Enhanced Mobile Broadband (eMBB)

The three main scenarios of 5G: Only one scenario can be used to its full extent in a network.

© Siemens

Today's 5G is based on Release 15. It is mainly focused on consumer needs and is supported by currently available handheld devices as well as some mobile network operators. eMBB aims to achieve peak data rates of 20 Gbit/s and real-world data rates of 1 Gbit/s per device. This is around 10 to 20 times faster than 4G.
To achieve this, 5G is adopting and improving the modulation scheme that 4G already uses, Orthogonal Frequency-Division Multiple Access (OFDMA). With OFDMA, information is sent via several sub-carriers. One of the main features is the sub-carrier spacing (SCS), i.e. the distance between two sub-carriers. There is a relationship between the symbol duration and the sub-carrier spacing. The connection between the sub-carrier spacing, the symbol duration and the slot-per-frames is called numerology and is indicated by the Greek letter µ. In contrast to 4G, 5G carrier spacing is dynamic. This allows it to be adapted to the needs of the user. As a result, 5G uses the spectrum more efficiently than 4G. The efficiency increases from 90 to 95%.

Each sub-carrier can be modulated using different modulation methods: QPSK (Quadrature Phase Shift Keying), 16-QAM (Quadrature Amplitude Modulation), 64-QAM and 256-QAM. The latter is not part of 4G. With higher modulation orders, more bits can be sent via an OFDM symbol, with a corresponding increase in the data rate. However, this requires better connection quality.
Depending on the number of available carriers, carrier aggregation is also possible. If a network uses two 20 MHz blocks, for example, two data streams can be sent to a terminal device at the same time, one via each carrier. The end device then recombines the data into one data stream.

The downlink data rate as a function of the frequency range: The values shown represent the maximum peak data rates theoretically achievable in a network with a single user.

© Siemens

The use of MIMO (multiple input multiple output) antennas is also possible with 5G. This means that a receiver can receive several data streams simultaneously via several antenna signals, which it recombines. The available frequency also has an influence on the achievable data rate. For 5G, this is currently 3.4 to 3.8 GHz - also known as the sub-6 GHz frequency spectrum - and the millimeter wave range around 26 GHz. Most of the frequencies already allocated for 5G in Europe are in the sub-6 GHz range.

In summary, this means that the maximum data rate depends on the number of carriers (carrier aggregation), the modulation, the number of MIMO layers and the frequency range. The data rates are calculated for the downlink under the assumption that the 256-QAM modulation method, the maximum eight MIMO layers and all slots for the downlink are used for a single user. They represent the theoretical maximum achievable peak data rates in a network with a single user.

In the current implementation of 5G, only Time Division Duplex (TDD ) is currently taken into account. This means that the allocated spectrum must be time divided between uplink, downlink and control channels, potentially reducing the above theoretical maximum throughput in a realistic environment.

It is true that 5G can reach around 20 Gbit/s in the downlink, but only at a bandwidth of 400 MHz, which will only be available in the millimeter wave spectrum. In the 3.7 to 3.8 GHz range, which has been allocated by the German government for private local applications, no more than 4.6 Gbit/s can be expected.
The most important scenario for the industry is URLLC. This main scenario will meet the demanding requirements of industrial IoT applications. It offers highly reliable communication with the lowest latency times and is intended to support applications ranging from automated guided vehicles (AGV) to wireless safety applications (e.g. emergency stop). These can be implemented with minimal latency.

Ultra-Reliable Low-Latency Communication (URLLC)

The latency and cycle times of radio traffic via 5G.

© Siemens

Latency is defined as the time it takes for a data packet to reach the user equipment (UE, end device) and be transmitted from the user plane function (UPF) to an external network or application. This is the end-to-end delay of the entire 5G system.

In 5G, flexible numerology (µ) can reduce the air transmission latency. To do this, an information slot is sent in less time. Self-contained slots and mini-slots within the slots are used. The 5G radio interface protocol is structured in frames of 10 ms. Each frame comprises 10 subframes of 1 ms each, which in turn are divided into slots containing 14 OFDM symbols. Depending on the selected numerology, a slot can last from 1 ms to a minimum value of 125 µs. In all cases, the number of symbols per slot is always 14, which means that in a subframe with an SCS of 30 kHz, a total of 28 symbols are transmitted within 1 ms. In a 60 kHz SCS slot, 56 symbols can be transmitted and so on.

These values apply to theory. Real values must be measured in a real stand-alone 5G network and only concern the physical part of the radio transmission. This delay must be added to all delays caused by the processing of the radio protocol and the delay in the core network.
Unlike the network nodes in older mobile radio technologies, the 5G core network is based on virtual functions. The advantage of virtual functions is that they can be easily deployed on an edge device, for example the software of the RAN (Radio Access Network), the UPF and possibly also as an application on the same machine, which results in lower latency.

Sub-carrier spacing and slot duration: The values mentioned are theoretical values. Real values must be measured in a real stand-alone 5G network and only relate to the physical part of radio transmission.

© Siemens

Industrial applications require communication latency to be as low as possible. A delay in communication leads to higher cycle times for the applications. The cycle time is the time required by a client application or device to send a request to a controller application and return the response. For deterministic and mission-critical industrial applications, it is therefore essential to keep the cycle time to a minimum. The 5G network introduces a delay in both the uplink and downlink, which does not necessarily have to be the same. The applications also require a certain processing and response time. The sum of all these factors is the cycle time.

Latency times and cycles

The 5G spectrum in Germany: The Federal Network Agency has reserved 100 MHz from 3.7 to 3.8 GHz for local use.

© Siemens

With 5G, several points contribute to latency: processing in the end device, wireless transmission, 5G protocol processing, the connection between the RAN and the core and the delay in processing the core network protocol. The overall performance also depends on the implementation of the 5G network.

In addition, the latency in private and public networks can be very different. If a company uses a public mobile network, it must expect higher latency times, as the data has to travel long distances outside the plant to reach the core network. In comparison, private networks are installed entirely on site. As a result, there are no long distances between the various network elements and the delay in the backbone is reduced to a minimum. In addition, the implementation of the 5G network can be optimally adapted to the applications. A private network therefore achieves better latency times by default.

Vincenzo Fiorentino is System Manager Industrial Wireless Communication at Siemens.

© Siemens

According to the specifications of 5G, the shortest Packet Delay Budget (PDB) has an upper limit of 5 ms, which is to be guaranteed by the use of 5G quality of service classes. However, there are currently no known measurements that prove such values. A realistic latency value of around 10 ms can therefore be assumed, at least for the first 5G deployments.

Advantages of private networks

Sander Rotmensen is Head of Product Management for Wireless Industrial Communication at Siemens.

© Siemens

Private 5G networks enable companies to take advantage of the benefits of 5G. A private network is more secure than a public network, as the wireless signal is only available where it is needed. In a privately owned network, the data remains on site and users can decide which data is processed where - in a cloud, for example. In addition, users can control their private 5G network themselves and are able to adapt it depending on the application. Sufficient spectrum must be available to the industry for such private networks.
Many countries are considering making local spectrum available either in the sub-6 GHz range or in the millimeter wave range or even in both. In Germany, the Federal Network Agency (BNetzA) has already reserved 100 MHz from 3.7 to 3.8 GHz for local use. This gives companies the opportunity to rent spectrum for a low annual fee.