Mobile Communications Chapter 2: Wireless Transmission

Mobile Communications
Chapter 2: Wireless Transmission
� Frequencies
� Signals
� Antenna
� Signal propagation
� Multiplexing
� Spread spectrum
� Modulation
� Cellular systems
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.1

twisted
pair
1 Mm
300 Hz
Frequencies for communication
10 km
30 kHz
coax cable
100 m
3 MHz
VLF = Very Low Frequency UHF = Ultra High Frequency
LF = Low Frequency SHF = Super High Frequency
MF = Medium Frequency EHF = Extra High Frequency
HF = High Frequency
VHF = Very High Frequency
UV = Ultraviolet Light
Frequency and wave length:
� = c/f
1 m
300 MHz
10 mm
30 GHz
wave length �, speed of light c � 3x10 8 m/s, frequency f
100 µm
3 THz
optical transmission
1 µm
300 THz
VLF LF MF HF VHF UHF SHF EHF infrared visible light UV
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.2

Frequencies for mobile communication
� VHF-/UHF-ranges for mobile radio
� simple, small antenna for cars
� deterministic propagation characteristics, reliable connections
� SHF and higher for directed radio links, satellite
communication
� small antenna, beam forming
� large bandwidth available
� Wireless LANs use frequencies in UHF to SHF range
� some systems planned up to EHF
� limitations due to absorption by water and oxygen molecules
(resonance frequencies)
� weather dependent fading, signal loss caused by heavy rainfall
etc.
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.3

Frequencies and regulations
ITU-R holds auctions for new frequencies, manages frequency bands
worldwide (WRC, World Radio Conferences)
Europe USA Japan
Cellular
Phones
Cordless
Phones
Wireless
LANs
GSM 450-457, 479-
486/460-467,489-
496, 890-915/935-
960,
1710-1785/1805-
1880
UMTS (FDD) 1920-
1980, 2110-2190
UMTS (TDD) 1900-
1920, 2020-2025
CT1+ 885-887, 930-
932
CT2
864-868
DECT
1880-1900
IEEE 802.11
2400-2483
HIPERLAN 2
5150-5350, 5470-
5725
Others RF-Control
27, 128, 418, 433,
868
AMPS, TDMA, CDMA
824-849,
869-894
TDMA, CDMA, GSM
1850-1910,
1930-1990
PACS 1850-1910, 1930-
1990
PACS-UB 1910-1930
902-928
IEEE 802.11
2400-2483
5150-5350, 5725-5825
RF-Control
315, 915
PDC
810-826,
940-956,
1429-1465,
1477-1513
PHS
1895-1918
JCT
254-380
IEEE 802.11
2471-2497
5150-5250
RF-Control
426, 868
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.4

Signals I
� physical representation of data
� function of time and location
� signal parameters: parameters representing the value of data
� classification
� continuous time/discrete time
� continuous values/discrete values
� analog signal = continuous time and continuous values
� digital signal = discrete time and discrete values
� signal parameters of periodic signals:
period T, frequency f=1/T, amplitude A, phase shift �
� sine wave as special periodic signal for a carrier:
s(t) = A t sin(2 � f t t + � t )
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.5

1
0
Fourier representation of periodic signals
g(
t)
=
1
2
c +
�
�
n=
1
a
n
sin( 2�nft)
+ �b
cos( 2�nft)
t t
ideal periodic signal real composition
(based on harmonics)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.6
�
n=
1
1
0
n

� Different representations of signals
A [V]
� amplitude (amplitude domain)
� frequency spectrum (frequency domain)
� phase state diagram (amplitude M and phase � in polar coordinates)
�
Signals II
t[s]
A [V]
f [Hz]
Q = M sin �
� Composed signals transferred into frequency domain using Fourier
transformation
� Digital signals need
� infinite frequencies for perfect transmission
� modulation with a carrier frequency for transmission (analog signal!)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.7
�
I= M cos �

Antennas: isotropic radiator
� Radiation and reception of electromagnetic waves, coupling of
wires to space for radio transmission
� Isotropic radiator: equal radiation in all directions (three
dimensional) - only a theoretical reference antenna
� Real antennas always have directive effects (vertically and/or
horizontally)
� Radiation pattern: measurement of radiation around an antenna
y
z
x
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.8
z
y x ideal
isotropic
radiator

