1. Introduction
A step forward toward the use of millimeter waves (mmWaves) in fifth-generation (5G) radio interface technology allows the use of larger bandwidth than previous mobile generations, allowing the possibility to deliver gigabit per second (Gb/s) wireless services. This significant increase in traffic data has been conceived to cover multiple usage scenarios, from enhanced mobile broadband to ultra-reliable low-latency communications (URLLC), going through massive Internet of Things (IoT) connections. New standards, such as the 5G New Radio (5G-NR) [
1,
2,
3], have included innovative techniques and procedures to overcome the unique challenges associated with mmWave transmissions. For example, mmWave transmitter antennas must be directional to take advantage of beamforming gains and cope with increased path loss and other propagation losses compared to sub-6 GHz frequency bands. Besides, 5G-NR benefits from a high flexibility level in several domains, such as the time domain (i.e., variable Time Division Duplex (TDD) schemes), frequency domain (i.e., bandwidth fractions), spatial domain (i.e., high adaptability in the implementation of beam sweeping or Multi-User Multiple Input Multiple Output (MU-MIMO) technology) and scalable numerology (enabling variations in subcarrier spacing as a function of the numerology parameter µ ranging from 0 to 5 from 15 kHz to 480 kHz and slot lengths given by 1/2 µ ms) in order to optimize the usage of temporal and spatial resources in the communication channel.
For indoor scenarios with a high user density (i.e., convention centers, event halls, concerts, indoor stadiums, etc.) and/or enterprise deployments (i.e., office buildings, shop floors, meeting rooms, auditoriums, libraries, etc.), 5G-NR mmWave can complement existing Wireless Local Area Network (WLAN) deployments with new and enhanced mobile broadband experiences, bringing a multi-Gb/s low-latency channel capacity, supporting devices beyond smartphones (i.e., tablets, always-connected laptops, augmented reality (AR)/virtual reality (VR), etc.) and therefore leveraging the existing infrastructure.
The advantages of 5G networks are well-discussed in the literature [
4,
5], and there is no doubt about the need for faster and more reliable wireless communication system deployments, with broadband data access in crowded locations. However, at the same time, the implementation of mmWave new technologies has increased the population concern over the possible impact on health and safety arising from the radiated electromagnetic field (EMF) exposure by these systems. This concern has led to the requirement of having accurate EMF simulation and measurement techniques to analyze the radiation exposure in the current and future wireless crowded scenarios. These techniques can verify compliance or not with the regulations from the point of view of radioelectric exposure of nonionizing radiation.
In the past few years, there has been a significant effort by the research community to provide clear EMF exposure insight with the presentation of different models for the RF-EMF assessment of 5G communication systems. From an engineering perspective, these RF-EMF exposure assessment models can aid in the design and deployment of 5G communication systems, achieving a good tradeoff between efficiency and operation, minimizing radiation exposure. On the one hand, most of the works are carried out in outdoor scenarios and present EMF assessments focused on the downlink (DL) of a 5G system. The work presented in Reference [
6] proposed a simulation technique to assess EMF exposure in 5G cellular systems operating at sub-6 GHz frequencies. The novelty of their work is that they propose a localization-enhanced pencil beamforming technique in which the traffic beams are tuned in accordance with the uncertainty localization levels of the User Equipment (UE) in the DL configuration. The authors in Reference [
7] presented a computational method based on ray tracing techniques (RT) to estimate human EMF exposure in DL 5G base stations at sub-6 GHz frequencies in outdoor macrocell environments. The same authors presented in References [
8,
9,
10] a combined numerical approach based on RT and Finite Difference Time Domain (FDTD) techniques, which estimates EMF exposure for 5G massive MIMO in different environments, such as an industrial environment [
8], an urban microenvironment (UMi) [
9] or an urban macro-environment (UMa) [
10], all of them at 5G frequencies below 6 GHz. The work in Reference [
11] presented a statistical approach to obtain realistic maximum power levels of 5G gNodeB (gNB) for the assessment of EMF exposure, employing massive MIMO for DL scenarios. An experimental and sequential statistical analysis for the assessment of EMF exposure from a massive MIMO 5G testbed was presented in Reference [
12]. The work focused on a 5G DL operating at sub-6 GHz frequency bands in an indoor empty scenario with low topological complexity. The impact of different beam profiles and number of users was assessed by means of a campaign of measurements. Moreover, an EMF exposure assessment from DL base station transmissions in a commercial 5G network was also analyzed in Reference [
13]. Finally, the work in Reference [
14] presented a comprehensive exposure assessment methodology to measure EMF radiation exposure from DL 5G NR base stations, using conventional spectrum analyzer equipment.
