Empirical Study of the Stochastic Nature of Electromagnetic Field Exposure in Massive MIMO Systems (5G)

2022 · Unclear · Low evidence

← Research stream

UnclearLow evidencePaper 749

Study snapshot

Publicationmanual · 2022

Study typeexposure_assessment

Reported directionUnclear

Evidence ratingLow

Abstract

Empirical Study of the Stochastic Nature of Electromagnetic Field Exposure in Massive MIMO Systems (5G) F. Héliot, T. H. Loh, D. Cheadle, Y. Gui, M. Dieudonne. An Empirical Study of the Stochastic Nature of Electromagnetic Field Exposure in Massive MIMO Systems. IEEE Access, vol. 10, pp. 63100-63112, 2022, doi: 10.1109/ACCESS.2022.3182236. Abstract Base stations (BSs) rely on the massive multiple-input multiple-output (mMIMO) technology in the fifth generation of mobile networks (5G). A technology having a major impact on the nature of the electromagnetic field (EMF) exposure in such systems. This work has used a fully reconfigurable mMIMO testbed (operating at 2.63 GHz), capable of mimicking realistic 5G new radio (NR) BS beamforming performance, to first gather experimental-based evidence of 5G BS EMF exposure within a real-world outdoor environment, to then analyze its stochastic behaviour, and to finally understand its impact on the definition of exclusion boundaries for 5G BSs. The exposure data of our testbed have been complemented by exposure data collected from a typical commercial 5G BS (operating at 3.65 GHz) to confirm the result trends and findings of our analysis. A robust metrology has been followed to obtain all the EMF exposure data. Our data and analysis indicate that significant exposure variations can be noticed according to the beam directions, i.e. the relative position of the exposure measurement location to the beam directions as well as the environment, confirming the stochastic nature of 5G BS exposure. The variance of the exposure tends to decrease as the number of users increase for a constant traffic load. Whereas the exposure grows sub-linearly with the traffic load, regardless of the number of users. As far as the exclusion boundary of 5G BS is concerned, its revised definition based on 95-th percentile seems still not flexible enough to accommodate the deployment of 5G BS in countries/places with stringent EMF exposure limits, as for instance in Belgium. Conclusion This paper has studied the stochastic nature of the EMF exposure and how it can impact the de finition of the exclusion boundary, especially under the constraint of stringent exposure limits, by relying on two different approaches for gathering experimental-based evidences. First we have used a fully reconfi gurable mMIMO testbed for evaluating the spatial and temporal variations of the RF-EMF exposure in a controlled outdoor scenario. Then, we have used a typical commercial 5G BS to complement/confirm some of the results/trends obtained via the mMIMO testbed. Several important insights can be inferred from our measurement campaign results: EMF exposure due to 5G BS grows linearly with the number of utilized RF chains at the BS. Significant exposure variations can result from the beam directionality, i.e. the relative position of the exposure measurement location to the beam direction has a significant impact on the measured exposure. This suggests a very dynamic/changing exposure environment driven by active UEs in 5G, in clear contrast with previous cellular technologies. Environmental effects, such as indirect propagation paths (e.g. re flecting off objects) or blocked paths can also create further variations. The variance of the exposure tends to decrease as the number of active UEs increases when the traffic remains constant, or in other words it becomes more deterministic. Whereas the exposure grows sub-linearly with the traffi c, regardless of the number of active UEs. The statistical distribution of EMF exposure for 5G BSs should be better understood and probably taken into account for defi ning the exclusion boundary. The results of our two measurement campaigns indicate that the statistics of different 5G BSs are not necessarily the same, i.e., the exposure of our mMIMO testbed BS follows a Poisson distribution while the commercial 5G BS exposure is closer to a normal distribution. In turn, this can impact the exclusion boundary size when it is defi ned as a particular percentile of the exposure cdf. Defining the BS exclusion boundary based on the 95-th percentile seems to not be a exible enough approach to accommodate the deployment of 5G BS in countries/places with stringent EMF exposure limits. Our measurement results correspond to a boundary size in the order of several tens of meters for an exposure limit of 6 V/m based on the 95-th percentile. The quality and quantity of field strength data collection need to be carefully considered to properly tune the extrapolation parameters in the SSB method as well as to accurately model the EMF exposure distribution of a particular 5G BS site. In the future, we would like to extend this work to mm-wave frequencies and build an empirical statistical model of the EMF exposure based on the same metrology approach. ieeexplore.ieee.org

Machine Enhanced Logic

AI evidence extraction

Structured fields below were generated from the paper’s abstract and metadata to make the evidence easier to inspect. Automated extraction may be incomplete or incorrect; verify important details against the original paper.