Antennas: simple dipoles
� Real antennas are not isotropic radiators but, e.g., dipoles with lengths
�/4 on car roofs or �/2 as Hertzian dipole
� shape of antenna proportional to wavelength
� Example: Radiation pattern of a simple Hertzian dipole
y
side view (xy-plane)
x
�/4
y
side view (yz-plane)
� Gain: maximum power in the direction of the main lobe compared to
the power of an isotropic radiator (with the same average power)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.9
�/2
z
z
top view (xz-plane)
x
simple
dipole

Antennas: directed and sectorized
Often used for microwave connections or base stations for mobile phones
(e.g., radio coverage of a valley)
y
side view (xy-plane)
z
top view, 3 sector
x
x
y
side view (yz-plane)
z
top view (xz-plane)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.10
z
top view, 6 sector
z
x
x
directed
antenna
sectorized
antenna

Antennas: diversity
� Grouping of 2 or more antennas
� multi-element antenna arrays
� Antenna diversity
� switched diversity, selection diversity
� receiver chooses antenna with largest output
� diversity combining
� combine output power to produce gain
� cophasing needed to avoid cancellation
ground plane
�/4
�/2
+
�/4
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.11
�/2
�/2
+
�/2

Signal propagation ranges
Transmission range
� communication possible
� low error rate
Detection range
� detection of the signal
possible
� no communication
possible
Interference range
� signal may not be
detected
� signal adds to the
background noise
sender
transmission
detection
interference
distance
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.12

Signal propagation
Propagation in free space always like light (straight line)
Receiving power proportional to 1/d_ in vacuum – much more in real environments
(d = distance between sender and receiver)
Receiving power additionally influenced by
� fading (frequency dependent)
� shadowing
� reflection at large obstacles
� refraction depending on the density of a medium
� scattering at small obstacles
� diffraction at edges
shadowing reflection refraction
scattering diffraction
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.13

Real world example
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.14

Multipath propagation
Signal can take many different paths between sender and receiver due to
reflection, scattering, diffraction
signal at sender
Time dispersion: signal is dispersed over time
LOS pulses multipath
pulses
signal at receiver
� interference with “neighbor” symbols, Inter Symbol Interference (ISI)
� pulses become wider, Delay Spread
The signal reaches a receiver directly and phase shifted
� distorted signal depending on the phases of the different parts
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.15

Effects of mobility
Channel characteristics change over time and location
� signal paths change
� different delay variations of different signal parts
� different phases of signal parts
� quick changes in the power received (short term fading)
Additional changes in
� distance to sender
� obstacles further away
� slow changes in the average power
received (long term fading)
power
short term fading
long term
fading
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.16
t

Multiplexing
Multiplexing in 4 dimensions
� space (s i )
� time (t)
� frequency (f)
� code (c)
Goal: multiple use
of a shared medium
Important: guard spaces needed!
channels k i
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.17
s 1
k 1
c
k 2 k 3 k 4 k 5 k 6
t
s 3
c
f
s 2
t
c
f
t
f

Frequency multiplex
Separation of the whole spectrum into smaller frequency bands
A channel gets a certain band of the spectrum for the whole time
Advantages:
� no dynamic coordination
�
necessary
works also for analog signals
k1 Disadvantages:
� waste of bandwidth
if the traffic is
distributed unevenly
� inflexible
� guard spaces
t
k 2 k 3 k 4 k 5 k 6
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.18
c
f

Time multiplex
A channel gets the whole spectrum for a certain amount of time
Advantages:
� only one carrier in the
medium at any time
� throughput high even
for many users
Disadvantages:
� precise
synchronization
necessary
t
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.19
c
k 1
k 2 k 3 k 4 k 5 k 6
f

Time and frequency multiplex
Combination of both methods
A channel gets a certain frequency band for a certain amount of time
Example: GSM
Advantages:
� better protection against
tapping
� protection against frequency
selective interference
� higher data rates compared to
code multiplex
but: precise coordination
required
t
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.20
c
k 1
k 2 k 3 k 4 k 5 k 6
f

Code multiplex
Each channel has a unique code
All channels use the same spectrum
at the same time
Advantages:
� bandwidth efficient
� no coordination and synchronization
necessary
� good protection against interference and
tapping
Disadvantages:
� lower user data rates
� more complex signal regeneration
Implemented using spread spectrum
technology
k 2 k 3 k 4 k 5 k 6
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.21
k 1
t
c
f

Modulation
Digital modulation
� digital data is translated into an analog signal (baseband)
� ASK, FSK, PSK - main focus in this chapter
� differences in spectral efficiency, power efficiency, robustness
Analog modulation
� shifts center frequency of baseband signal up to the radio carrier
Motivation
� smaller antennas (e.g., �/4)
� Frequency Division Multiplexing
� medium characteristics
Basic schemes
� Amplitude Modulation (AM)
� Frequency Modulation (FM)
� Phase Modulation (PM)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.22

digital
Modulation and demodulation
analog
demodulation
radio
carrier
analog
baseband
signal
data digital
analog
101101001 modulation
modulation
analog
baseband
signal
radio
carrier
synchronization
decision
digital
data
101101001
radio transmitter
radio receiver
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.23

