ELE 451: Wireless Communications

Mobile Radio Propagation: Large

Scale Path Loss

Dr. Mahmoud H. Ismail Ibrahim

Professor
Department of Electrical Engineering
American University of Sharjah
ACKNOWLEDGMENT

Lots of the course contents presented here are

based on lecture notes provided by Prof.
Mohamed El-Tarhuni. His help is gratefully
acknowledged.

08-Mar-23 2
INTRODUCTION
 The mobile radio channel has a significant effect on the
performance of wireless communication systems. The
transmission path varies from a simple line-of-sight to a severely
obstructed path.
 Radio channels are extremely random as opposed to wired
channels which are stationary and predictable.
 Modeling of radio channels has been based on statistical and
measurements characterization.
 Modeling of radio propagation for cellular networks was
investigated in the 1970s.
 In the 1980s, modeling of indoor radio propagation for cordless
telephony and wireless LAN applications.
 Currently, modeling of the wireless channel at the 28 and 60 GHz
range (mmWave) is being investigated.
08-Mar-23 3
WHY RADIO PROPAGATION STUDIES?
 For appropriate design, deployment, and management of
wireless networks.
 Radio propagation is heavily site-specific and can vary
significantly depending on
 Terrain
 Frequency of operation
 Velocity of the mobile terminal
 Accurate characterization of the radio channel is important for
 Predicting signal coverage and achievable data rates
 Performance of alternative signaling and reception schemes
 Interference analysis from different systems
 Determining optimum location of base station antennas

08-Mar-23 4
CHALLENGES IN RADIO PROPAGATION STUDIES
 Radio propagation is different in different cell types
 Femto-, pico-, micro-, and macrocells.
 Radio propagation in open areas is different from radio
propagation in indoor and urban areas.
 Open area
The signal strength falls with the square of the distance.
There is a line-of-sight (LOS) path between Tx and Rx.
 In indoor and urban areas
The signal strength falls at a much higher rate.
Non line-of-sight (NLOS) propagation and the signal is
carried by a multiplicity of indirect paths (called
multipath).
 Radio propagation is affected by the frequency of operation.
08-Mar-23 5
RADIO PROPAGATION MECHANISMS
 Three basic mechanisms are considered to describe radio
propagation: Reflection, Diffraction and Scattering.

08-Mar-23 6
RADIO PROPAGATION MECHANISMS (CONTD.)
 Reflection
 Electromagnetic waves impinge on obstructions larger than
the wavelength.
 Rays incident upon the ground, walls of buildings, the
ceiling, and the floor.
 Diffraction
 Rays incident upon edges of buildings, walls, and other large
objects can act as a secondary source.
 Scattering
 Caused by irregular objects such as walls with rough
surfaces and furniture and vehicles especially when the
object dimensions are close to or less than the wavelength.
08-Mar-23 7
PROPAGATION MODELS
 Propagation models focus on predicting the average received
signal strength at a given distance from the transmitter, as well as
the variability of the signal strength in close spatial proximity to a
particular location.
 Propagation models that predict the mean signal strength for an
arbitrary transmitter-receiver (T-R) separation distance are useful
in estimating the radio coverage area of a transmitter and are
called large-scale propagation or path loss models, since they
characterize signal strength over large T-R separation distances
(several hundreds or thousands of meters).
 On the other hand, propagation models that characterize the
rapid fluctuations of the received signal strength over very
short distances (few wavelengths) or few seconds are called small
scale or fading models.
08-Mar-23 8
PROPAGATION MODELS (CONTD.)
Small-scale and large-scale fading

