Instrumentation & Measurement Magazine 23-9 - 7

A further extrapolation technique can be applied for evaluating this quantity, starting from the knowledge of Emax:

E24 h  Emax *  24 h (2)

where α24h is a coefficient evaluated using (3):

1 24 Pi
(3)
24 
i 1 Pmax

 24 h 

where Pi is the average emitted power during the ith hour in a
day, and Pmax is the maximum power at the antenna terminals.
This coefficient is usually provided by the network operator.

Measurement Equipment and Evaluation
Technique of 3G Measurements
Typical dosimeters (PEMs) may be used for the long-term
monitoring of EMF temporal variations (in terms of the Efield or power density). Generally, PEMs are worn on the body
(e.g., on the belt), but they can also be used under isolated conditions of the unperturbed field ([3], [11]). Additional focus
has been the evaluation of 3G emissions through the integral-based-measure [10]. Because there is no 'referent day' to
compare the exposure with, the spectrum of the power density
fluctuations is seen as a convenient and compact way of describing very complex system behavior. Therefore, a different
approach (dosimetric) is used to quantify the effect of cumulative exposure to RF-EMF which estimates the contributions
of UMTS sources over the whole day (24 hours) and through
the week. Such a measure is described through the " exposed
energy density " parameter (expressed in J/m2), which is a cumulative sum of the instantaneous power density values in
time (see (4)). This parameter represents the amount of energy density transferred through a certain place over a certain
time, where any kind of absorption in the human tissues is not
considered.
i

W i  
t Sk , i 
1,  , m (4)
k 1

In (4), m is the number of measurement samples per day,
Δt is the time step of the sampling interval and Sk is the instantaneous power density value related to the kth measurement.

Experimental Results
Measurement Campaign and Instrument Set-up
for Narrowband LTE Measurements
The experimental campaign was carried out for several weeks
during November and December 2019, by monitoring three
different network operators in two specific LTE downlink
bandwidths (800 MHz, 1800 MHz). To maintain operator privacy, exact center frequencies are not provided in this paper.
In detail, seven days per operator and per bandwidth were
adopt-ed as an observation period.
To perform such measurements, an ad-hoc set-up was designed. Its sketch is reported in Fig. 1. The elements composing
the set-up are an active directional antenna, pre-characterized
December 2020

Fig. 1. The developed set-up to perform E-field measurements.

coaxial cables, a power splitter, a Rohde & Schwarz FSH8
Spectrum Analyzer (SA) working in a zero-span mode, and
a Rohde & Schwarz FSH4 Spectrum Analyzer (CP-SA) working in channel power mode. Both SAs are characterized by a
very high sensitivity (< -141 dBm/Hz), with a preamplifier (<
-161 dBm/Hz). In the set-up, they are both connected to a PC,
used as a bridge to save the acquired data on the cloud. Unwanted effects (both systematic and random) due to the set-up
arrangement were minimized by placing the antenna three
meters away from any metallic surface, as required by technical standards, and on the roof of the Engineering building at
the University of Cassino and Southern Lazio, Italy.
The antenna is placed 1.5 m above the trampling level.
Furthermore, no roof access was permitted to people during
experiments to avoid fading effects. As explained in the previous sections, two different analyses were carried out: daily
and 6-minute ones. For daily analysis, we used only SAs because we aimed to investigate the stability of the signal during
the daytime and not to compare it with some reference. Conversely, for a 6-minute analysis, it was important to have
two instruments connected in parallel to work in zero-span
mode (required by the extrapolation technique and technical
standard) and channel power mode (used as reference measurement). In this last case, a CP-SA was adopted to measure
channel power, obtained by integration of power spectral density on the operator downlink bandwidth. To deal with it,
once fixed, the integration bandwidth (10 MHz in the case of
the 800 MHz carrier and 20 MHz in the case of the 1800 MHz
carrier), a resolution bandwidth (RBW) equal to 300 kHz and
video bandwidth (VBW) equal to 3 MHz were automatically
set by the instrument. On the other hand, according to instrument settings required by [4], to work in zero-span mode, the
SAs were tuned with a RBW equal to 1 MHz and VBW equal to
equal to 3 MHz. Finally, both instruments used RMS as a detector and the MaxHold as Trace Mode. The whole measurement
chain introduces a 95% extended uncertainty equal to 2.5 dB.

LTE Narrowband Measurements: Experimental
Results
In Fig. 2, the daily results, reported for a representative day,
are depicted for all operators and for the 800 MHz-band. The
E-Field is obtained through the measurement of the received
power and by taking into account the antenna factor, the cable attenuation, and the input impedance of the measuring

IEEE Instrumentation & Measurement Magazine 7

Instrumentation & Measurement Magazine 23-9

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 23-9