Physical Downlink Shared Channel (PDSCH

For the parameterization of the Downlink, we analyzed the different sections of the standard. Here, only the FRCs (Fixed Reference Channels) will be considered, which are the channels with fixed characteristics. This type of information is quite useful, as the Physical Downlink Shared Channel parameters are used in several UE (User Equipment) tests, including:

• UE receiver requirements.

• UE maximum entry-level test.

Based on the standard, the first variable to choose will be the frequency band in which it will operate (table 1), since the options for the following parameters will be derived from this, such as the modulations used, the bandwidths, sub-carrier spacing, among others.

Once the Frequency Range has been determined, a bandwidth (BW) and a sub-carrier spacing (SCS) can be chosen to consider the combinations defined in the standard. Bearing in mind that, according to 3GPP, there are cases in which the combinations cannot be used and do not apply (ETSI, 2021), as for modulations, QPSK, 64QAM and 256QAM are used in the downlink FRCs (3GPP, 2020). Duplexing modes in 5G-NR are defined both in time and frequency, so FDD (Frequency Division Duplex) and TDD (Time Division Duplex) are used according to the specific needs of the channel to be configured. It is recommended that for transmission in paired and unpaired frequencies, FDD and TDD should be used, respectively.

Once the parameters to be used in the PDSCH were studied and defined, the graphical interface was designed and programmed (figure 2). The parameterizable variables are presented as “Drop Down” containers. The containers are used to deliver the values to the user in accordance with the standard and automatically change the options that can be accessed. That is, changing one parameter will influence others, and the interface will automatically adapt to this change.

Figure 2: PhysicalDownlink Shared Channel User Interface. 

Physical Uplink Shared Channel (PUSCH

With the Physical Uplink Shared Channel, there are types of channels well defined by the standard, called FRCs (Fixed Reference Channels) where, despite being fixed, they can be reconfigured in such a way as to provide some flexibility in the parameters, according to potential design needs. As in the PDSCH, the first variable to choose will be the frequency band in which it will operate, since we will derive the workable options for the following parameters from this, and it limits the bandwidths, sub-carrier spacing, and so on.

Once the frequency range has been determined, the bandwidth, sub-carrier spacing, and modulations such as QPSK, 16QAM and 64QAM can be chosen. Those parameters vary according to the type of FRC. The base properties were taken from the standard. It could be mentioned that in the uplink, it is appropriate to specify an N-cell-id (ETSI, 2021).

After the concepts were reviewed and the PUSCH parameters were defined, the graphical interface was designed in the development environment (figure 3), where the variables are presented sequentially as “Drop Down” containers. Similarly, containers are used to deliver the values to the user following the standard specifications.

Figure 3: Physical Uplink Shared Channel User Interface. 

In the case of the PUSCH, the interface is arranged so that the user starts with the base parameters, and according to this, the software suggests a Fixed Reference Channel that meets the requested characteristics. The types of FRCs used as a base can later be modified with additional parameters, such as bandwidth, duplex mode, and the N-Cell-ID that will serve as the identifier of a cell.

Urban Planning - SINR Maps

The test environment guidelines for 5G technologies reuse the test network design of its 4G predecessor. Therefore, the guidelines are defined in Section 8.3 of the Report ITU-R M.2135-1(ITU, 2010), where in the rural/high-speed, urban coverage and micro-cell cases, specific topographical details are not considered.

The base stations are placed uniformly, following a hexagonal layout. In this hexagonal design, three cells are located per site, where the direction of the antenna, cell range and ISD (Inter-site Distance) are also defined. The simulation will be a surrounding configuration of 19 sites, each with three cells. Users are assumed to be dispersed throughout the area (ITU, 2010). The initial default values were taken from the ITU-R M.IMT-2020.EVAL (ITU, 2020).