On the other hand, other works in the literature focused on the EMF assessment of the uplink (UL) from the UE in a 5G system. These works usually focused on the exposure assessment of a unique user device and its interactions with the human body. The authors in References [
15,
16,
17,
18] presented the impact of EMF exposure from mmWave phased arrays in mobile devices for 5G communication systems. The work presented in Reference [
19] investigated the maximum Effective Isotropic Radiated Power (EIRP) that can be achieved at 28 and 39 GHz considering beamforming UE under the constraints of the incident power density regulation limits. Colombi et al. presented in Reference [
20] the analysis of radio frequency (RF) energy absorption by biological tissues from mmWave 5G wireless devices in near-field conditions, showing low radiation effects of the near-field body interactions when evaluating EMF compliance at mmWave frequencies. In addition, the authors in Reference [
21] presented actual 5G UE output power levels operating in the current commercial 5G networks below 6 GHz, showing that the time-averaged output power levels were, in all cases, well below the regulation limits.
Table 1 summarizes the different proposed methodologies for the RF-EMF exposure assessment of 5G communication systems presented in recent years. As it can be seen from the table, there were few works in the literature that presented an approximation in which simulation estimations were simultaneously combined with measurement results. References [
6,
7,
8,
9,
10,
11] presented different methodologies based only on simulation techniques, all for DL and sub-6 GHz frequency bands. Conversely, the works in References [
12,
13,
14] presented EMF exposure assessment methodologies based on measurements campaigns, again at sub-6 GHz frequency bands and DL assessments. Nevertheless, less attention has been given in the literature to the assessment of human EMF exposure considering both DL and UL in complex heterogeneous real-world crowded environments, where multiple wireless communication systems coexist, which is the focus of this work, providing a multi-Gb/s low-latency channel capacity and enhanced mobile broadband experiences to the users.
Accordingly, based on an in-house implemented deterministic 3D ray launching (3D-RL) approach, a novel enhanced simulation tool for RF-EMF exposure assessment is presented, allowing the EMF characterization of complex context-aware scenarios with high node density heterogeneous networks. In this sense, combined wireless communication scenario setups with both DL and UL connections in crowded environments can be analyzed considering realistic operation conditions. The previous version of the simulation tool has already been validated for the assessment of EMF exposure of the current wireless cellular technologies in complex environments [
22,
23,
24]. In Reference [
22], the spatial characterization of UL personal RF-EMF exposure in public transportation buses was presented, where worst-case studies considering different user densities and distributions for sub-6 GHz cellular communication systems were evaluated in terms of legislation compliance. The work in Reference [
23] reported a simulated and experimental comprehensive analysis of the UL from a 2G–5G cellular system exposure assessment within a public tramway. Although both scenarios, the bus and the tram wagon car, could be considered as complex indoor scenarios in terms of radio wave propagation, the metal structure influence of the tram, as well as the supplying lines and towers and, specifically, its presence in the city central urban districts with huge passenger affluence, involved much more challenging propagation phenomena and, thus, presented higher exposure average levels. Reference [
24] presented an environmental RF-EMF radiation exposure assessment from an empirical and simulation approach in public shopping malls, focusing on the current wireless communication systems at sub-6 GHz frequency bands. The main differences of these works, as well as the differences with the presented work, are shown in
Table 1.
Table 1.
Different methodologies for the assessment of RF-EMF exposure for 5G.
Table 1.
Different methodologies for the assessment of RF-EMF exposure for 5G.