Study typeexposure_assessment

Effect directionunclear

PopulationNot supplied

Sample sizeNot supplied

ExposureRF

Evidence strengthLowConfidence: 74% · Peer reviewed: yes

Main findings

Using an outdoor measurement campaign with a reconfigurable mMIMO testbed at 2.63 GHz and complementary measurements from a commercial 5G base station at 3.65 GHz, the authors report significant exposure variations with beam direction and environmental effects, consistent with stochastic exposure. Exposure variance decreased as the number of active users increased for constant traffic load, while exposure increased sub-linearly with traffic load; defining exclusion boundaries using the 95th percentile was reported as potentially insufficiently flexible under stringent limits (example given: 6 V/m in Belgium), yielding boundary sizes on the order of several tens of meters.

Outcomes measured

RF-EMF exposure variability (spatial/temporal) around 5G massive MIMO base stations
Relationship of exposure with beam direction, number of users, traffic load, and number of utilized RF chains
Statistical distribution of measured exposure (e.g., Poisson vs normal)
Implications for defining 5G base-station exclusion boundaries (e.g., 95th percentile; boundary size under 6 V/m limit)

View raw extracted JSON

{
  "study_type": "exposure_assessment",
  "exposure": {
    "band": "RF",
    "source": "base station",
    "frequency_mhz": null,
    "sar_wkg": null,
    "duration": null
  },
  "population": null,
  "sample_size": null,
  "outcomes": [
    "RF-EMF exposure variability (spatial/temporal) around 5G massive MIMO base stations",
    "Relationship of exposure with beam direction, number of users, traffic load, and number of utilized RF chains",
    "Statistical distribution of measured exposure (e.g., Poisson vs normal)",
    "Implications for defining 5G base-station exclusion boundaries (e.g., 95th percentile; boundary size under 6 V/m limit)"
  ],
  "main_findings": "Using an outdoor measurement campaign with a reconfigurable mMIMO testbed at 2.63 GHz and complementary measurements from a commercial 5G base station at 3.65 GHz, the authors report significant exposure variations with beam direction and environmental effects, consistent with stochastic exposure. Exposure variance decreased as the number of active users increased for constant traffic load, while exposure increased sub-linearly with traffic load; defining exclusion boundaries using the 95th percentile was reported as potentially insufficiently flexible under stringent limits (example given: 6 V/m in Belgium), yielding boundary sizes on the order of several tens of meters.",
  "effect_direction": "unclear",
  "limitations": [],
  "evidence_strength": "low",
  "confidence": 0.74,
  "peer_reviewed_likely": "yes",
  "keywords": [
    "5G",
    "massive MIMO",
    "mMIMO",
    "base station",
    "beamforming",
    "RF-EMF exposure",
    "outdoor measurements",
    "stochastic exposure",
    "exclusion boundary",
    "95th percentile",
    "traffic load",
    "users",
    "RF chains",
    "2.63 GHz",
    "3.65 GHz",
    "Belgium"
  ],
  "suggested_hubs": [
    {
      "slug": "5g-policy",
      "weight": 0.62,
      "reason": "Discusses implications of exclusion-boundary definitions (95th percentile) under stringent national exposure limits (example: Belgium) for 5G base-station deployment."
    }
  ]
}

AI can be wrong. Always verify against the paper.

AI-extracted fields are generated from the abstract and metadata and may be incomplete or incorrect. This material is for informational purposes and is not medical advice.

Identifiers and source

DOI: 10.1109/access.2022.3182236
PubMed: Not supplied

Read the original paper →

Related papers

← Back to the Research Library