The bandwidth and the rest of the remaining parameters depend on the usage scenario. For the Dense Urban-eMBB test environment, the default bandwidth is 20 MHz with the transmitted power of 44 dBm. The frequency designated for the test environment is 4 GHz, and the height defined for base stations is 25 meters, as detailed in (ITU, 2020, p. 18). However, the simulator allows changing those parameters.

Also, different antenna radiation patterns can be used (figure 4 figure 5, figure 6, and figure 7). For each antenna pattern, a 3D graphic shows the directivity variation scale, intending to have a better idea of the antennas that are to be used in the simulations.

Figure 4: Standard antenna radiation pattern. 

Figure 5: Radiation pattern corresponding to the patch antenna. 

Figure 6: Radiation pattern corresponding to the standard 8x8 antenna array. 

Figure 7: Radiation pattern corresponding to the 8x8 array of patch antennas. 

The entire map will be built based on the location of the central antenna, and from this, together with the ISD values, the network diagram will be built. Two test locations have been defined, which are in the center of Riobamba and the center of Quito. Additionally, different locations can be tested defining the longitude and latitude values in decimal degrees.

The graphical interface was designed as shown in figure 8. The containers are used to deliver the values that are necessarily fixed to the user, while the editable fields are configured with their default value. These parameters can be changed at the user’s discretion to experiment with different scenarios.

Figure 8: Main parameters window for SINR maps. 

Also, to provide more information to the user, an information window was located, where the default parameters and some important aspects of the simulation can be verified, as shown in figure 9.

Figure 9: Information window about SINR maps. 

Parameters for Ray Tracing and Coverage

The proposed test environments for 5G use and consider different criteria from Report ITU-R M.2412 (ITU, 2021), which contains guidelines for the evaluation of radio interfaces for these technologies. After analyzing the physical structure of 5G-NR and the tentative SINR Maps, a deeper step would be conducted using the ray-tracing method.

Like in any telecommunications test environment, the basis will be the working frequency of the transmitter and receiver that will be displayed on a map, and from these, the corresponding calculations will be made. The main parameters considered are shown in table 2.

To construct a realistic scenario, the Site viewer tool was used, which allows importing external maps, such as the Open Street 3D building maps. These maps are created by a large community of people worldwide and are updated daily. Maps that can be downloaded from its own portal www.openstreetmap.org. These maps are the basis of the Ray Tracing simulator. With the imported map of the area of interest, transmitters or receivers can be placed and attribute their own characteristics. The MATLAB© antenna toolbox allows for creating sites and implementing antennas within the site viewer with different properties that can be modified in each item.

In the simulator’s case that was implemented, to calculate each of the reflections it is important to know the permittivities and conductivities of the external materials of the constructions and terrains, which can be calculated simply, as detailed in the ITU-R P.2040-1 recommendation (ITU, 2015). This equation is used to get the relative permittivity, where the calculated value depends on the working frequency and two constants:

Where a and b are constants characterizing the material, f is the frequency in GHz, and (' i’ dimensionless.

Physically, conductivity is interpreted as the facility offered by a material to establish a flow of electric current in it. According to their conductivity, materials are classified into conductors, semiconductors, and insulators. The international system (SI) units of conductivity are siemens/m. (^{Paz Parra, 2013}). Equation (2) described in recommendation ITU-R P.2040-1 (ITU, 2015) was used to calculate the conductivity.

Being f the frequency in GHz and ( is expressed in S/m. The values of c and d are constants characterizing the material.

One of the most important steps in the calculations for free space loss (FSL) is to determine the losses in clear-sky conditions, as described in equation (3). These are the losses that remain constant, and the value of the distance can be obtained by the Ray-Tracing algorithm and together with the permittivity and conductivity at each reflection, a more accurate model can be obtained.

When the frequency f is represented in MHz and the distance d is in km.

The location of the transmitter will be given in geographic coordinates, defined in their longitude and latitude. The height of the transmitter will depend on the test case; for the parameterizable simulator, an initial tower height value of 10 meters will be taken. The working frequency is a crucial point of the parameterizable simulator. The proposed software allows simulate the high frequencies used in 5G-NR; the default value is 28 GHz. The radiated power will depend on the user's approach and will be in Watts [W]. This parameter, like others, is left open to the user's criteria.

