Propagation Characterization of an Office Building in
the 60 GHz Band
Jonathan Lu1, Daniel Steinbach2, Patrick Cabrol2, Phil Pietraski2 and Ravikumar V. Pragada2
1 Polytechnic Institute of New York University, USA, lushiaoen@gmail.com
2 InterDigital Communications, LLC, USA, Daniel.Steinbach@interdigital.com, Patrick.Cabrol@interdigital.com,
philip.pietraski@interdigital.com and Ravikumar.Pragada@interdigital.com
Abstract—This work investigates the propagation
characteristics of an office building in the 60 GHz band. From
reflection and scattering measurements of several painted and
un-painted common building materials recorded at 60 GHz, the
complex permittivity and Lambert’s Law scattering coefficient of
each material are extracted. Diffraction measurements at a
building corner at 60 GHz are also presented and analyzed.
Lastly, power angular profiles of building penetration and
scattering at 60 GHz are presented and used to characterize the
outdoor to indoor propagation, and the significant scatterers on
a building surface, respectively.
Index Terms—60 GHz, Buildings, Diffraction, Millimeter
Wave, mmW, NLOS, Penetration, Reflection, Scattering, Urban.
I. INTRODUCTION
Many cellular, back-haul and peer-to-peer applications [1]-
[3] have been proposed for the 57–64 GHz millimeter wave
(mmW) unlicensed band, also known as 60 GHz band. These
applications will utilize directive antenna arrays at the
transmitter and receiver to perform beamforming and achieve
multi-gigabit data rates. To assist in the simulation and
planning of such applications, a measurement campaign was
carried out to investigate various wave propagation
interactions with an office building.
Reflection and scattering measurements at 60 GHz of
several painted and un-painted building materials are presented
in this work and then used to extract each material’s complex
permittivity and scattering coefficient. Past works [4]-[9] have
mostly focused on reflection and transmission characteristics
of various building materials at 60 GHz and have largely
neglected their rough surface scattering characteristics which
cause angular dispersion in the received signal. A scattering
profile from a brick at 60 GHz was given in [8] and empirical
investigations of scattering select materials at higher
frequencies (69-74 GHz, 94 GHz) were presented in [10], [11].
For outdoor applications, the presence of a building can
cause shadowing or conversely, provide an alternate ray path
to beamform along. Limited work has been performed for 60
GHz band. Past works for corner diffraction have focused on
cellular applications at 28 [12] and 40 GHz [13], and
diffraction from metal and wooden wedges at 60 GHz [14].
Two sets of building corner diffraction measurements [15] are
presented and then compared to the absorbing screen
diffraction model.
Figure 1. Block diagram of 60 GHz measurement setup.
In addition to the diffraction from building corners, the
columns, window frames, etc. on building surfaces are
expected to give significant contributions to the received
signal. Knowing where the pertinent scatterers on the surface
are, can speedup beam-finding times and enhance
performance. Our findings through a set of measurements
show that the majority of power arrives in the horizontal plane
containing the transmitting and receiving antennas.
For femtocell applications in which the transmitter is inside
a building and receiver is outside, or vice versa, understanding
the propagation into and out of buildings is paramount.
Limited work was found in the literature. Most works focus on
transmission through uniform slabs of building material [9].
We will present angular power profile measurements of a radio
link where one end of the link is located in an office building
near and far from the windows.
The organization of the paper is as follow. The reflection
and scattering properties of select building materials are
characterized in Section II. Measurements of corner diffraction
and building scattering are presented and analyzed in Section
III. Lastly, in Section IV, outdoor to indoor propagation
measurements are presented.
II. BUILDING MATERIAL MEASUREMENTS
A. Measurement Equipment and Setup
The measurement equipment setup illustrated by the block
diagram in Fig. 1 was used to record the received power at 60
GHz. In the individual measurement scenarios, the antennas,
their heights and positioning systems varied. To model the
effect of antenna arrays at the transmitter (TX) and receiver
(RX), the measurements used directive vertically-polarized V-
band horn antennas and lens antennas with 24 and 34 dBi gain,
and 11 degree and 4 degree 3-dB-beamwidths, respectively.