Digital modulation
Modulation of digital signals known as Shift Keying
� Amplitude Shift Keying (ASK):
� very simple
� low bandwidth requirements
� very susceptible to interference
� Frequency Shift Keying (FSK):
� needs larger bandwidth
� Phase Shift Keying (PSK):
� more complex
� robust against interference
1 0 1
1 0 1
1 0 1
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.24
t
t
t

Advanced Frequency Shift Keying
� bandwidth needed for FSK depends on the distance between
the carrier frequencies
� special pre-computation avoids sudden phase shifts
� MSK (Minimum Shift Keying)
� bit separated into even and odd bits, the duration of each bit is
doubled
� depending on the bit values (even, odd) the higher or lower
frequency, original or inverted is chosen
� the frequency of one carrier is twice the frequency of the other
� Equivalent to offset QPSK
� even higher bandwidth efficiency using a Gaussian low-pass
filter � GMSK (Gaussian MSK), used in GSM
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.25

data
even bits
odd bits
low
frequency
high
frequency
MSK
signal
Example of MSK
1 0 1 1 0 1 0
No phase shifts!
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.26
t
bit
even 0 1 0 1
odd 0 0 1 1
signal h n n h
value - - + +
h: high frequency
n: low frequency
+: original signal
-: inverted signal

Advanced Phase Shift Keying
BPSK (Binary Phase Shift Keying):
� bit value 0: sine wave
� bit value 1: inverted sine wave
� very simple PSK
� low spectral efficiency
� robust, used e.g. in satellite systems
QPSK (Quadrature Phase Shift Keying):
� 2 bits coded as one symbol
� symbol determines shift of sine wave
� needs less bandwidth compared to
BPSK
� more complex
Often also transmission of relative, not
absolute phase shift: DQPSK -
Differential QPSK (IS-136, PHS)
11 10 00 01
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.27
A
10
00
1
Q
Q
0
I
11
I
01
t

Quadrature Amplitude Modulation
Quadrature Amplitude Modulation (QAM): combines amplitude and
phase modulation
� it is possible to code n bits using one symbol
� 2 n discrete levels, n=2 identical to QPSK
� bit error rate increases with n, but less errors compared to
comparable PSK schemes
Q
0010
0011
a
_
0001
0000
I
1000
Example: 16-QAM (4 bits = 1 symbol)
Symbols 0011 and 0001 have the same phase _,
but different amplitude a. 0000 and 1000 have
different phase, but same amplitude.
� used in standard 9600 bit/s modems
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.28

Hierarchical Modulation
DVB-T modulates two separate data streams onto a single DVB-T stream
� High Priority (HP) embedded within a Low Priority (LP) stream
� Multi carrier system, about 2000 or 8000 carriers
� QPSK, 16 QAM, 64QAM
� Example: 64QAM
� good reception: resolve the entire
64QAM constellation
� poor reception, mobile reception:
resolve only QPSK portion
� 6 bit per QAM symbol, 2 most
significant determine QPSK
� HP service coded in QPSK (2 bit),
LP uses remaining 4 bit
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.29
10
00
Q
000010 010101
I

Spread spectrum technology
Problem of radio transmission: frequency dependent fading can wipe out
narrow band signals for duration of the interference
Solution: spread the narrow band signal into a broad band signal using a
special code
protection against narrow band interference
power interference spread
signal
power
Side effects:
detection at
receiver
protection against narrowband interference
signal
f f
� coexistence of several signals without dynamic coordination
� tap-proof
Alternatives: Direct Sequence, Frequency Hopping
spread
interference
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.30

i)
iii)
dP/df
dP/df
Effects of spreading and interference
ii)
f
sender
iv)
f
receiver
dP/df
dP/df
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.31
f
f
v)
user signal
broadband interference
narrowband interference
dP/df
f

channel
quality
channel
quality
Spreading and frequency selective fading
1
narrow band
signal
spread
spectrum
2
3
4
guard space
2
1
2
2
2
2
5 6
frequency
frequency
narrowband channels
spread spectrum channels
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.32