Ref. | EMF Assessment | Carrier Frequency | Beamforming | Environment | Simulation | Measurements | Channel Model Analysis | Description |
---|
DL | UL | Sub-6 GHz | mmWave | Fixed Beams | Flexible Beams |
---|
[6] | ✓ | ✗ | ✓ | ✗ | ✗ | ✓ | UMi | ✓ | ✗ | 3GPP UMi-Street Canyon Model Release 16 [25] | Localization-enhanced pencil beamforming technique, in which the traffic beams are tuned in accordance with the uncertainty localization levels of User Equipment (UE). |
[7] | ✓ | ✗ | ✓ | ✗ | ✓ | ✓ | UMa | ✓ | ✗ | RT | Computational method to estimate human EMF exposure in DL 5G base stations in outdoor macrocells environments. |
[8] | ✓ | ✗ | ✓ | ✗ | ✓ | ✗ | Industrial env. | ✓ | ✗ | Hybrid approach: RT/FDTD | Numerical approach for massive MIMO human exposure assessment in industrial environments. |
[9] | ✓ | ✗ | ✓ | ✗ | ✓ | ✓ | UMi | ✓ | ✗ | Hybrid approach: RT/FDTD | Numerical approach that estimates EMF exposure for 5G massive MIMO considering the effects of electromagnetic coupling between a user and the receiving device. |
[10] | ✓ | ✗ | ✓ | ✗ | ✗ | ✓ | UMa | ✓ | ✗ | Hybrid approach: RT/FDTD/Network planning methods | Novel method to design massive MIMO 5G networks under power consumption and EMF constraints. |
[11] | ✓ | ✗ | ✓ | ✓ | ✓ | ✗ | LoS conditions | ✓ | ✗ | - | Model for time-averaged realistic maximum power levels gNBs based on a statistical approach. |
[12] | ✓ | ✗ | ✓ | ✗ | ✗ | ✓ | Indoor empty room | ✗ | ✓ | - | Statistical assessment from experimental measurements in DL 5G at sub-6 GHz frequency bands considering an empty room with low topological complexity. |
[13] | ✓ | ✗ | ✓ | ✗ | ✗ | ✓ | Dense urban area | ✗ | ✓ | - | EMF exposure assessment based on real network data from base stations in a commercial 5G network. |
[14] | ✓ | ✗ | ✓ | ✗ | ✓ | ✓ | Urban | ✗ | ✓ | - | Exposure assessment methodology for measure with common spectrum analyzer equipment 5G NR base stations DL exposure. |
[21] | ✗ | ✓ | ✓ | ✗ | ✗ | ✓ | Dense urban area/Urban area | ✗ | ✓ | - | EMF exposure assessment based on real network data from 5G UE operating in commercial 5G networks. |
[23] | ✗ | ✓ | ✓ | ✓ | ✓ | ✗ | Indoor vehicle | ✓ | ✓ | RL [26] | Deterministic model to assess RF-EMF exposure of different systems within indoor metallic vehicles with different users’ densities and distributions, and comparison with current cellular technologies. |
[24] | ✓ | ✗ | ✓ | ✗ | ✓ | ✗ | Shopping malls case study | ✓ | ✓ | RL [26] | Empirical and deterministic model to assess RF-EMF exposure on sub-6 GHz shopping malls case study. |
This work | ✓ | ✓ | ✗ | ✓ | ✓ | ✓ | Indoor complex env. | ✓ | ✓ | RL [26] | Empirical and deterministic model to assess RF-EMF exposure on mmWave high-node density complex heterogeneous environments, with high topological complexity where all the scatterers are included. |
In this work, a step further is proposed for the EMF exposure assessment technique by considering mmWave frequency bands, allowing beamforming emulation with flexible beams, single-user MIMO (SU-MIMO) and multi-user MIMO (MU-MIMO) for both DL/UL (although, nowadays, MU-MIMO is only a DL feature, the simulation tool can consider MU-MIMO also in UL to emulate a potential technology development taking advantage of cooperative receiver schemes), different heterogeneous wireless communication technology analyses, complex indoor environments considering all the scatterers of the scenario, UE and gNB scenario densification and multifrequency operation, etc. The enhanced 3D-RL simulation technique has been implemented for the EMF exposure assessment of a potential complex indoor crowded scenario where two wireless communication systems will coexist in the mmWave frequency range. The selected scenario is a two-floor partial area of a new library building located on a university campus in which two different systems have been implemented: 5G personal mobile communications at frequency range 2 (5G-FR2) and wireless data access services WLAN 802.11ay at 60 GHz. The novel aspects of this work in relation with the previous works are the following:
- -
High-node user density environments: Software implementations have been performed in the 3D-RL algorithm kernel to allow user densification scenarios. By means of multiple integrated simulations considering high-node user density scenarios of increased complexity, the raw data is merged and collected in a new module to provide accurate final results. Following this procedure, the enhanced simulation tool is able to adequately reproduce the behavior and influence of environmental RF-EMF radiation exposure, considering different approaches: from a specific local exposure assessment in a particular communication beam to the overall exposure distribution in all the selected scenarios.
