters
This slicing allows operators to offer differentiated service levels on the same physical infrastructure. For businesses, this means a mobile network that adapts to their needs, not the other way around.
Phasing Out 2G and 3G in France: 5G in a Transition Context
One cannot discuss 5G deployment without mentioning what is happening in parallel. French operators have initiated the gradual phasing out of 2G and 3G networks to free up frequencies and reallocate them to 4G and 5G. Arcep is closely monitoring this transition.
This shift has immediate practical consequences. Some older equipment, such as alarms, emergency call systems, or payment terminals, still operate exclusively on these old networks. If these devices do not support 4G or 5G, they will simply stop working.
For an individual with a recent smartphone, the change is seamless. For a business managing a fleet of connected devices deployed a few years ago, checking compatibility becomes a priority.
5G-Advanced and Non-Terrestrial Networks: What Comes After the Current Deployment
The 5G we use today is not static. The standard evolves in stages, called Releases, defined by the international consortium 3GPP. The next step is called 5G-Advanced, stemming from Release 18.
The announced improvements are not limited to increasing speeds. Three main areas are emerging:
- Energy optimization of the network, a central issue given the operating costs of 5G antennas
- Integration of artificial intelligence to dynamically manage radio resources based on actual load
- Support for non-terrestrial networks, meaning communication via satellites integrated directly into the 5G standard
This last point deserves attention. The specifications of Releases 17 and 18 provide for satellites to relay the 5G signal in areas where no terrestrial antenna is viable: sea, mountains, very remote regions. The continuity of mobile service would no longer depend solely on ground antennas.
The first compatible components have been presented, but feedback varies on the actual timeline for commercial availability. Having a component in a terminal does not mean that the antenna or the plan on the other side already supports these functions.
Private 5G Networks: An Underappreciated Use Outside the General Public
5G is not limited to mobile plans. Companies are deploying their own private 5G networks on their industrial, logistics, or hospital sites. The idea is to have a dedicated network, isolated from the public network, with guarantees of speed, latency, and security that are impossible to achieve through a standard plan.
Specific certification procedures exist for these non-public networks, regulated in particular by the GCF. This market remains discreet but is progressing rapidly, driven by the needs for automation and reliable connectivity in constrained environments.
5G, in its practical version, is not just a logo on a phone screen. Between frequency bands, network slicing, the phasing out of old networks, and the arrival of 5G-Advanced, the entire mobile infrastructure is evolving. The most useful thing is to check what your local antenna is actually emitting, rather than relying on a marketing promise.