DSSS (Direct Sequence Spread Spectrum) I
XOR of the signal with pseudo-random number (chipping sequence)
� many chips per bit (e.g., 128) result in higher bandwidth of the signal
Advantages
� reduces frequency selective
fading
� in cellular networks
� base stations can use the
same frequency range
� several base stations can
detect and recover the signal
� soft handover
Disadvantages
� precise power control necessary
0 1
0 1 1 0 1 0 1 0 1 1 0 1 0 1
user data
chipping
sequence
resulting
signal
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.33
t b
t c
0 1 1 0 1 0 1 1 0 0 1 0 1 0
t b : bit period
t c : chip period
XOR
=

DSSS (Direct Sequence Spread Spectrum) II
received
signal
user data
chipping
sequence
radio
carrier
X
demodulator
spread
spectrum
signal
transmitter
receiver
lowpass
filtered
signal
radio
carrier
chipping
sequence
modulator
transmit
signal
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.34
X
correlator
products
integrator
sampled
sums
decision
data

FHSS (Frequency Hopping Spread Spectrum) I
Discrete changes of carrier frequency
� sequence of frequency changes determined via pseudo random number
sequence
Two versions
� Fast Hopping:
several frequencies per user bit
� Slow Hopping:
several user bits per frequency
Advantages
� frequency selective fading and interference limited to short period
� simple implementation
� uses only small portion of spectrum at any time
Disadvantages
� not as robust as DSSS
� simpler to detect
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.35

f
f 3
f 2
f 1
f
f 3
f 2
f 1
FHSS (Frequency Hopping Spread Spectrum) II
t b
0 1
t d
t d
0 1 1 t
t b : bit period t d : dwell time
user data
slow
hopping
(3 bits/hop)
fast
hopping
(3 hops/bit)
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.36
t
t

FHSS (Frequency Hopping Spread Spectrum) III
user data
hopping
sequence
received
signal
modulator
transmitter
demodulator
frequency
synthesizer
narrowband
signal
narrowband
signal
modulator
frequency
synthesizer
demodulator
spread
transmit
signal
hopping
sequence
receiver
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.37
data

Cell structure
Implements space division multiplex: base station covers a certain
transmission area (cell)
Mobile stations communicate only via the base station
Advantages of cell structures:
� higher capacity, higher number of users
� less transmission power needed
� more robust, decentralized
� base station deals with interference, transmission area etc. locally
Problems:
� fixed network needed for the base stations
� handover (changing from one cell to another) necessary
� interference with other cells
Cell sizes from some 100 m in cities to, e.g., 35 km on the country side
(GSM) - even less for higher frequencies
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.38

Frequency planning I
Frequency reuse only with a certain distance between the base
stations
Standard model using 7 frequencies:
Fixed frequency assignment:
� certain frequencies are assigned to a certain cell
� problem: different traffic load in different cells
Dynamic frequency assignment:
� base station chooses frequencies depending on the frequencies
already used in neighbor cells
� more capacity in cells with more traffic
� assignment can also be based on interference measurements
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.39
f 4
f 3
f 5
f 1
f 2
f 3
f 6
f 7
f 2
f 4
f 5
f 1

f 3
f 1
f 2
f 3
f 2
f 3
f 1
Frequency planning II
f 3
f 1
f 2
f 3
f2 f2 f1 f
f1 3 h2 f3 h h
g 1 g 1
2 h3 2
g1 g
g 1
3 g3 f 2
f 3
f 1
h 2
h 3
f 1
f 3
f 1
f 2
f 3
f2 f3 g2 g1 g3 3 cell cluster
3 cell cluster
with 3 sector antennas
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.40
f 2
f 4
f 3
f 6
f 5
f 1
f 2
f 3
f 6
f 7
f 5
f 2
f 4
f 3
7 cell cluster
f 7
f 5
f 1
f 2

Cell breathing
CDM systems: cell size depends on current load
Additional traffic appears as noise to other users
If the noise level is too high users drop out of cells
Dr.-Ing. Jean-Pierre Ebert MC-I SS07 2.41

Andreas Willig
• The Cellular Concept
• System Capacity
• Channel Allocation
• Handover
• Paging / Location Update
Overview
Cellular System Fundamentals, slide 2

Andreas Willig The cellular concept
The cellular concept
• cellular systems evolved from the first systems supporting wireless and
mobile telephony
• initially their design was focused towards telephony services, data services
were added later on
Cellular System Fundamentals, slide 3

Andreas Willig The cellular concept
Some important milestones
• 1946: the very first analog systems for public mobile telephony are
introduced in some cities of the US
• 1983: the analog AMPS system for wireless telephony is introduced
• 1987-1991: development phase of GSM [9], a digital cellular network
• 2001: GPRS becomes publicly available (packet-switching over GSM)
• 1993: IS-95 [5] is the first commercial cellular CDMA system
Cellular System Fundamentals, slide 4