SMF100A
10 GHz
SMZ90
(w 25dB pad)
LNA
FSQ26
Harmonic
Mixer
Vector Signal
AnalyzerV-Band Horn
N12-3387
Gain = 20dB
NF = 4dB
Input Range:
10-15GHz
+7dBm
7.4GHz – 14.9GHz
x6
WR12WR12
FS-Z9024dBi
V-Band
Horn
24dBi
Multiplier
Signal
Generator
The max EIRPs of the transmitter were 15.6 and 25.6 dBm,
respectively. For more details on the setup refer to [15].
The measurement testbed used for our small-scale building
model measurements [16] was used to perform the reflection
and scattering measurements on the building materials listed in
Table I. To perform scattering measurements for an incident
angle θi, the transmit antenna was moved to the desired θi at
distance ri from the focal point of the setup, while the receive
antenna was moved in an arc of constant distance rs from the
focal point, so that θs spanned a desired range of angles as
shown in Fig. 2(a). To perform reflection measurements, both
antennas were moved so that θs = θi.
All power measurements are normalized to the received
power from a large metal sheet when θs = θi. These relative
measurements are called the Normalized Power (NP). Note
that from image theory, the received power from a large metal
sheet is equivalent to the free-space power for the antenna
separation rs + ri = 1 + 1 = 2 m. For reflection measurements,
NP is approximately equal to the magnitude squared of the
reflection coefficient |Γslab|2.
B. Reflection Measurements and Relative Permittivities
Reflection coefficient |Γslab| measurements were recorded at
60 GHz for the building materials listed in Table I. As an
example, measurements for a ceiling tile and a concrete block
(CMU) wall are scatter plotted versus reflection angle θr = θi =
θs in Figs. 3(a)-(b). To extract the complex relative permittivity
εr of each material from the measurements, we utilize the same
approach used in [4]-[7] where the material under test is
treated as an infinite homogeneous slab. Using this
assumption, the reflection coefficient Γslab can be expressed as
2
,2 2
,
1 ( , )
1 ( , )
j
slab TE TM r rj
TE TM r r
e
e
δ
δ ε θε θ
−
−
−
Γ = Γ
− Γ
, (1)
where 𝛿 = !!!! 𝜀! − 𝑠𝑖𝑛!(𝜃!), λ = 5 mm is the wavelength
and ∆ is the thickness of the slab in meters. ΓTE,TM (εr, θr) is
either the transverse electric (TE) or transverse magnetic (TM)
Fresnel reflection depending on the polarization of the incident
wave relative to the slab. For our measurements, the incident
wave was TE polarized.
The estimated complex relative permittivity εr was then
found from minimizing the prediction error from (3) to
measurements. The estimated εr for all materials are listed in
Table I. Example predictions for the ceiling tile and CMU wall
using the estimated εr are also plotted in blue in Figs. 3(a)-(b).
From these figures, it is seen that the homogeneous slab model
gives good prediction of a materials Γslab. This is also true for
the other materials listed in Table I.
To investigate the effect of paint, measurements were
recorded for painted and unpainted drywall. The semi-gloss
paint with primer appeared to affect the estimated εr of drywall
as seen in Table I. While the flat-paint which was not applied
with primer, did not appear to have any effect. It remains to be
determined whether the thickness of the dried paint and/or the
composition of the paints caused this change.
(a)
(b)
Figure 2. (a) Building material testbed , (b) horn antenna on
pan-tilt motor.
