Cellular Network

UNIT-IV Wireless System Planning

By
Prof. Kantilal B. Kharat
Department of Electronics & Telecommunication Engineering
NBN Sinhgad School of Engineering, Ambegaon(Bk), Pune-41
 Conten
t nit- 4 -Wireless System Planning
U
 Introduction (8.1 to 8.7 of T2)
 Free Space model
 Okumura model
 Hata Model
 Log Normal Shadowing
 Receiver noise computation
 Link-Budget Analysis (8.8 of
T2)
 Tele-traffic Theory(8.9)
 Tele-traffic System Model (8.10)
2  Steady State Analysis(8.11)
 Objective
s To make students understand Various aspects of wireless-system planning.


Outcome
• Describe and analyse the aspects which are necessary to characterize the
overall performance of a large wireless network

3
 Wireless System
Planning
Introduction
To plan the installation and deployment of a comprehensive wireless
network, it needs to characterize the performance of the communication
system in terms of
• The transmitted power
• The total load in terms of users that can be supported by the
network.

4
 • The strength of the transmitted wireless radio signal decreases as the distance of
propagation increases.
• Radio waves follow an inverse square law for power density: the power density is
proportional to the inverse square of the distance.
• Every time we double the distance, we receive only one-fourth the power.
• The signal strength at the mobile is a function of the distance d. Hence, we need
models which predict the mean signal strength at the receiver as a function of the
separation between the transmitter and the receiver. These models are also termed
large-scale propagation models

5
 How to predicts the received signal
strength ?
The Friis free-space equation, which gives the received power Pr (d) as a function of the distance
d, given by

where Pt - the transmitted power

Gt is the transmit antenna
gain Gr is the receive antenna
gain
λ is the wavelength
L ≥ 1 the system-loss factor
Lets assume that the power received at a reference distance d0

6
That the received power decays as the inverse of d^2. The number 2 in the exponent
of d is also termed the path-loss exponent.
In db

The path-loss exponent varies from scenario to scenario and is typically greater than 2, especially
in
7the presence of reflectors.
 Consider the ground-reflection scenario, The total received signal ETot is given
as

8
 E0 = the transmitted signal amplitude at a reference
distance d0,
dLOS = the propagation distance for the
LOS component.
denotes the phase lag of the carrier at the
receiver arising due to the propagation
delay

• The signal amplitude is inversely proportional to

the distance of propagation

Assuming dLOS ≈ d, for the large-scale path loss, where d is the ground distance
between the transmitter and the receiver

9
 Eg corresponding to the ground-reflection
component
− ve sign arises because of the phase inversion from ground
reflection

10
 Evaluating
Δd

11
12 r
 Substituting
Δd

Therefore, the net magnitude |E Tot | of the received signal is inversely proportional
to 1/d^2 , it follows that the net received power PTot, which is proportional to the
squared magnitude, decreases as 1/ d^4 , indicating a path-loss exponent n = 4.
Thus, the path-loss exponent can vary depending on the scenario and, therefore, the
received power decays as 1/ d^n , where n = 2 for free space, while n can be
significantly greater than 2
Okumura Model
A model developed to accurately model the received signal strength in
practical wireless scenarios,
Okumura model, proposed by the Japanese engineer Yoshihisa
Okumura in 1968
• Used to predict signal strength in urban/suburban areas.
• Is valid roughly in the 150-to-1920 MHz range

Technique used is
• Evaluation of free space loss
• Determine median attenuation
• Add correction gain factors for transmitter and receiver antenna
heights (from curves or equation) and area gain factor (from curves)
• Okumura developed a set of curves giving the median attenuation relative to free
space (Amu), in an urban area over a quasi-smooth terrain with a base station
effective antenna height (ht) of 200 m and a mobile antenna height (hr) of 3 m.
• These curves were developed from extensive measurements using vertical omni-
directional antennas at both the base and mobile, and are plotted as a function of
frequency in the range 100–1920 MHz.
• To determine path loss using Okumura's model, the free space path loss between the
points of interest is first determined, and then the value of Amu(f, d) is added to it.
The model can be expressed as
L50- 50th percentile -i.e median value of propagation loss between Tx and
Rx measured in dB.
LF denotes the free-space-propagation
loss Amu (fc, d) is the correction factor
G (hte), G (hre) gain factors corresponding to the transmit and receive antenna
heights the hte, ( Base station antenna height) , hre(Mobile station antenna
height) respectively;
GArea is the gain due to the environment
Base station effective antenna height, hte = 200m and hre =3 m mobile
station height
gain factors G (hte), G (hre) respectively are given as