08-Mar-23 9
LARGE SCALE (PATH LOSS) PROPAGATION MODELS

 The path-loss model relates the loss of signal strength to the

distance between two terminals  estimates signal coverage.
 Since a wireless network spans a variety of environments, a
unique channel model cannot describe the radio propagation
in all environments.
 The following cases will be considered
 Free space propagation.
 Two-Ray model.
 Log-Normal Shadowing.
 Empirical Path loss models for macrocells, microcells,
picocells, and femtocells.
08-Mar-23 10
1. FREE SPACE PROPAGATION
 In general, the radio signal strength falls as a power n of the
distance, called the path-loss exponent or path-loss gradient.
 Let the transmitted power be Pt, then after a distance d meters,
the signal strength will be proportional to 𝑃𝑡 𝑑−𝑛
 In free space, n = 2.
 The received signal power at a distance d is given by
2
𝜆
𝑃𝑟 (𝑑) = 𝑃𝑡 𝐺𝑡 𝐺𝑟
4𝜋𝑑
where 𝐺𝑡 and 𝐺𝑟 are the transmitter and receiver antenna
gains in the direction from Tx to Rx, and 𝜆 = 𝑐/𝑓 is the
wavelength of the carrier (c = 3 x 108 m/s and f is the
frequency). This is also called the Range Equation.
08-Mar-23 11
1. FREE SPACE PROPAGATION (CONTD.)
 The quantity 𝑃𝑡 𝐺𝑡 is called the effective isotropic radiated power
(EIRP).
 Let the received signal strength at distance d0 (used as a reference
for finding the power at distance d) be 𝑃𝑟 (𝑑0 ):
2
𝜆
∴ 𝑃𝑟 (𝑑0 ) = 𝑃𝑡 𝐺𝑡 𝐺𝑟
4𝜋𝑑0
 The distance d0 must lie in the far field of the antenna given by
𝑑𝑓 = 2𝐷 2 /𝜆 where D is the largest physical dimension of the
antenna so we need to have 𝑑0 ≥ 𝑑𝑓 .
 The received power at distance d in terms of that at 𝑑0 is
2
𝑑0
𝑃𝑟 𝑑 = 𝑃𝑟 (𝑑0 )
𝑑
𝑑
in dB: 𝑃𝑟 𝑑 (dB) = 𝑃𝑟 𝑑0 (dB) − 𝟐𝟎 log 𝑑0
.
08-Mar-23 12
1. FREE SPACE PROPAGATION (CONTD.)
The free space path loss is clearly given by
2
𝑃𝑡 1 4𝜋𝑑
𝐿𝑝 (𝑑) = =
𝑃𝑟 (𝑑) 𝐺𝑡 𝐺𝑟 𝜆
In dB, and putting the distance in km and the frequency in MHz,
we have (NOTICE THE UNITS)

𝐿𝑝 (𝑑)(dB)
= 32.45 + 20 log 𝑓𝑐 𝑀𝐻𝑧 + 𝟐𝟎 log 𝑑 𝑘𝑚 − 10 log 𝐺𝑡 − 10 log 𝐺𝑟

 There is a 20 dB/decade loss in signal strength as a function of

distance in free space.
𝑑
 The transmission delay is 𝜏 = 𝑐 = 3.3𝑑 nsec, i.e., 3.3 nsec per
meter.
08-Mar-23 13
1. FREE SPACE PROPAGATION (CONTD.)