Andreas Willig The cellular concept
Some important milestones II
• 1989-today: development, deployment and operation of UMTS:
– UMTS = Universal Mobile Telecommunications System [8]
– third-generation digital cellular CDMA system supporting data and
voice services
– since 1999 the UMTS specification is controlled by the 3GPP (3rd
Generation Partnership Project) – all UMTS specifications are publicly
available under
www.3gpp.org
Cellular System Fundamentals, slide 5

Andreas Willig The cellular concept
Design Assumptions and Requirements
• the available amount of spectrum is limited
• a large number of users in a large geographical area must be supported,
otherwise the system is not accepted
business people term this “networking effect”: it becomes more and more attractive for an individual to
subscribe to cellular services the more people he/she can reach this way
Cellular System Fundamentals, slide 6

Andreas Willig The cellular concept
Design Assumptions and Requirements II
• the area is subdivided into cells:
– each cell has at its center a base station (BS)
– each cell contains a number of mobile stations (MS)
– all communication from/to an MS is relayed through a BS, there is no
peer-to-peer communication
– the BS’s are interconnected and connected to fixed networks / other
cellular networks through a backbone or core network and gateways
• two important notions indicate the direction of communication in a cell:
– downlink: from the BS to a MS
– uplink: from the MS to a BS
Cellular System Fundamentals, slide 7

Andreas Willig The cellular concept
Design Assumptions and Requirements III
1
2
2 1
Backbone
BS 1 BS 2
¡ ¡¡¡¡¡¡¡¡¡¡
¡ ¡¡¡¡¡¡¡¡¡¡
¡ ¡¡¡¡¡¡¡¡¡¡
Interworking Function /
Gateway
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤
¤ ¤ ¥¥ ¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
¥ ¥
Other
Networks
¢ ¢¢¢¢¢¢¢¢¢¢¢ £ ££££££££££
¢ ¢¢¢¢¢¢¢¢¢¢¢ £ ££££££££££
¢ ¢¢¢¢¢¢¢¢¢¢¢ £ ££££££££££
Cellular System Fundamentals, slide 8

Andreas Willig The cellular concept
Cell Sizes and Frequency Reuse
• in systems like GSM a number of channels or frequency bands is allocated
to each BS
• the cell size is determined from the transmit power of the BS and the
receive threshold of the MS:
– as long as an MS can decode the signal of the BS, it is inside the cell
but even outside the cell the BS signal may cause interference
– the stronger the BS’s tx power the larger the cell / interference area
– the reuse distance of a frequency f used by a BS A is the geographical
distance where A’s signal only causes negligible interference, and f
may be re-used by another BS B
Cellular System Fundamentals, slide 9

Andreas Willig The cellular concept
Cell Sizes and Frequency Reuse II
• choosing a small transmit power in BS’s thus:
=⇒ reduces cell sizes and reuse distances
=⇒ increases the number of users which may use the same
frequency in a given area (at places separated by at least
reuse distance), and thus increases the system capacity
the larger the system capacity the more revenue is possible
=⇒ increases the number of base stations and thus the
system costs
Cellular System Fundamentals, slide 10

Andreas Willig The cellular concept
Cell Sizes and Frequency Reuse III
• typically network operators choose different cell sizes:
– large cells (macrocells) in sparsely populated rural areas
– small cells (microcells or picocells) in densely populated urban areas
– large cells overlaying small cells (umbrella cells) to support highly
mobile users
• network operators have to do proper cell planning [3] to:
– achieve full coverage of a given area
– accommodate the expected number of users / user densities
– minimize the number of base stations needed
cell planning results in locations of base stations and their cell size
Cellular System Fundamentals, slide 11

Andreas Willig The cellular concept
Mobility and Handover
• mobile users reach the boundary of a cell and move into the next cell
from time to time
• ongoing calls should be maintained when crossing cell boundaries, the
call should be handed over from the old cell to the new cell
• a handover procedure involves exchange of signalling messages between:
– MS
– old BS and new BS
– some further network elements in the backbone, e.g. the gateway to
the fixed network
Cellular System Fundamentals, slide 12

Andreas Willig The cellular concept
Mobility and Handover II
Gateway to PSTN
Cellular System Fundamentals, slide 13