**Base station antennas above 200m

gib=ves positive gain and below
200m give negative gain, hence
200m is considered
• The quantity Amu (fc, d) is a correction factor as a function of the carrier
frequency fc and distance d.This has been plotted for several values of fc, d

For example, from the Okumura model, the factor Amu at distance d = 5 km and fc
=
1.8 GHz is given as

Amu (1.8 GHz, 5 km) = 28 dB

• The quantity GArea is a correction factor for various environments. It has been
computed and plotted for different frequencies for suburban, quasi-open, and open
areas.
•Example, the factor GArea for a suburban area at 1.8 GHz is given as
GArea = 12 dB
Hata Model
The Hata model is another popular model
• Signal strength prediction
• Proposed by the Japanese engineer Masaharu Hata in his 1980
• Hata model presents an analytical approximation for the
graphical- information-based Okumura model
Okumura model and Hata Model are based on extensive
Measurements applied for path loss analysis
Content Beyond
Syllabus
Log-Normal Shadowing
• The surrounding environment at different locations is very different in a
wireless scenario.
• Example, users can be shadowed by large objects such as walls or buildings.
• Thus, the net received signal strength is basically a random variable with the
mean predicted by the path loss.
• This random dB deviation about the mean signal strength
canbe modelled as a Gaussian random variable- Xσ ∼ N (0,
σ^2 ).

• Xσ, the deviation about the mean is Gaussian or normally distributed.

• Xσ, in dB, is logarithmically related to the signal power.
• Hence, Log of the received signal power is Gaussian distributed, OR
• log-normal
Xσ is a log normal shadowing.Gaussian random variable with variance σ^2. This is termed as
31
• (TakPreo.f aR RntIukamkraerical)
 Path loss in dB is given by
Where, L50 denotes the median-path

loss Therefore, the probability that this path loss is greater than the threshold value γ is given

as

where Q (·) denotes the Gaussian Q-

function.
Example: Consider the previous example where L50 = 167 dB. Let the deviation σ of
the log-normal shadowing be given by σ = 6 dB. Compute the threshold γ such that
the path loss (PL) is greater than γ at only 5% of the locations in the cell. In other
words, at 95% of the locations the path loss is less than γ.
 Path loss is greater than 177 dB only 5% of the time. In other words, 95% of the time,
the path loss is lower than 177 dB.
i.e reliability is ρ = 95% = 0.95.
The required margin M dB is given as, σQ−1 (1 − ρ)
=6 dB × Q^−1 (0.05) = 6 dB × 1.65 ≈ 10 B01/28/2022
34
dB
 Receiver-Noise Computation
• With Path loss in propagation there is a noise at the receiver arises due to thermal
effects and is also known as thermal noise.

• The noise Power Spectral Density (PSD) η0 denotes the noise power per hertz of
bandwidth.

Noise power = η0 × B

noise power spectral density η0 is given by, η0 = kT

F Noise power = kT BF

where k = 1.38 × 10−23 is the Boltzmann constant, T is the temperature in Kelvin,

and F is the noise figure.(Take a Numerical)
35
 Link-Budget Analysis of a wireless link
Systematic listing
• Power losses
• Gains
of different intermediate components in the transceiver
chain