08-Mar-23 14
1. FREE SPACE PROPAGATION (CONTD.)
 Let us solve the examples on page 29 and 32 in the textbook.
 It is also possible to link the free space path loss at any
arbitrary distance 𝑑 to that at a reference distance 𝑑0 as
follows:
𝐿𝑝 𝑑 dB = 𝑃𝑡 dB − 𝑃𝑟 𝑑 dB
𝑑
= 𝑃𝑡 dB − 𝑃𝑟 𝑑0 dB + 𝟐𝟎 log
𝑑0
𝑑
= 𝐿𝑝 𝑑0 dB + 𝟐𝟎 log
𝑑0
 Note that 𝑃𝑇 and 𝑃𝑟 can be expressed in dBm instead of dB and
the conversion is: 𝑃 dBm = 𝑃 dB + 30.
 However, the path loss can ONLY be expressed in dB since it is
a ratio between powers and is thus unitless!
08-Mar-23 15
1. FREE SPACE PROPAGATION (CONTD.)
 Example: What is the received power (in dB and dBm) in the
free space of a signal whose transmit power is 1 W and carrier
frequency is 2.4 GHz if the receiver is 1 mile from the
transmitter? Assume that the antenna gains are 1.6. What is
the path loss in dB? What is the transmission delay in ns?
 Solution:
𝑐
𝜆 = = 0.125 m
𝑓
2
0.125
𝑃𝑟 = 10 log 1 × 1.6 × 1.6 × = −100 dB = −70 dBm
4𝜋 × 1600
 The path loss is given by the difference between 10 log(𝑃𝑡)
and 10 log(𝑃𝑟) (where both are in dB or dBm). This is 100 dB.
 The transmission delay is 3.3 x 1600 = 5280 ns = 5.28 µs.
08-Mar-23 16
1. FREE SPACE PROPAGATION (CONTD.)
 The receiver sensitivity 𝑃𝑠𝑒𝑛𝑠 is the minimum signal strength
that can be detected and decoded properly by the receiver.
 Given 𝑃𝑠𝑒𝑛𝑠 , we can calculate the maximum range (coverage) of
the data link for free-space propagation
𝑃𝑡 𝐺𝑡 𝐺𝑟 𝜆2
𝑑𝑚𝑎𝑥 =
4𝜋 2 𝑃𝑠𝑒𝑛𝑠
 Example (textbook p. 33): Suppose a transmitting base station
emits an EIRP of 100 W. The handset receiver has a sensitivity
of 𝑃𝑠𝑒𝑛𝑠 = −90 dBm. What is the maximum range for the
mobile unit given the following selected antennas: a dipole
having gain of 2.15 dB, and an omnidirectional having gain of 0
dB? The operating frequency is 860 MHz. Neglect system
losses.
08-Mar-23 17
2. TWO-RAY MODEL FOR MOBILE RADIO ENVIRONMENTS

 In free space, the signal travels from Tx to Rx along a single

path.
 In all realistic environments, the signal reaches the receiver
through several paths. Thus, the free space model does not
apply.
 The two-path or two-ray model is used for modeling the land-
mobile radio.
08-Mar-23 18
2. TWO-RAY MODEL FOR MOBILE RADIO ENVIRONMENTS (CONTD.)
 Let the base station (BS) antenna height be hb and the mobile
station (MS) height be hm above the earth.
 The earth is assumed as a flat surface between the BS and MS.
 There is a LOS component which carries the signal as in free
space and there is also another path that results from signal
reflection form the flat surface of the earth.
 The two signals travel different distances and they may add
either constructively or destructively at the receiver.
 The received signal strength for the two-ray model is
approximated as:
ℎ𝑏2 ℎ𝑚
2
𝑃𝑟 (𝑑) = 𝑃𝑡 𝐺𝑡 𝐺𝑟 4 , 𝑑 ≫ ℎ𝑏 ℎ𝑚
𝑑
08-Mar-23 19
2. TWO-RAY MODEL FOR MOBILE RADIO ENVIRONMENTS (CONTD.)
 The signal strength falls as the fourth power of the distance,
i.e., 40 dB per decade or 12 dB per octave.
 The path loss in dB is

𝐿𝑝 (𝑑)(dB)
= 𝟒𝟎 log 𝑑 − 20 log ℎ𝑏 − 20 log ℎ𝑚 − 10 log 𝐺𝑡 − 10 log 𝐺𝑟

 The two-Ray model is good for systems with tall towers (>50
m) and for LOS microcell channels in urban environments.
 Let us solve the example on page 84 in the textbook.