- -
Beamforming techniques: The consideration of 5G MIMO antennas and beamforming is provided by means of a new enhanced beamforming strategy that has been implemented using postprocess modules and multiple integrated simulations in order to simulate real antenna operations under future expected severe UL and DL conditions in complex high-node user dense context-aware heterogeneous environments.
- -
Complex heterogeneous environments at mmWaves: The consideration of heterogeneous environments in terms of electromagnetic fields and, specifically, in the exposure assessment and dosimetric characterization, is pivotal due the advent of new wireless communication systems and their unstoppable widespread use. It is a reality that current and future communication systems have become and will be increasingly heterogeneous, providing services and applications relying on coexisting merged heterogeneous networks in order to comply with coverage/capacity relations. Technically, when considering a complex heterogeneous system in the proposed enhanced simulation tool, a full analysis of the spectrum use needs to be performed in order to provide an adequate exposure assessment and evaluation based on the fact that multiple systems are operating at the same time in different frequency ranges, increasing the overall spectral usage. Thus, this corresponding electromagnetic influence is characterized in a new synchronized and integrated module by implementing merging techniques, as more intensive spectrum uses with multiple systems have unique characteristics. In addition, these challenging environments need to consider high complex scenario spatial designs in terms of the morphology and topology, with special attention on the scatterers’ frequency dispersive material properties. In this sense, the material properties’ database library of the new enhanced version of the simulation tool has been upgraded to consider material properties up to mmWaves.
The rest of the paper is organized as follows.
Section 2 presents the Materials and Methods, where the enhanced EMF exposure tool is presented, as well as the scenario description with the considered simulated study cases and the measurement campaign in order to validate the proposed EMF-enhanced technique. Then,
Section 3 reports the simulation results with the analysis of the different study cases and the measurement results. Finally, the conclusions are presented in
Section 4.
4. Conclusions
In this work, environmental RF-EMF exposure was assessed from an empirical and modeling approach in a complex heterogeneous indoor environment when considering dense personal mobile communications operating at 28 and 60 GHz. For that purpose, an enhanced in-house deterministic 3D-RL simulation tool for a RF-EMF exposure assessment was proposed, allowing E-field and incident power density level characterizations in different complex heterogenous scenarios of increased complexity. Realistic operation conditions were considered, as well as different user distributions and densities, emulating advanced mmWave communication systems with directive antennas and MIMO beamforming techniques.
From the obtained results, a discussion regarding the contribution, impact and health effects of the coexistence of multiple heterogeneous networks and services was provided. The main conclusion that must be stated is that wireless communication systems operating at both the 28 and 60 GHz frequency bands considering realistic and worst-case scenarios in terms of user densities, as well as directive antennas and beamforming techniques, generated environmental E-field exposure levels lower than 2 V/m and incident power density levels lower than 0.1 W/m
2 for all the analyzed cases. Therefore, the compliance with the current established international regulation limits with the exposure levels for the general population far below the aforementioned limits (10 W/m
2 [
54]) was verified even in the worst-case conditions and, at the same time, denying the hypothesis of an increase in the total RF-EMF exposure by the use of mmWave frequencies or directive antennas in complex dense heterogeneous indoor environments. In this sense, these results guaranteed that E-field distributions and received power levels within the complete scenario under analysis were below the thresholds and, hence, complied with the current regulatory frameworks in relation with the health assessment.
Moreover, the proposed simulation methodology was validated with a complete empirical campaign of measurements, showing good agreement with the experimental results. Thus, the obtained measurement datasets and simulation estimations, along with the presented enhanced 3D-RL simulation tool, could be a reference approach for the design, deployment and exposure assessment of the current and future wireless communication systems, where complex context aware scenarios with massive high-node density heterogenous networks are expected in the mmWave frequency range.