The types of antennas used can be generated from the same Phased Array System Toolbox™ that has integrated MATLAB software. A custom antenna from Report ITU-R M.2412, designated for the evaluation of 5G radio technologies, will be used in the simulator (ITU, 2021) (figure 10). An isotropic antenna was also used for more general tests, where all rays will be traced with the same gain in all directions.

Figure 10: Radiation pattern of ITU-R M.2412 antenna. 

For the simulator, an always fixed transmitter (gNodeB) was considered (figure 11). The antenna azimuth and elevation can be configured, which can be entered by the user or automatically assigned with the direction towards the receiver (beamforming mode).

Figure 11: ITU-R M.2412 Antenna in Transmitter. 

There are likely scenarios in which there will be no LOS (Line of Sight) between the transmitter and the receiver. Here, the antenna will be directed at the angle of the ray with the slightest delay. In figure 12, the user interface is presented. Two preloaded maps of the city of Riobamba and Quito were integrated into the simulator, plus two default transmitter locations for test scenarios. However, it is possible to upload a map in ‘.osm’ format that can be downloaded from different web portals.

Figure 12: Transmitter parameters window at the interface. 

The 5G receiver UE (User Equipment) can be placed in any position on the map, and all reflections with power above the threshold of -120 dBm, will be considered. Also, the UE has 2 types of behaviors: i) Fixed Receiver and ii) Mobile Receiver (Doppler Effect). If it acts as a mobile receiver, the receiver’s coordinates must be entered, and additionally, the coordinates of the direction movement (figure 13). It allows the simulator to calculate the respective Doppler shift for each contribution.

Figure 13: Receiver parameters window in the parameterizable simulator. 

While the empirical models are based on in situ measurements, the theoretical models are based on the fundamental principles of electromagnetic wave propagation phenomena. The propagation models predict the loss that can occur along the path that a radio frequency signal follows, between a base station and one or more receivers (UE) that can be fixed or mobile (^{Camargo, 2009}). The SBR (shooting and bouncing rays) algorithm is a hybrid method that combines PO (Physical Optics) and Geometrical Optics (OG). Electromagnetic propagation is first represented by optical rays’ reflection, refraction, and divergence. Millions of ray tubes are launched from the source to the scatterer, and the paths of each ray tube bouncing off the scatterer’s surfaces are plotted according to Snell’s law. Next, the electromagnetic properties of magnitude, direction, and phase are added to the ray traces to mimic the properties of waves. Then, the contribution of each ray tube to the stray field is calculated independently. Finally, the scatterer under the irradiation of the incident wave can be expressed as the superposition of the individual contributions of all the ray tubes (^{Xu, Dong, Zhao, Yin, & Chen, 2021}).

For a more intuitive use of the software, three switches were used, as shown in figure 14, one for each propagation model incorporated in the simulator, where the ‘Ray-Tracing SBR’ model will be activated by default. Besides this, it will be possible to activate rain models (ITU, 2005) or mist (ITU, 2019).

Figure 14: Propagation model parameters window in the parameterizable simulator. 

Recommendation ITU-R P.840-8 provides methods to predict attenuation because of clouds and fog, considering that it is necessary to give guidelines to engineers for the design and simulation of telecommunication systems in frequencies higher than 10 GHz. Clouds and fog are composed entirely of tiny droplets, generally less than 0.01 cm, so the Rayleigh approximation is valid for frequencies up to 200 GHz; the specific attenuation inside a cloud or fog is expressed in dB/Km (ITU, 2019). For the use of this fog and cloud propagation model within the parameterizable simulator, the working frequency of the transmitter and an extra parameter of the density of liquid water in the fog will be necessary, by default, it will be 0.5 g/m^3.