Andreas Willig The cellular concept
Mobility and Handover III
=⇒ the smaller the cell sizes the more handover events!
=⇒ increased signalling traffic
=⇒ since handovers take some minimum time, a too high
handover rate can lead to loss of connection
=⇒ there is a tradeoff between cell sizes and supported
mobile speeds
Cellular System Fundamentals, slide 14

Andreas Willig The cellular concept
Overview
• The Cellular Concept
• System Capacity
• Channel Allocation
• Handover
• Paging / Location Update
Cellular System Fundamentals, slide 15

Andreas Willig System Capacity
System Capacity
• in certain cellular systems the allocated spectrum is subdivided into a
number N of equal-sized frequency channels or channels
• these have to be allocated to the BS such that:
– co-channel interference is minimized
– adjacent channel interference is minimized
– reuse distance is properly considered
• a channel (or portions of it) is assigned to a MS for the duration of a call
• a connected set of M base stations / cells in which each of the N
frequencies is assigned exactly once, is called a cluster
Cellular System Fundamentals, slide 16

Andreas Willig System Capacity
Visualization and Modeling of Cells
case a) omnidirectional antenna
case b) sectorized antenna
Cellular System Fundamentals, slide 17

Andreas Willig System Capacity
Visualization and Modeling of Cells – A Clustering
Example
G
F B
G
E
A
D
F
E
C
E
G
A
D
F
A
D
B
C
B
C
Cellular System Fundamentals, slide 18

Andreas Willig System Capacity
Visualization and Modeling of Cells II
• in reality:
– cells have no regular shape
– two cells typically overlap by ≈ 10% to 15% to enable handover
– a MS in the overlap region of two cells belongs to either cell with
some probability (soft cell boundaries)
• for the hexangular cell layout only certain cluster sizes M are feasible
(i.e. can create a plane tiling) – these satisfy the relation:
M = i 2 + i · j + j 2
solutions are M = 1,3, 4, 7,9, 12,13, . ..
i, j ∈ N0
Cellular System Fundamentals, slide 19

Andreas Willig System Capacity
Estimation of System Capacity
• let us make the following assumptions:
– we adopt the hexagonal cell model
– the overall number of available channels is N
– each cluster consists of M cells, to each cell of a cluster k frequencies
are assigned (k · M = N)
– each cell has radius R
– be D the distance between two BS using the same frequency (reuse
distance)
– we consider only downlink direction, co-channel interference at a MS
has its source in transmissions of other BS than the current one
– the path loss exponent is n, valid for the whole system area
Cellular System Fundamentals, slide 20

Andreas Willig System Capacity
Estimation of System Capacity II
• the ratio
Q = D
R
gives the normalized reuse distance, Q is also denoted as co-channel
reuse ratio
for the hexagonal cell layout one can show that:
Q = √ 3 · M
Cellular System Fundamentals, slide 21

Andreas Willig System Capacity
Estimation of System Capacity III
• if Q is large, we have:
– less interference
=⇒ smaller bit-/symbol error rates
=⇒ better speech quality
– less channels per cell
=⇒ decreased capacity
• conversely, a small Q leads to higher interference and higher capacity
Cellular System Fundamentals, slide 22

Andreas Willig System Capacity
Estimation of System Capacity IV
• let us fix one interior cell and look at its signal-to-interference ratio (SIR)
as experienced by a MS at the fringe of the cell:
where:
S
I =
Pr(R)
�
i∈I Pr(Di)
– I is the set of all interferers, i.e. the set of all BS using the same
frequency
– Di is the distance between the MS and the i-th interferer
for maintaining a good speech quality a SIR of ≥ 18 dB should be used
Cellular System Fundamentals, slide 23

Andreas Willig System Capacity
Estimation of System Capacity V
• if we take into account that:
Pr(d) ≈ Pr(d0) ·
� d
d0
� −n
and assume d0 = 1 (in appropriate units), we have:
S
I =
R −n
�
i∈I D−n i
Cellular System Fundamentals, slide 24

Andreas Willig System Capacity
Estimation of System Capacity VI
• if we take only the closest I0 interferers into consideration and assume
that they all have the same distance D we have
S
I
R−n
= =
I0 · D−n � �n D
R
· 1
I0
=
�√3 �n · M
=⇒ for larger n (e.g. in urban areas) and for fixing a minimal
SIR (e.g. of 18 dB) we can decrease M (increase number
k of frequencies per cell) and thus make smaller clusters,
thus increasing capacity
I0
Cellular System Fundamentals, slide 25

Andreas Willig System Capacity
Overview
• The Cellular Concept
• System Capacity
• Channel Allocation
• Handover
• Paging / Location Update
Cellular System Fundamentals, slide 26

