5G in Poland – the current state
5G in Poland has been a reality for some time, not just plans. The operators coped with the delay in frequency allocation in the 3.4-3.8 GHz band and launched their 5G networks using various techniques. In fact, we are seeing two methods of providing 5G transmission: Orange, Play, and T-Mobile share dynamically channels with a bandwidth of 10 or 15 MHz from the 2.1GHz frequency band between 5G and LTE, and Plus decided to use the 40MHz channel with 2.6GHz frequency band – previously intended for LTE networks with time division duplex (TDD). The latter solution, although less popular in Poland, but quite commonly used in other parts of the world, allows for a more effective division of frequency resources between downlink and uplink directions, which may translate into higher data transmission speeds.
All four 5G networks currently available in Poland operate in the 5G NSA (Non-Stand Alone) mode, which means that a phone using the 5G network also uses the LTE network, and frequencies assigned to it, in parallel. This means that when we observe the results of the 5G network, they are, in fact, to some extent the results of the LTE network. The impact of older technology is greater for operators who allocate channels with a smaller width for 5G (Orange, Play, T-Mobile), and is less important for Plus, which uses the 40 MHz channel.
Let’s check it how it is in practise
There are numerous promises and announcements associated with the introduction of 5G technology. The first and the loudest ones concern the transmission speed – declarations say about reaching hundreds of Mbps or even speeds of Gbps. Here the expectations should be tempered – without a dedicated band with a much larger width, we will stay much closer to what we know from LTE.
An equally important aspect – although evoking much fewer emotions – is the expectation that 5G will reduce the delay in sending packets between the phone and the server (Round Trip Time latency). Here, the limiting factor is not the bandwidth available, but more the network architecture. Since all operators use the 5G NSA mode (based on the architecture of the older LTE network), the results should also be closer to what we know from LTE.
So what 5G network quality can be expected?
To fairly assess all operators, Systemics-PAB performed an independent 40-hour measurement in motion (Drive Test) covering the entire area of Warsaw (map below) from August 23-27, 2021.
Samsung Galaxy S21+ phones with firmware adequate for the local Polish market were used for the measurements. This terminal ensures a fair assessment of all operators, enabling the aggregation of all frequency bands currently owned by Polish operators. Measurement phones were kept in conditions of uniform, controlled temperature, and important parameters such as processor load or processor/battery temperature were monitored continuously. In the case of measurements – every detail is important – even such a small element as forcing a constantly illuminated screen on the measuring phone (some manufacturers – including Samsung – do not allow the phone to use 5G when the screen is turned off as part of energy usage optimization mode).
The range of the network was measured with a Rohde & Schwarz TSME6 radio scanner. The telephones and the scanner were controlled by the Rohde & Schwarz SmartBenchmarker v20.3.95 measurement system.
How to assess the 5G quality?
Comprehensive assessment of the quality of the 5G network requires an adequate methodology. The simplest solution is to “extend” the methodology known from LTE by tightening the requirements for transfer speeds, delays, and other basic parameters.
Such a shortcut, however, leads to nowhere. We are facing a generational change – and although it is still difficult to see it, it is enough to remember how we used the phone before the era of smartphones to understand the consequences.
The Systemcis-PAB methodology for assessing the quality of 5G has been created from scratch. The results of the European project METIS II (Mobile and Wireless Communications Enablers for the 2020 Information Society-II) were used as the base. This project assumes that the quality of the network will be assessed in 5 different ways of using the network by the user (Use Case):
Use Case 1 – access to information in networks with high density traffic: Fast-moving city user, using services of medium intensity: watching a video in 1080 or 4k quality
Use Case 2 – office in virtual reality: Although it may have been an abstract concept until recently, in the era of widespread remote work, this way of using the network will not surprise anyone anymore. Desktop user, huge demand for download/upload speed
Use Case 3 – broadband access to data transmission services everywhere: A user who moves in a suburban area, accepting the use of services of lower quality than in the urban area (UC1). Continuity of access to services is crucial.
Use Case 4 – the mass presence of the Internet of Things: Up to 1 million IoT devices within 1km2, small data volumes generated by a single device. The delay tolerances can be variable
Use Case 5 – connected cars: The key is error-free and reliable data transmission while maintaining a small delay – even at very high speeds.
Each of the above uses is defined by a set of qualitative indicators and the minimum values that the indicators should achieve to ensure the possibility of using at a good level the services appropriate to each application case. These requirements include high availability of services (network coverage and connection reliability), high data transmission speed from and to the network, low transmission delays, and the ability to connect many devices.
Since the current 5G networks are not yet 5G networks in the full sense of the word, we are aware that they cannot fully meet the expectations set for them. Lack of access to dedicated 5G services does not help the task of assessing network preparedness. The developed methodology assumes that we test 5G networks to verify whether the quality parameters communicated by the telecommunications industry can be achieved in the tested network.
The research used the iPerf3 and Interactivity tests, which allow for the flexible generation of data streams corresponding to various types of services. The generated traffic streams used both the UDP protocol, simulating the transfer of multimedia streams (e.g. for UC2) and TCP, simulating smaller data streams with the requirement for error-free data transmission.
The results of the study carried out from August 23-27, 2021 are presented in the form of a radar chart, which shows the degree of fulfilment of expectations for individual application cases. The values given in the diagram are expressed as a percentage in relation to the values of the qualitative indicators expected for the particular application examples. The table next to the graph shows the results with the best and the weakest results color-coded for each of the Use Cases.
As can be seen from the above list of 5G network results, Polish operators are currently best prepared to provide services belonging to the UC3, UC4, and UC5 application cases. However, there is still a bit of a gap to meet expectations. The situation is worse for UC1 and UC2 applications. The results will certainly be better when the frequency bands to be allocated under the expected 5G auction are implemented in the network. An important element of development will also be the expansion of the network with additional base stations and the transition from the 5G Non-Stand Alone architecture to 5G Stand Alone. The network ranges represented by the UC4 results correspond to the overall coverage on the streets and inside buildings near windows. For broadband access and good indoor coverage, the coverage is much smaller.
The study shows that for three out of five tested Use Cases, the operator Orange took the first place. It is certainly a surprise that Plus in four out of five application cases took fourth place due to lower transmission rates when sending data and greater transmission delay than the competition.
The values of the main qualitative indicators from the tests for general broad Internet access are as follows:
Detailed results for individual indicators within the application cases are summarized in the chart below. From the chart, we can learn how the individual indicators contributed to the results for each application case.
Systemics-PAB is a leading provider of independent Quality of Experience benchmarking services for mobile operators and regulators. Our mission is to assist customers to understand and address the variety of issues affecting the quality in mobile and fixed telecommunication networks.
We have conducted various benchmarking campaigns of mobile networks in more than 50 countries to maintain the high-quality services as they have transitioned towards 2G/3G/4G/VoLTE and now 5G.
Systemics-PAB is a member of the ETSI working group developing guidance for the quality of services testing in mobile networks and is a member of the 5G PERFECTA project.