TABLE I. BUILDING MATERIAL PROPERTIES
Building
Material
Thickness
∆
[mm]
Complex
Permitivitty
εr
Attenuation
α
[dB/cm]
Scattering
Coefficient
γ
[dB]
Cement
Cinderblock
Wall
193.7 3.3
-‐
j0.38 11.3 -‐17.1
Brick
Wall 92 2.55
-‐
j0.43 14.7 -‐17.3
Cement
Backerboard
12.7 3.0
-‐
j0.54 17 -‐
Drywall 12.7 2.26
-‐
j2.4*10-‐3 0.09 -‐12.5
Drywall
with
semigloss
paint
12.8 2.77
-‐
j1.9*10-‐2 0.6 -‐
Drywall
with
flat
paint
12.8 2.28
-‐
j2.4*10-‐3 0.09 -‐14.2
Plexiglass 2.2 2.70
-‐
j0.26 8.6 -‐
Glass 12.7 6.55
-‐
j0.20 4.3 -‐
Wood 5.1 2.8
-‐
j4.0*10-‐2 1.3 -‐
Ceiling
Tile 15.9 1.55
-‐
j2.6*10-‐2 1.12 -‐11.9
(a)
(b)
Figure 3. Reflection coefficient magnitude |Γslab| measurements
and predictions for (a) ceiling tile and (b) CMU wall.
0 10 20 30 40 50 60 70 80 90
0
0.2
0.4
0.6
0.8
1
θr [degrees]
| Γ
sla
b |
Measurement
Theory
0 10 20 30 40 50 60 70 80 90
0
0.2
0.4
0.6
0.8
1
θr [degrees]
| Γ
sla
b |
Measurement
Theory
Also listed in Table I is the attenuation parameter α of each
material which was calculated from its εr [6]. α is the rate of
absorption loss incurred by propagating through a material.
Note that this loss does not include the transmission and
reflection loss at the material-air boundaries. The CMU and
brick walls commonly used in exterior walls were both found
to have very high attenuation with α > 11 dB/cm. While glass
was found to have much lower attenuation with α = 4 dB/cm.
Accounting for the smaller thickness of glass, these results
suggest that dominant contribution from waves propagating
into and out of buildings are those that propagate through
windows rather than walls.
For interior walls, drywall has a small α which promotes
penetration, while cement backerboard, which is commonly
used as a substrate for tiles, has very high attenuation. This
suggests the feasibility of non-line of sight (NLOS)
communication between certain rooms separated by drywall,
though a more complete study which includes the inner
structure (eg., multi-layers [4], insulation and studs) should be
performed.
C. Scattering Measurements and Lambertian Model
Scattering measurements at 60 GHz were recorded while
the transmitting antenna at a fixed θi, while the receiving
antenna was moved over a range of θs. An example for a CMU
wall with θi = 45° is shown in Fig. 4 where the normalized
power NP in dB is plotted versus θs. Also plotted in blue is the
power contribution from a single reflected ray using ϵr = 3.3 –
j0.38 from Table I. In Fig. 4, the scattering is seen to be
dominant in the region θs < 20°, where the measured power is
greater than that predicted by reflection. Note that the shape of
the blue curve is partially dependent on the antennas used, so
that for a set of more directive antennas, the blue curve would
have a narrower shape.
To predict the average total scattered power for the θs < 20°
region, the surface was segmented into K patches each of area
A m2. The bi-static scattering equation [17] was then applied to
each patch assuming a Lambert’s Law model (σ = γ cos(θi)
cos(θs)). The average power <Pr> assuming each patch had
uniformly random phase, was then the power sum over all
patches. Thus,
2
3 2 2
1(4 )
K
r t
k kTX kRX
AP P
R R
λ σ
π =
= ∑ . (2)
where Pt is the transmitted power in watts, RkTX is the distance
in meters from the TX antenna to the kth slab and RkRX is the
distance from the kth slab to the RX antenna.
To extract γ, a 0-watt-error criterion between predicted
scattered power and measurements was enforced. Note that, a
locally ergodic assumption was used where the variation
caused by different realizations of the random surface was
approximated by the local variation in θs of the received signal.
For the example in Fig. 4, a value of γ = 0.02 was
computed and used to plot the green curve in Fig. 4. The level
of the scattered contribution relative to the reflected
contribution in and near the specular direction (θs = θi = 45°)
suggested that the rough scattering contribution (2) was
relatively weaker to the reflected contribution. This resulted in
very little angular dispersion in and near the specular direction.
Similar results were found for the other materials in Table I.