With the implemented simulator, to calculate each one of the reflections, it is important to know the permittivity of the exterior materials of the constructions, which can be calculated merely, as detailed in the recommendation ITU-R P.2040-1 (ITU, 2015).

The graphical interface of the exterior material for the constructions is designed so that one of the materials described can be chosen (figure 15). In addition, a custom material can be entered, defined by its permittivity and conductivity. This process is carried out analogously for terrain materials.

Figure 15: Parameter window for exterior materials in the parameterizable simulator. 

Once the parameters of the transmitter, receiver, and a propagation model have been defined, they can be combined with functions such as ‘coverage’. This function performs the computation as for a point, but in a particular area, considering the same attributes of each one of the elements. The button to generate the coverage graphic is also at the same interface. The last window with all the functions to the user is shown in figure 16.

Figure 16: Ray Tracing main window. 

The general graphic interface (figure 17) contains the cover and four buttons that access each of the mentioned parts of the project. It allows a single window of each type to be opened at a time, and when it is closed, all windows derived from it are closed.

Figure 17: Main menu of the 5G-NR parameterizable simulator. 

Finally, the parameterizable simulator was exported as an autonomous executable or ‘standalone’ for Windows and Mac OS (figure 18), and a MATLAB application (figure 19), being the distribution formats to the end users. These versions can be used for free, under the agreements specified by MathWorks®.

Figure 18: Executable and installation icon. 

Figure 19: Installation process MATLAB application. 

Physical Downlink Shared Channel

After the parameters are adjusted at the user’s discretion, the software will design the channel. In this case, the first to appear is figure 20, which is the magnitude of the FRC baseband waveform. The values of each of the samples are generated to satisfy the chosen options, and thus complete all the resource blocks.

Figure 20: Magnitude graph of the FRC baseband waveform. 

These samples, after going through the post-process, can be displayed as Resource Blocks, where the relationship of the symbols concerning their use is represented by colors. These blocks can be designated for the PDCCH (Physical downlink control channel), PDSCH, Coreset (Control resource set) or SS Burst (Synchronization signal burst). The length frame in 5G is 10 ms for FDD and 20 ms for TDD. Each frame comprises 10 subframes, and these subframes comprise their respective Slots, which will depend on the chosen numerology. With 30KHz, each 1 ms subframe comprises 2 slots of 14 OFDM symbols. An easy way to calculate the number of total symbols per 1ms subframe would be by multiplying the number of Slots defined in the numerology multiplied by the 14 OFDM symbols that make up each slot. A total of 280 symbols on the X-axis is obtained, as shown in figure 21.

Figure 21: Graph of the relationship between Resource Blocks and Symbols in FR1. 

Figure 22 illustrates the frequency bandwidth to be used, centered at 0 Hz and having the channel edges in dotted lines. The channels must have guard bands that are shown in red; the center point of the resource blocks is represented by 𝑘 0 and the center frequency by 𝑓 0 .

Figure 22: Graph of blocks of resources in bandwidth belonging to FR1. 

Figure 23: Graph of the relationship between the Subcarriers and the symbols in FR1. 

Each RB (Resource Block) is made up of 12 Sub-carriers. Therefore, the total number of Sub-carriers will be 288, having a maximum of 24 RBs for the chosen configuration. The relationship between each of the symbols and the Sub-carriers can be seen in figure 23.

Physical Uplink Shared Channel

In the case of the PUSCH, the interface is arranged so that the base parameters are entered first, and according to this, the software will suggest an FRC that adapts to the user's requirements.

Figure 24: Magnitude graph of the FRC baseband waveform. 

Additionally, the interface has additional parameters, which can be changed at the discretion of each user and will be helpful for testing. Figure 24 illustrates the relationship between magnitude and the symbols that will be entered into the resource blocks.

Figure 25: Relationship between the RBs and the symbols. 

For the chosen base FRC 'G‘F‘1-A2-1',’t’e standard states that there will have a total of 25 occupied RBs, which is why figure 25 shows only 25 occupied of the 52 total.