08-Mar-23 20
GENERAL DISTANCE-POWER RELATIONSHIP
In general, the received signal power is related to the distance
such that
𝑛
𝑑0 𝑑
𝑃𝑟 𝑑 = 𝑃𝑟 𝑑0 ⇒ 𝑃𝑟 𝑑 dB = 𝑃𝑟 𝑑0 dB − 10𝑛 log
𝑑 𝑑0
where 𝑛 is the path loss exponent.
 For indoor and urban radio channels, the distance-power
relationship will change with
 Building and street layouts
 Construction material
 Density and height of buildings in the area
 In general, variations in 𝑛 in different outdoor areas are smaller
than variations observed in indoor areas.
08-Mar-23 21
GENERAL DISTANCE-POWER RELATIONSHIP (CONTD.)
 Measurements of indoor channels show values of 𝑛 less than 2
for corridors or large open indoor areas and values as high as 6
for metallic buildings.
 Example: (Coverage with a general path loss exponent) What
is the coverage of a base station that transmits a 2 kW signal
given that the receiver sensitivity is –100 dBm, the path loss at
the first meter is 32 dB, and the path loss exponent is 𝑛 = 4?
 Solution: 10 log 𝑃𝑡 = 10 log 2000 ∗ 103 = 63 dBm. The receiver
sensitivity is 10 log 𝑃𝑟 = −100 dBm. The total path loss that is
allowed is 10 log 𝑃𝑡 − 10 log 𝑃𝑟 = 63 − −100 = 163 dB. There is a
32 dB loss at the first meter, then the loss due to the distance is
at most 163 – 32 = 131 dB. Thus, 10𝑛 log(𝑑) = 131  𝑑 = 1883
meters. The coverage of the cell is 1.883 km
08-Mar-23 22
MEASUREMENTS OF THE PATH LOSS EXPONENT
 The receiver is fixed at one location.
 The transmitter is placed at a number of locations with
different distances between Tx and Rx.
 The received power or the path loss in dB is plotted against
the distance on a logarithmic scale.
 The slope of the best-fit line through the measurements is
taken as the path loss exponent n.
 The figure in the next slide shows such a measurement for
an indoor area at distances from 1 to 20 meters, and the
best-fit line.

08-Mar-23 23
MEASUREMENTS OF THE PATH LOSS EXPONENT (CONTD.)

08-Mar-23 24
3. SHADOW FADING
 In practice, the received signal strength for the same
distance from the Tx will be different depending on the exact
location of the Rx.
𝑑
 The relation 𝑃𝑟 𝑑 dB = 𝑃𝑟 𝑑0 dB − 𝟏𝟎𝒏 log 𝑑0
in fact
provides the mean or average value of the signal strength
that can be expected for a distance d between the Tx and Rx.
 The variation of received signal strength due to location is
referred to as shadow fading.
 The fluctuations around the mean value are due to signal
blocking by buildings (in outdoor areas), walls (inside
buildings), and other objects in the environment.
08-Mar-23 25
3. SHADOW FADING (CONTD.)

Measured data and MMSE path loss model for many cities in
08-Mar-23 Germany 26
3. SHADOW FADING (CONTD.)
aN
a1 a3
Tx a2
Rx

 Total attenuation = 𝑎1 . 𝑎2 . 𝑎3 … . 𝑎𝑁
 Total attenuation (in dB) = 10 log 𝑎1 + 10 log 𝑎2 + ⋯ + 10 log 𝑎𝑁 .
 Summing N random variables can be approximated by a random
variable with Gaussian distribution.
 The path loss has to be modified to include this effect by adding a
random component
𝑑
𝐿𝑝 𝑑 = 𝐿𝑝 𝑑0 + 10𝑛 log +𝑋
𝑑0
where X is a zero-mean Gaussian distributed random variable
(in dB) with standard deviation 𝜎𝑋 (also in dB).
08-Mar-23 27
3. SHADOW FADING (CONTD.)
 Measurements indicate that the path loss is a log-normally
distributed random variable with mean value of 𝜇 = 𝐿𝑝 𝑑0 +
𝑑
10𝑛 log and variance 𝜎𝑋2 .
𝑑 0