36
 link-budget expression for the SNR required is given as

SNRreq = Pt ( dB) + Gt ( dB) − L50 ( dB) − M dB + Gr ( dB) − Lc ( dB) − (N +

1) dB

From the above the required transmit power is given as

Pt ( dB) = SNRreq − Gt ( dB) + L50 ( dB) + M dB − Gr ( dB) + Lc ( dB) + (N +

1) dB

37
Consider a wireless-signal-propagation scenario with cell radius d = 8 km, carrier
frequency fc = 2.1 GHz, transmit antenna height hte = 40 m, and receive antenna height
hre = 2 m. Let the standard deviation σ = 6 dB for the log-normal shadowing and it is
required to achieve a reliability of ρ = 95%. The temperature T = 293 K, bandwidth B
= 30 kHz, noise figure F = 5 dB. Further, the wireless link has a receive antenna gain of
5 dB, cabling losses of 3 dB and a transmit antenna gain of 12 dB. Consider a scenario
with level of interference equal to the noise power, and a Rayleigh fading channel
with average power unity. Through a link-budget analysis, compute the transmit power
required to achieve a bit-error rate of 10^−4 at the receiver for BPSK modulation.
Pt ( dB) = SNRreq − Gt ( dB) + L50 ( dB) + M dB − Gr ( dB) + Lc ( dB) + (N +
1) dB
38
39
• The median-path loss = L50 ( dB) = 167 dB.
• For a reliability of ρ = 95% = 0.95, (as seen in above
Examplethe required margin M dB = 10 dB.
• the noise power at 293 K and bandwidth 30 kHz
is 3.84 × 10^−16 = −154 dB
• Given that the interference power I is equal to that of the
noise power N,
i.e., I = N = 3.84 × 10^−16.
Hence, N + I = 2 × 3.84 ×
10^−16 (N + 1) dB
= −154 dB + 3 dB = −151 dB
40
 Gt ( dB) =
12dB M dB =
10 dB Gr ( dB)
= 5 dB Lc
( dB) = 3 dB
Pt = 37 − 12 + 167 + 10 − 5 + 3 − 151 = 49 dBW (db per
watt or we can calculate dB/mW)

41
 • Link Budget is a way of quantifying the link performance.
• The difference between the minimum received signal level and the
actual received power is called as Link Margin.
• The link margin should be a positive value should be maximized(at
least should be 10dB or more for reliable link)

42
43
 Teletraffic Theory
• Cellular systems employ the principle of trunking to meet the demands
of a large number of users, with a limited number of
channels.
• one-to-one or dedicated channel allocation results in wastage of
resources as the probability that all users are active at a given instant of
time is low.
• Practically, only a few channels are necessary to meet the demands of
users as they are random in nature.
• The large number of users share these limited number of channels,
which is basically statistical multiplexing, i.e., assigning the
channels to the various users based on demand
44
45
 In a cellular
•system
less channels in each cell are shared by a large number of subscribers the cell.
In a landline or a PSTN network
• few trunck lines at the exchange for a large number of subscriber
• customers on these lines have to share the limited number of outgoing lines
at the telephone exchange.
• Thus, since the number of users is much greater than the number of
channels available in the cell, there is always a finite probability that all the
lines are occupied.
• Hence, when a new user requests a channel for communication, his call is
blocked as there are no channels available for communication. This
probability with which calls are blocked is termed blocking
46
probability, or in cellular systems, is also termed grade of service 01/28/2022
• Blocking probability derived using teletraffic theory
i.e traffic per user A0

Traffic Intensity = A0 = (user call rate) × (average call duration)

Consider call rate of 2 calls per hour & 2-minute average holding time for
each.

E denotes the unit erlang of traffic. For N number of

users
Traffic Intensity = A0 *N
 Teletraffic System Model
In Teletraffic the random-call-arrival and s and call terminations areindependent
random events, modeled by the Poisson distribution.

Probability that k calls arrive in a time duration t is given as

Consider now a cellular system with N available channels.

• The state of the wireless system that the number of channels

occupied at a given point of time is given by a state diagram
Consider states S0, S1, . . . , SN ,
where Si denotes the state the i channels are occupied.
• In the state S0, all the channels are vacant,
• In the state SN , all the N channels are occupied.
• Any new call arrival in the state SN will, be blocked.
• Let Pk denote the probability of the system being in the state k, i.e., k of the N
channels are occupied.
• Therefore, PN denotes the probability of the system being in state N, in which any
arriving call is blocked.
• PN denotes the blocking probability of the system.
 Steady State analysis
Consider an infinitesimally small time interval Δt.
The probability that one call arrives in time interval Δt is given
by

consider the average call time as T. the call departure can be modeled as
a Poisson process with departure μ = 1/T.
Pr(one call departs) = μΔt
If k channels occupied. Hence, probability of call departure in the state
k is
= kμ (Δt).
1. The system is in the state Sk−1 at the time t and one call arrives in Δt.

The probability of this event is given as Prob = Pk−1 × (λΔt)

2.The system is in the state Sk+1 at the time t and one call departs in Δt.

The probability of this event is given as Prob = Pk+1 × (k + 1) (μΔt)

3. .The system is in the state k, and call neither arrives or departs.

The probability of this event is given as Prob = Pk × (1 − λΔt − (k + 1)

μΔt)
Therefore, the probability Pk can be expressed as
Pk = Pk−1 × (λΔt) + Pk+1 × (k + 1) (μΔt) + Pk × (1 − λΔt − (k + 1)
μΔt)

Simplifying the above expression,

(λ + kμ) Pk = λPk−1 + (k + 1) μPk+1

As no further calls can depart from S0, the expression for P0 can be derived as
P0 = P0 (1 − λΔt) + P1μΔt & P1 = λ/ μ P0
Sim