Andreas Willig Channel Allocation
Channel Allocation
• in practice the allocation of channels to cells depends on:
– the expected traffic load per cell / per area unit
∗ example: urban vs. rural areas
– certain performance measures
=⇒ often different cells have different numbers of channels
assigned, to accommodate differences in traffic load
Cellular System Fundamentals, slide 27

Andreas Willig Channel Allocation
Call Blocking and Call Dropping
• each cell i in a cellular system has a finite number ki of channels
• if a MS requests a new call in a full cell, the call is blocked
• if a MS moves from a neighbored cell into a full cell, the call is dropped
• there are two important performance measures for a cellular system:
– call blocking probability
– call dropping probability
Cellular System Fundamentals, slide 28

Andreas Willig Channel Allocation
Call Blocking and Call Dropping II
• the call blocking probability:
– should be low (typical target: < 1%), blocked customers are unhappy
– can be computed under specific assumptions (Poisson arrivals,
exponential call holding times) from simple queueing theory results
(Erlang loss formulas)
• the call dropping probability:
– should be even lower, call dropping makes customers even more angry
=⇒ proper channel allocation strategies are needed to fulfill
these requirements for a given traffic load
Cellular System Fundamentals, slide 29

Andreas Willig Channel Allocation
Fixed Channel Allocation (FCA)
• channels are assigned to cells / BS on a permanent basis and the BS
assigns them to the MS for the duration of a call
=⇒ FCA is susceptible to call blocking / call dropping
• the following constraints have to be considered in the assignment:
– number of available frequencies
– avoiding adjacent-channel interference:
∗ do not assign neighbored frequencies to a single cell
∗ do not assign neighbored frequencies to neighbored cells
– avoid co-channel interference: keep a minimum distance between two
cells using the same frequency
– accommodate expected traffic load
Cellular System Fundamentals, slide 30

Andreas Willig Channel Allocation
Fixed Channel Allocation (FCA) II
• a number of heuristic techniques have been developed to solve FCA
problems [6]
• an assignment is only valid as long as the traffic load distribution does
not change (much) – otherwise the call dropping/blocking rate increases
and the allocation must be re-computed
Cellular System Fundamentals, slide 31

Andreas Willig Channel Allocation
FCA with Borrowing
• if a new call or handover call arrives to a crowded cell, the BS might ask
its neighbor BS to borrow a channel for the call duration:
– if successful, the channel is temporarily used by the accepting BS
– it is not used by the donating BS for the duration of the call
– after the call finishes the accepting BS returns the channel
still the interference constraints have to be obeyed
Cellular System Fundamentals, slide 32

Andreas Willig Channel Allocation
Dynamic Channel Allocation
• several cells are grouped into a cluster
• a cluster possesses a clusterhead (CH)
• channel allocation is done dynamically by the CH:
– if a new or handover call arrives to a BS, the BS requests a channel
from the CH and issues this to the MS
– after the call finished, the channel is returned to the CH
this approach allows to explore statistical multiplexing gains!!
• a CH can coordinate with neighbored CH’s to minimize co-channel
interference
Cellular System Fundamentals, slide 33

Andreas Willig Channel Allocation
Overview
• The Cellular Concept
• System Capacity
• Channel Allocation
• Handover
• Paging / Location Update
Cellular System Fundamentals, slide 34

Andreas Willig Handover
Handover
• a handover becomes necessary if a MS with an ongoing call:
– is about to leave its current cell, and
– is about to enter a neighbored cell
• possible causes for handovers:
– mobility of the MS
– signal degradation to current BS due to moving obstacles
• types of handovers (w.r.t. data, not to signalling connections!):
– hard handover: MS has a data connection to at most one BS
– soft handover: MS can communicate with several BS simultaneously
Cellular System Fundamentals, slide 35

Andreas Willig Handover
Handover II
Gateway to PSTN
Cellular System Fundamentals, slide 36

Andreas Willig Handover
Handover III
• necessary actions in a hard handover:
– the new BS must allocate a channel and assign it to the MS
– the old BS must deallocate the channel
– if the call is going through a PSTN gateway the connection between
the gateway and the old BS must be re-routed to the new BS
• important requirements:
– no noticeable degradation of speech quality during handover
– no actions required from the user
• important questions:
– when is a handover initiated?
– who initiates the handover?
Cellular System Fundamentals, slide 37

Andreas Willig Handover
Handover Initiation
• let us assume that:
– the BS/MS need a minimum signal power Pmin to maintain a call at
an acceptable level of speech quality
– if signal power drops below this level (the MS moves out of the range
of the BS) the call is canceled
– no handoff is initiated as long as the signal level is above
Pmin + δ
with the safety margin δ > 0
– a handover process takes some minimum time tmin
Cellular System Fundamentals, slide 38