 The log-normal probability density function (pdf) of the path

loss is given by
2
1 𝑙𝑝 − 𝜇
𝑓𝐿𝑝 𝑙𝑝 = exp −
𝜎𝑋 2𝜋 2𝜎𝑋2

 In practice, 𝜎𝑋2 is estimated from measurement data similar

to the setting used to estimate 𝑛 by measuring the variance
of the received signal strength at different distances.
08-Mar-23 28
DETERMINATION OF PERCENTAGE OF COVERAGE AREA
 Since 𝑃𝑟 𝑑 [𝑑𝐵𝑚] = 𝑃𝑡 𝑑𝐵𝑚 − 𝐿𝑝 (𝑑)[𝑑𝐵] = 𝑃𝑡 𝑑𝐵𝑚 − 𝐿𝑝 𝑑0 −
𝑑
10𝑛 log − 𝑋 = 𝑃𝑟 𝑑 [𝑑𝐵𝑚] − 𝑋 , we see that the received
𝑑 0
power is also log-normally distributed with mean 𝑃𝑟 𝑑 [𝑑𝐵𝑚]
and variance 𝜎𝑋2 .
 The probability that the received signal at distance d is greater
than a certain threshold value 𝛾 is
∞ 2
1 𝑥 − 𝑃𝑟 𝑑
𝑃𝛾 𝑑 = 𝑃 𝑃𝑟 𝑑 > 𝛾 = න exp − 𝑑𝑥
𝛾 𝜎𝑋 2𝜋 2𝜎𝑋2
 Define the Q-function
∞ 𝜁2
1
𝑄 𝑥 = න 𝑒 − 2 𝑑𝜁
2𝜋 𝑥
𝛾−𝑃𝑟 𝑑
 Hence, 𝑃𝛾 𝑑 = 𝑃 𝑃𝑟 𝑑 > 𝛾 = 𝑄 𝜎𝑋
. Note that 𝑄 −𝑥 =
1−𝑄 𝑥 .
08-Mar-23 29
DETERMINATION OF PERCENTAGE OF COVERAGE AREA (CONTD.)
 The probability that the signal falls below the threshold is
1 − 𝑃𝛾 𝑑 and is known as the outage probability.
 Example: Let the average received signal strength at the radius d
be –100 dBm and assume that the Rx sensitivity is –110 dBm. The
measured standard deviation for the log-normal shadowing is 10
dB. The probability that the received signal at d exceeds the
satisfactory level is
−110 − (−100)
𝑃𝛾 𝑑 = 𝑄 = 𝑄 −1 = 1 − 𝑄 1 = 0.84
10
 As seen from the example, due to the random effects of
shadowing, some locations within the coverage area will have a
signal below the desired threshold.
 We would like to relate the boundary coverage to the percentage
of area covered within the boundary.
08-Mar-23 30
DETERMINATION OF PERCENTAGE OF COVERAGE AREA (CONTD.)
 For a circular coverage area of radius R, we are interested in
computing 𝑈(𝛾), the percentage of useful service area where
received signal is equal to or greater than 𝛾, given a known
probability of coverage at the cell boundary.
 It can be shown that
1 1 − 2𝑎𝑏 1 − 𝑎𝑏
𝑈 𝛾 = 1 − erf 𝑎 + exp 1 − erf
2 𝑏2 𝑏
𝛾 − 𝑃𝑡 + 𝐿𝑝 𝑑0 + 10𝑛 log 𝑅/𝑑0 10𝑛 log 𝑒
𝑎= ,𝑏 =
𝜎 2 𝜎 2
erf 𝑥 = 1 − erfc 𝑥 = 1 − 2𝑄 2𝑥
 The above equation may be evaluated for different values of σ and
n as shown in the next figure.
 For example, if n = 4 and σ = 8 dB, a 75% boundary coverage (75%
of the time the signal is to exceed the threshold at the boundary),
then the percentage of coverage area is 94%.
08-Mar-23 31
DETERMINATION OF PERCENTAGE OF COVERAGE AREA (CONTD.)

Probability of coverage at the cell boundary

08-Mar-23 32
DETERMINATION OF PERCENTAGE OF COVERAGE AREA (CONTD.)
 Example: 4 received power measurements at different distances
are summarized in the table below. The path loss model is 𝐿𝑝 𝑑 =
𝑑
𝐿𝑝 𝑑0 + 10𝑛 log + 𝑋, where 𝑑0 = 100 m.
𝑑 0