Note that the ceiling tile has the greatest γ value in Table I.
This was because the size of the ceiling tile was not large
relative to the antenna pattern footprint, so the diffraction from
the edges was significant. All other materials in Table I were
large enough so that the edge effects were negligible.
III. BUILDING CORNER DIFFRACTION AND
SURFACE SCATTERING
The TX antenna and RX antenna were each placed on push
carts for the building corner diffraction, scattering and
penetration measurements. The pointing of the antennas was
performed using pan-tilt motors shown in Fig. 2(b) that have
0.0032⁰ angle resolution.
A. Corner Diffraction
Diffraction coefficient magnitude |D| measurements for
two different corners were recorded when the RX antenna was
in the shadowed region (θ < 0° in Fig. 5) of each corner [15].
The measurements are scatter plotted in Fig. 6.
Figure 4. Scattering measurements from a CMU wall (ϵr = 3.3
– j0.38) and simulated reflection and scattering curves.
Figure 5. Diffraction at a building corner.
Figure 6. Comparison of diffraction coefficient |D|
measurements and absorbing screen diffraction coefficient.
0 20 40 60 80 100
-60
-50
-40
-30
-20
-10
0
Scattering Angle θs [degrees]
No
rm
ali
ze
d
Po
we
r N
P
[d
B]
Measurements
Reflected Contribtuion
Scattered Contribution
-70 -60 -50 -40 -30 -20 -10 0
-40
-30
-20
-10
0
Di
ffr
ac
tio
n
Co
ef
fic
ien
t |
D(θ
)|
[d
B]
θ [deg]
Corner 1 Measurements
Corner 2 Measurements
Abosrbing Screen
To model the angular dependence of the diffraction, we
compared these measurements to predictions from a semi-
infinite absorbing screen with a knife edge corner whose D is
( ) 2 1 cos( ) 2 2
2sin( )2
S S SD f jg
k
π θ
θ
θ π ππ
⎡ ⎤⎛ ⎞ ⎛ ⎞+
= − +⎢ ⎥⎜ ⎟ ⎜ ⎟⎜ ⎟ ⎜ ⎟⎢ ⎥⎝ ⎠ ⎝ ⎠⎣ ⎦
. (3)
Here θ is the angle the diffracted ray makes with the shadow
boundary as seen in Fig. 5, and f, g, and S are defined in [18].
The absorbing screen predictions are plotted as the dotted line
in Fig. 6 and are seen to give good comparison to the
measurements.
As an example, consider an urban scenario in which the TX
and RX are located on adjacent sides of a building, so that θ ≈
90°. From the measurements and predictions in Fig. 6, corner
diffraction involves a large loss with |D| ~ -40 dB. Thus, it is
the scattering from the clutter in the street (e.g., cars and
lamppost) which will provide the dominant contribution to the
received signal [15].
Conversely, for the case in which the TX and RX are
located on the same street and the line-of-sight (LOS) path is
blocked (e.g., humans [15], etc.), the TX and RX can use a
corner diffracted ray to beamform along. Though, only
shallowly diffracted ray paths (θ < 10°) should be considered
because of the large loss. Note that due to the symmetrical
property (3), |D(θ = 10°)| ≈ -20° as seen in Fig. 6.
B. Building Surface Scattering
Building scattering measurements were recorded using the
setup depicted in Fig. 7. The TX and RX antennas were
pointed at different portions of the wall using the pan-tilt
antennas. The received power from the point of view of the
RX antenna normalized relative to the LOS power for 7.5 m is
plotted in Fig. 8. The majority of power was observed to arrive
in the horizontal plane (elevation angle = 0°) containing the
TX and RX antennas at 0°, 25° and 40°. The 0° and 40° humps
correspond to the vertical columns on the building surface
shown in Fig. 7. The specularly reflected contribution
corresponds to the hump at 25°. Other humps outside of the
horizontal plane correspond to features on the building surface
such as window sills. These 60 GHz results are similar to UHF
band measurements [19], in which features on the building
surface such as balconies, contributed significantly to the
received power angle profile.