Figure 26: Resource blocks in bandwidth belonging to FR1. 

Figure 26 depicts the maximum number of RBs that can be accommodated within the guard bands of the selected bandwidth. The center point of the resource blocks is represented by 𝑘 0 and the center frequency by 𝑓 0 .

Figure 27: Relationship between the Subcarriers and the symbols in FR1. 

Figure 27 shows the relationship between the Subcarriers and the symbols of the chosen configuration. Note that the dark blue ones are not hosted blocks, and in the case of the blue ones that are inside the channel, they work as pilot symbols.

SINR maps for urban cell planning

As an example, a specific simulation using the parameters detailed in table 3 was performed.

In this case, when choosing the type of antenna, the radiation pattern of the chosen option will be displayed. After the parameters of table 3 were simulated, the distribution of the towers in the center of the city of Riobamba is shown as in figure 28, and it is visualized that since it is not such a directional antenna, the SINR tends to be lower.

Figure 28: SINR topology simulated on a satellite map. 

In the ‘topographic map’ type view, there is a better appreciation of the coverage area of the cells. Figure 29 presents a simulation made with the standard 8x8 antenna. Because of the antenna has better directivity, it can be seen how the cells and the area have better reception C/I, compared with figure 28.

Figure 29: Top view of SINR map within a topographic map. 

Ray Tracing in urban cells

A specific scenario inside the “ESCUELA SUPERIOR POLITÉCNICA DEL CHIMBORAZO (ESPOCH)” was simulated. Table 4 presents the simulation parameters.

In the simulation, the radiation patterns of both the transmitter and the receiver are displayed. The names of the elements and their properties are shown when clicked, including the received power, coordinates, and the antenna height, as can be seen in figure 30.

Figure 30: Top view of simulated Ray Tracing. 

Also, the graph of the PDP (Power Delay Profile) is shown, where each one of the rays received by the UE is characterized. In the upper graph of the figure 31, the power received from each of the rays are represented, and in the lower graph, is shown the normalized and excess delay PDP that arrives at the receiver.

Figure 31: PDP in the receiver obtained from Ray Tracing. 

In the second case, where the receiver has movement, each element can be viewed in the ‘Site Viewer’. As can be seen in figure 32, the direction is represented with a blue arrow towards a second location called ‘Place of Displacement’.

Figure 32: Ray tracing and visualization of the place of displacement of the receiver. 

Similarly, the PDP graph is shown, where each one of the rays received by the UE is characterized (figure 33). In the upper graph of the figure, the received power and the delay of each ray is represented, and in the lower graph, the PDP normalized to the first traced ray that reaches the receiver.

Figure 33: PDP in the mobile receiver obtained by Ray Tracing. 

With a mobile receiver advancing towards the ‘Place of Displacement’ at 5 Km/h, will cause the so-called ‘Doppler Effect’, which causes a shift in the frequency components to higher or lower frequencies, depending on the angle of arrival of the corresponding ray. Figure 34 shows the shifts in frequency produced by the movement of the receiver, represented in the Scattering function.

Figure 34: Scattering function in the simulation. 

Coverage in urban cells

For coverage, the same Ray Tracing criteria are considered, but in a particular area. As can be seen in figure 35, the transmitter, receiver model and coverage footprint are represented on the map, as well as an arrow between the transmitter and receiver model to verify LOS.

Figure 35: Coverage with Antenna ITU-R M.2412 and 0 reflections. 

Figure 36 presents the coverage map for the same case shown in figure 35, but with a single reflection. It can be seen the improvement in the reception area since more reflections were considered. These show the importance of reflections, especially in urban environments.

Figure 36: Coverage with antenna ITU-R M.2412 and 1 reflection. 

As mentioned above, the same ray-tracing criteria are considered for coverage, but in a particular area. In figure 36, the simulation of a transmitter with an isotropic antenna can be observed.

Figure 37: Coverage with isotropic antenna with 0 reflections. 