(a) Find the minimum mean square error (MMSE) estimate of the
path loss exponent n.
(b) Calculate the standard deviation about the mean value.
(c) Estimate Pr at 𝑑 = 2 km using the resulting model.
(d) What is the likelihood that the
received signal at 2 km is > −60 dBm?
(e) Predict the percentage of coverage
area within a 2 km radius cell that
receives signals > −60 dBm with the
probability in (d).
08-Mar-23 33
4. EMPIRICAL PATH LOSS MODELS
 Provide a means for predicting path loss for a particular
application environment.
 Any model has a limited range of applicability and will
provide only an approximate characterization for a specific
propagation environment.
 Most models include effects such as terrain profile, heights of
antennas, carrier frequency, ..etc.
 These models are obtained via extensive measurements in a
number of cities and locations of received signal strength.
 We will be discussing path loss models for macro, micro and
picocells.
08-Mar-23 34
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS
 Macrocells are large cells spanning few kilometers to tens of
kilometers.
 Traditional cells used in cellular telephony base stations.
 Frequency of operation is around 900 MHz and 1800 – 1900
MHz.
(1) Okumura Model:
 Based on measurements done by Okumura in 1968 around
Tokyo for frequency range between 100 MHz and 1,920 MHz,
distances between 1 km and 100 km with a base station
antenna height of 200 m and a mobile station antenna height of
3 m.
 The model has been extrapolated to 3000 MHz.
08-Mar-23 35
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
 Is generally applicable for signal prediction in Urban areas.
 Correction factors are used to adjust the median path loss for
differences in antenna heights, diffraction and terrain profile.
 The median path loss is given by
𝐿 𝑑𝐵 = 𝐿𝑝 + 𝐴𝑚𝑢 𝑓, 𝑑 − 𝐺 ℎ𝐵𝑆 − 𝐺 ℎ𝑀𝑆 − 𝐺𝐴𝑅𝐸𝐴
where
– 𝐿𝑝 is the free space loss,
– 𝐴𝑚𝑢(𝑓, 𝑑) is the median attenuation relative to free space
(given in the form of curves),
– 𝐺(ℎ𝑏𝑠) is the base station height correction factor,
– 𝐺(ℎ𝑚𝑠) is the mobile antenna height correction factor,
– 𝐺𝐴𝑅𝐸𝐴 is the correction factor depending on the type of
environment (open, quasi open, suburban).
08-Mar-23 36
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)

08-Mar-23 37
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
 The antenna correction factors are given by:
ℎ𝐵𝑠
𝐺 ℎ𝐵𝑆 = 20 log , 30 m < ℎ𝐵𝑆 < 1000 m
200
ℎ𝑀𝑆
𝐺 ℎ𝑀𝑆 = 10 log , ℎ𝑀𝑆 ≤ 3 m
3
ℎ𝑀𝑆
𝐺 ℎ𝑀𝑆 = 20 log , 3 m < ℎ𝑀𝑆 ≤ 10 m
3
 Example: Find the median path loss using Okumura’s model
for d = 50 km, hBS = 100 m, hMS = 10 m in a suburban
environment. If the base station transmitter radiates an EIRP
of 1 kW at a carrier frequency of 900 MHz, find the power at
the receiver (assume a unity gain receiving antenna).
08-Mar-23 38
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
(2) Okumura-Hata (Hata) Model:
 Okumura model suffers from being purely empirical and
being based on correction factors from curves, which makes it
difficult to incorporate into computer-based design tools.
 Hata model is an analytical tool that is based on Okumura’s
data. It produces path loss predictions that compare favorably
with those produced by Okumura model.
 Well-suited for computer implementation as an alternative to
Okumura model.
 The general relationship developed by Hata is (for Urban
Areas):

08-Mar-23 39
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
 In the previous equation, fc is in MHz, hb and hm are in meters,
and d is in km (> 1 km), a(hm) is a correction factor for effective
mobile antenna height, which is a function of the size of the
coverage area.

−
08-Mar-23 40
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
 Example:
Determine the path loss of a 900 MHz cellular system operating in
a large city from a base station with the height of 100 m and a
mobile station installed in a vehicle with antenna height of 2 m.
The distance between the mobile and the base station is 4 km.
(3) PCS Extension of Hata Model:
 Hata model has been extended to cover personal
communications systems (PCS) (around 1.8 GHz) with cell radius
of the order of 1 km.
 The European Cooperative for Scientific and Technical research
(EURO-COST) developed a modified formula for Hata model
called the COST-231 model.