IV. BUILDING PENETRATION
To investigate the propagation into and out of buildings at
60 GHz, two received power angular profiles were recorded
when the TX was located at ground level outside a building
while the RX was located on the third floor of the building at
the two locations illustrated in Fig. 9. The first location
denoted by RX1 in Fig. 9 was near a window, so the RX
antenna had LOS with the TX antenna. The second RX
location was away from the window as shown in Fig. 9, so that
the TX and RX antennas were NLOS and there were 3 office
cubicles separating the RX antenna and the nearest window.
The power angular profile recorded when the RX antenna
was located near the window is plotted in Fig. 10(a). There are
three dominant contributions seen in the profile. The largest
contribution is the direct contribution (5° Azimuth, -15°
Elevation) that penetrates the window. Lesser yet still
significant contributions were found to scatter off the top and
sides of the window frame. The profile recorded when the RX
antenna was located further away from the window is plotted
in Fig. 10(b). For this case, the dominant contributions arrive
in the vertical plane (VP) containing the TX and RX antennas.
The dominant contribution (5°, 20°) was found to propagate
through the window bisected by the VP and then reflect/scatter
from the ceiling. A smaller contribution (5°, -5°) diffracted
from the window frame and then was further attenuated by the
clutter in the environment before reaching the RX antenna.
Note that in both RX locations, other arrivals diffracted from
other windows away from the VP or those penetrating the
walls were found to be severely attenuated and did not affect
the angular dispersion of the received signal.
Figure 7. Building scattering measurement setup
Figure 8. Scattering profile from a building surface.
Figure 9. Outdoor to indoor measurement setup for RX
antenna located near to (RX1) and away from (RX2) window.
V. CONCLUSION
In this work, several different sets of measurements were
presented and analyzed. The reflection and scattering
properties of a variety of building materials were extracted
from measurements. Semigloss paint with primer was found
to have an effect on the relative permittivity of a material.
Outer wall building materials were seen to have very high
attenuation coefficients, so that building penetration is
expected to be least attenuated through windows. Rough
surface scattering from the considered building materials was
found to have little effect on the angular dispersion of the
received signal.
Diffraction measurements were recorded for two building
corners. The absorbing screen model was found to model
relatively well the angular dependence of the diffraction. For
some cases, the diffraction loss for turning corners was found
to be very large, so that the scattering form street furniture
(e.g., lamposts) was expected to be dominant.
From a building scattering profile of an office building, the
dominant contributions to received power were found to
arrive mostly in the horizontal plane containing the TX and
RX antennas. Several lesser out-of-plane contributions were
found to come from window ledges.
For propagation into a building, when the TX and RX
antennas have LOS, the dominant arrival was the direct ray.
Though scattering from the window frame gave significant
contributions to the received signal. When the RX was located
away from the window, the dominant contribution was ceiling
diffracted and lied in the vertical plane.
(b)
(a)
Figure 10. Power angular profiles when RX is located on the
third floor (a) near to; (b) away from the window
REFERENCES
[1] R. C Daniels, J. N. Murdock, T. S. Rappaport and R. W. Heath Jr.,
“60GHz Wireless: Up close and Personal”, IEEE Microwave Mag., Dec.
2010, pp. S44-S50,.
[2] 802.11ad-2012- IEEE Standard for Information technology-Tele-
communications and information exchange between systems-Local and
metropolitan area networks--Specific requirements-Part 11: Wireless
LAN Medium Access Control (MAC) and Physical Layer (PHY)
Specifications Amendment 3: Enhancements for Very High Throughput
in the 60 GHz Band
[3] M. R. Akdeniz, Y. Liu, S. Rangan, and E. Erkip, “Millimeter wave
picocellular system evaluation for urban deployments,” arXiv preprint,
Mar. 2013.
[4] J. Ahmadi-Shokouh, S. Noghanian, E. Hossain, M. Ostadrahimi, J.
Dietrich, “Reflection coefficient measurement for house flooring
materials at 57-64 GHz,” in IEEE Global Tel. Conf., Nov. 2009.