In the case of 10 reflections with the isotropic antenna (figure 37) and (figure 38) a better coverage can be seen in all directions and an increased power, even in not so accessible places; this because of the more significant number of contributions that will exist for each point.

Figure 38: Coverage with an isotropic antenna with 10 reflections. 

Comparison with a professional simulation tool

As part of the applicability evaluation, the parameterizable simulator was contrasted with other software on the market, which is an advanced engineering tool that provides multiple options for planning and designing cellular networks. Table 5 shows the parameters used for the comparison.

One model in the professional simulation tool is the standard propagation model, which is a model deduced from the Hata formula, and from which various studies confirm that it is a very suitable model for UMTS, CDMA2000, WiMAX and LTE, so it will be able to work at the test frequency of 2.8 GHz (^{Popoola, 2017}). In the case of the proposed parameterizable simulator, it will work with the Ray Tracing SBR propagation model that can perform the calculation for up to 10 reflections in any building scenario imported in 3D.

Figure 39: Coverage map with the Professional Tool. 

As can be seen in figure 39, the professional simulation tool also allows the use of propagation models that adapt to the needs of a user, being able to locate both transmitting and receiving antennas, and displaying results within the map. In the version of the professional simulation tool in which the simulations were carried out, there is no specific model for 5G, nor a ray-tracing propagation model, for which a frequency that fits in the FR2 of 5G was used for comparisons.

Figure 40: Coverage map with the proposed simulator. 

Figure 40 shows that the simulator generates similar results to professional simulation tools. Therefore, the simulator proposed applies for basic planning environments. In addition, it could be also used for education through the designed laboratory guides and support documents.

In situ measurements

Complementary to the development of the simulator, measurements in situ were taken. In the scenario, two Hyperlog antennas were used, an Anritsu MS2427C spectrum analyzer for the receiver, an Anritsu MG3690C signal generator besides cables and connectors. The collected data was helpful to know the difference between the simulations carried out and the measurements. The same geo-referenced locations were used in the simulator and in the measurement places. The proposed scenario is displayed in figure 41.

Figure 41: Measurement scenario. 

This test scenario was established with a frequency of 5.9 GHz and a transmission power of 18 dBm, a system loss margin of 10 dBm was set and the antennas have around 5 dBi gain. For the simulator, similar radiation patterns to real antennas were used, as shown in figure 42.

Figure 42: Antenna locations and radiation patterns. 

Once the samples were taken at the geo-referenced locations, they were compared with each of the results obtained at the same coordinates within the simulator.

The measurement and simulated results are presented in table 6, getting a difference of between +6.35 dBm and +8.04 dBm between the measurements and the results obtained by simulation, with a mean difference of +7.11 dB and standard deviation of 0.5579 (figure 43). Showing that the results are quite suitable for urban planning environments, considering that differences between calculated and measured values are inferior to in other popular propagation models, like those evaluated by ^{Sharma & Singh (2010}).

Figure 43: Normal distribution of the collected data. 

Applicability Survey

After the consolidation of the simulator, a qualitative survey focused on evaluating its applicability and collecting the points of view of the users was applied. This survey was carried out by 24 students of the subject of mobile communications belonging to the ESPOCH Telecommunications career. The technique used was the Likert Scale with five possible answers, from strongly agree to strongly disagree. The estimated completion time was 4 minutes.

The survey results showed that the users strongly agreed that the use of simulations is a valuable teaching method for learning, where 75% said they strongly agreed and 25% agreed.

Respondents from their level of knowledge perceive that the scenarios where the simulations are developed are close to reality, with 71% of respondents agreeing and 29% strongly agreeing. Showing that the use of 3D maps and visual resources helps to improve the understanding and interpretation of results.

There was a strongly positive response, since the students felt they could apply previously gained knowledge to experiment with the simulator. Having 54% of those surveyed agree and 38% strongly agree with this idea. Along with other results that helped to improve the software and reflected that students of these topics have a great interest in simulated environments as a complement to theory.