08-Mar-23 41
4A. PATH-LOSS MODELS FOR MACROCELLULAR AREAS (CONTD.)
 The median path loss is given by
L p  46.3  33.9 log f c  13.82 log hb  a(hm )  [44.9  6.55 log hb ] log d C M
where a(hm) is calculated as before and CM = 0 for medium sized
and suburban areas or CM = 3 dB for metropolitan areas.

 The COST-231 extension is restricted to the following:

 Frequency range: 1500 MHz – 2000 MHz
 Base station antenna height: 30 m – 200 m
 Mobile station antenna height: 1 m – 10 m
 Distance: 1 km – 20 km

08-Mar-23 42
4B. PATH-LOSS MODELS FOR MICROCELLULAR AREAS
 Microcells span hundreds of meters to a kilometer range.
 Base station antennas are usually mounted below rooftop level
(lampposts or utility poles).
 The shape of the microcell is no longer circular because streets
in urban areas create canyons.
 Little or no propagation through buildings and the cell looks like
a cross or a rectangle.
 Propagation through reflection from buildings and ground and
scattering from nearby cars. Diffraction of obstructed paths is
also important.
 Empirical path-loss models were developed by different
researchers.
08-Mar-23 43
4B. PATH-LOSS MODELS FOR MICROCELLULAR AREAS (CONTD.)
 The parameters are defined as follows:
 d is the distance between the mobile terminal and
transmitter (in km).
 hb is the base station antenna height in meters and hm is the
height of the mobile antenna from the ground in meters.
 fc is the carrier frequency in GHz (from 0.9 to 2 GHz).
 The height of the nearest building above the height of the
receiver antenna is Δℎ𝑚 .
 The relative height of the base station antenna compared
with the average height of buildings is called Δℎ.

08-Mar-23 44
4B. PATH-LOSS MODELS FOR MICROCELLULAR AREAS (CONTD.)
 In LOS situations, there are two slopes of the path loss curve
with a breakpoint distance given by 𝑑𝑏𝑘 = 4ℎ𝑏 ℎ𝑚 /1000𝜆.

L p  81.14  39.40 log f c  0.09 log hb  [15.80  5.73 log hb ] log d for d  d bk
L p  [ 48.38  32.1 log d ]  45.7 log f c  (25.34  13.9 log d ) log hb
 [32.10  13.90 log hb ] log d  20 log(1.6 / hm ) for d  dbk
08-Mar-23 45
4B. PATH-LOSS MODELS FOR MICROCELLULAR AREAS (CONTD.)
 In NLOS situations, there are two cases:
1. Low Rise buildings
L p  [139.01  42.59 log f c ]  [14.97  4.99 log f c ] sgn(h) log(1  h )  [40.67  4.57 sgn(h)] log(1  h ) log d 
20 log(hm / 7.8)  10 log(20 / rh )

2. High Rise buildings (hm = 1.6m)

 Streets perpendicular to LOS streets
L p  135.41  12.49 log f c  4.99 log hb  [46.84  2.34 log hb ] log d

 Streets parallel to LOS streets

L p  143.21  29.74 log f c  0.99 log hb  [47.23  3.72 log hb ] log d

08-Mar-23 46
4B. PATH-LOSS MODELS FOR MICROCELLULAR AREAS (CONTD.)
 Example:
Determine the path loss between the BS and MS of a 1.8 GHz PCS
system operating in high-rise urban area. The MS is located in a
perpendicular street to the location of the BS. The distances of the
BS and MS to the corner of the street are 20 m and 30 m,
respectively. The BS height is 20 m.
 Solution:
The distance between the BS and MS is
202  302  36.05m
The path loss is
L p  135.41  12.49 log 1.8  4.99 log 20  [46.84  2.34 log 20] log(36.05 / 1000)
 68.89 dB
08-Mar-23 47
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS
 Picocells cover a building or parts of buildings spanning 30 m
to 100 m such as in WLANs, wireless PBX systems, and PCS
operating in indoor areas.
 Measurements were conducted to have empirical path loss
models.
 In multi-storey buildings, signal is attenuated by the floors.
(1) Model 1:
 The path loss is given by L p  L0  nF F  10 log d where F is
the attenuation due to each floor, L0 is the path loss at the
first meter, nF is the number of floors through which the
signal passes, and d is the distance between the BS and MS in
meters.
08-Mar-23 48
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS (CONTD.)