Andreas Willig Handover
Handover Initiation II
• if δ is too large the handover is initiated early and unnecessarily
=⇒ increased signalling traffic
• if δ is too small:
– if the mobiles speed v is so large that it moves out of the cell before
the handover is completed (tmin) the old connection drops and there
is no chance to set up the connection to the new BS
Cellular System Fundamentals, slide 39

Andreas Willig Handover
Handover Initiation III
• to determine the signal strength, measurements must be taken over a
timespan sufficiently long to average over fast fading
• the measurements can either be done by the BS or by the MS
• in mobile-assisted handover (MAHO):
– the MS measures the signal strength of surrounding BS (e.g. by
evaluating signal strength of specific beacon packets)
– the MS reports the measurement values to its current BS or other
stations in the network
– handover if another BS is significantly stronger than the current BS
such a behavior is called hysterese: choosing the BS such that at any time the MS is connected to
the best one would likely produce many handovers
Cellular System Fundamentals, slide 40

Andreas Willig Handover
Handover Initiation IV
• MAHO is used both in GSM and UMTS
• GSM:
– hard handover
– execution of a handover after making the decision takes one to two
seconds
• UMTS:
– soft handover and softer handover
Cellular System Fundamentals, slide 41

Andreas Willig Handover
Mechanisms for Call Dropping Avoidance
• to avoid dropping of handover calls, the BS treats channel allocation for
new calls and handover calls differently
• the guard channel concept:
– the BS puts aside some channels and allocates these exclusively to
handover calls
=⇒ reduced capacity
=⇒ can be effectively combined with DCA, to avoid
allocating guard channels in all cells of a cluster
Cellular System Fundamentals, slide 42

Andreas Willig Handover
Mechanisms for Call Dropping Avoidance II
• queueing of handover requests:
– the BS or CH puts handover requests into a queue
– as soon as an ongoing call ends or roams away, the corresponding
channel is assigned to a handover request from the queue
– in this case tmin can be large, depending on the number of mobiles
Cellular System Fundamentals, slide 43

Andreas Willig Handover
Umbrella Cells
• if the mobiles have vastly different speeds there is no best choice of δ:
– for small speeds the cells can be small while maintaining a small
handover rate
– for high speeds small cells would lead to a very high handover rate
• solution: put slow mobiles into small cells and fast mobiles into large
umbrella cells (with higher antennas and larger tx powers)
=⇒ the cells overlap spatially, but use different channels
• the decision can be made by CH from observing a MS’s handover rate
Cellular System Fundamentals, slide 44

Andreas Willig Paging / Location Update
References
[1] Manuel Duque-Anton. Mobilfunknetze – Grundlagen, Dienste und Protokolle. Verlag Vieweg,
Braunschweig / Wiesbaden, Germany, 2002.
[2] Jose M. Hernando and F. Perez-Fontan. Introduction to Mobile Communications Engineering. Artech
House, Boston, 1999.
[3] Ajay R. Mishra. Fundamentals of Cellular Network Planning and Optimisation: 2G/2.5G/3G... Evolution
to 4G. John Wiley & Sons, 2004.
[4] Theodore S. Rappaport. Wireless Communications – Principles and Practice. Prentice Hall, Upper
Saddle River, NJ, USA, 2002.
[5] Arthur H. M. Ross and Klein S. Gilhausen. Cdma technology and the is-95 north american standard.
In Jerry D. Gibson, editor, The Communications Handbook, pages 199–212. CRC Press / IEEE Press,
Boca Raton, Florida, 1996.
[6] Harilaos G. Sandalidis and Peter Stavroulakis. Heuristics for solving fixed-channel assignment problems.
In Ivan Stojmenovic, editor, Handbook of Wireless Networks and Mobile Computing, pages 51–70. John
Wiley & Sons, New York, 2002.
[7] Mischa Schwartz. Mobile Wireless Communications. Cambridge University Press, Cambridge, GB, 2005.
Cellular System Fundamentals, slide 53

Andreas Willig Paging / Location Update
[8] B. Walke, P. Seidenberg, and M. P. Althoff. UMTS – The Fundamentals. John Wiley and Sons,
Chichester, UK, 2003.
[9] Bernhard Walke. Mobile Radio Networks – Networking, Protocols and Traffic Performance. John Wiley
and Sons, Chichester, 2002.
Cellular System Fundamentals, slide 54