[5] K. Sato, T. Manabe, T. Ihara, H. Saito, S. Ito, T. Tanaka, K. Sugai, N.
Ohmi, Y. Murakami, M. Shibayama, Y. Konishi and T. Kimura,
“Measurements of reflection and transmission characteristics of interior
structures of office building in the 60-GHz band,” IEEE Trans. on
Antenna and Propagation, Vol. 45, No. 12, pp. 1783-1792, Dec. 1997.
[6] L. M. Correa and P. O. Frances, “Estimation of Materials
Characteristics from Power Measurements,” in Proc. IEEE Int. Symp.
on Personal Indoor and Mobile Comm. (PIRMC), pp. 510-513, 1994.
[7] J. Lahteenmaki and T. Karttaavi, “Measurement of dielectric parameters
of wall materials at 60 GHz band,” Electron. Lett., vol. 32, no. 16, pp.
1442–1444, Aug. 1996.
[8] B. Langen, G. Lober, and W. Herzig, “Reflection and transmission
behaviour of building materials at 60 GHz” in Proc. IEEE Int. Symp. on
Personal Indoor and Mobile Comm. (PIRMC), pp. 505-509, Sept. 1994.
[9] C. Anderson and T. Rappaport, “In-building wideband partition loss
measurements at 2.5 and 60 ghz,” IEEE Trans. Wireless Comm., vol. 3,
no. 3, pp. 922–928, May 2004.
[10] M. Kyro, V. Semkin, and V. Kolmonen, “Empirical characterization of
scattering pattern of built surfaces at mm-wave frequencies,” in Proc. of
European Conf. on Ant. and Prop. (EuCAP), pp. 112-115, April 2013.
[11] Li, E.S., and K. Sarabandi, "Low Grazing Incidence Millimeter-Wave
Scattering Models and Measurements for Various Road Surfaces", IEEE
Trans. on Ant. and Prop., Vol. 47, No. 5, pp. 246-253, May 1999.
[12] P. A. Tenerelli "Diffraction by building corners at 28 GHz:
Measurements and modeling", in Proc. IEEE Int. Symp. Pers., Indoor
Mobile Radio Commun., Vol. 3, pp. 1166 -1169, Sep. 1998
[13] A. Alejos, M. Sanchez, and I. Cuinas, “Measurement and analysis of
propagation mechanisms at 40 GHz: Viability of site shielding forced
by obstacles,” IEEE Trans. Vehicular Technology, Vol. 57, No. 6, pp.
3369–3380, Nov. 2008.
[14] M. Jacob, S. Priebe, R. Dickhoff, T. Kleine-Ostmann, T. Schrader, T.
Kürner, “Diffraction in mm and sub-mm Wave Indoor Propagation
Channels”, in IEEE Trans. on Microwave Theory and Techniques, Vol.
60, No. 3, pp.833-844, Mar. 2012.
[15] J. S. Lu, P. Cabrol, D. Steinbach, and R. V. Pragada, “Measurement and
Characterization of Various Outdoor 60 GHz Diffracted and Scattered
Paths,” in Proc. of Military Comm. Conf. (MILCOM), San Diego, C.A.,
Nov. 2013.
[16] J. S. Lu, H. L. Bertoni, P. Cabrol and D. Steinbach, “60 GHz
Investigation of Building Scattering at 2 GHz Using a Scale Model,”
IEEE APS-URSI Symp., Orlando, FL, USA, July 7-13, 2013.
[17] E. F. Knott, J. F. Shaeffer and M. T. Tuley, Radar Cross Section, Artech
House London, 1993
[18] H. L. Bertoni, Radio Propagation for Modern Wireless Applications,
Upper Saddle River, NJ: Prentice Hall, PTR, 2000.
[19] M.R.J.A.E. Kwakkernaat and M.H.A.J. Herben, “Diagnostic analysis of
radio propagation in UMTS networks using high-resolution angle-of-
arrival measurements,” in IEEE Antennas and Propagation Mag., Vol.
53, No. 1, pp. 66-75, Feb. 2011.