 F is the value that minimizes the MMSE between the data

and the best-fit line.
 Typical values for F are 10 dB and 16 dB for 900 MHz and 1.7
GHz, respectively
 The log-normal distribution is the best model for the
shadowing caused by interior objects (furniture,
equipment,…)
 The standard deviation of the log-normal shadowing is about
4 dB.

08-Mar-23 49
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS (CONTD.)
(2) The JTC Indoor Model:
 The path loss is not a linear function of the number of floors (as
was assumed in the previous relationship).
 The Joint Technical Committee (JTC) has proposed the following
model:
L p  A  L f (nF )  B log d  X
where Lf(nF) is the function relating the power loss with the
number of floors, and X is a log-normally distributed random
variable.
 The table below gives a list of suggested parameters at a carrier
frequency of 1.8 GHz (Base and portable device are inside the
same building).
08-Mar-23 50
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS (CONTD.)

(3) Path-loss Models Using Building Material:

 The partition-dependent model uses the basic free space
model and adds losses for each obstruction that is
encountered by a straight line connecting the Tx and Rx.
 Sometimes referred to as Rappaport or Motley-Keenan
models.

08-Mar-23 51
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS (CONTD.)

 The pathloss in this model is given by

𝐿𝑝 (𝑑) = 𝐿𝑝 (𝑑0 ) + 20 log 𝑑 + ෍ 𝑚𝑖 𝑊𝑖 + ෍ 𝑛𝑗 𝐹𝑗

𝑖 𝑗

where 𝑚𝑖 is the number of partitions of type 𝑖 and 𝑊𝑖 is loss

associated with that partition in dB, 𝑛𝑖 is the number of floors
of type 𝑗 and 𝐹𝑗 is the loss encountered with that floor
and 𝐿𝑝 𝑑0 is the pathloss at a reference distance (usually 1 m
in indoor scenarios).

08-Mar-23 52
4C. PATH-LOSS MODELS FOR PICOCELLULAR INDOOR AREAS (CONTD.)
The table shows
some loss values at
2.4 GHz.

Example:
 The straight line connecting the Tx and
Rx intersects two brick walls and one
cubicle partition.
∴ 𝐿𝑝 = 𝐿0 + 20 log 𝑑 + 2𝑊𝑏𝑟𝑖𝑐𝑘 + 𝑊𝑐𝑢𝑏𝑖𝑐𝑙𝑒
08-Mar-23 53
4D. PATH-LOSS MODELS FOR FEMTOCELLULAR AREAS
 Femtocells span from a few meters to a few tens of meters.
 Deployed within individual residences and use low power
devices.
 The JTC model for residential area (previous model) may be
used to find the path loss at 1.8 GHz.
 At 2.4 GHz and 5 GHz, no path loss models are available.

08-Mar-23 54
CELLULAR CAD TOOLS
 CAD deployment tools are widely used to provide prediction
of coverage and plan/tune the network.
 Use Geographic Information Systems (GIS) terrain data base,
along with vehicle traffic/population density overlays and
propagation models.
 CAD Tools provide a variety of propagation models: free
space, Okumura-Hata, …etc.
 Output map with cell coverage with various signal levels and
interference values. This is important to plan out cell
coverage area, cell placement, handoff areas, interference
level, frequency assignment, …etc
08-Mar-23 55
CELLULAR CAD TOOLS (CONTD.)

08-Mar-23 56
CELLULAR CAD TOOLS (CONTD.)

 Let us play with https://www.cellmapper.net/

08-Mar-23 57