CORE1240 Electronic & Information Technology

Chapter 5 - Transmission of Signals

Wireless Communications Technology

P.1
Lecture 18 - Introduction to wireless
communications
- Radio wave as information carrier and frequency
allocation
- Wireless communications:
2G/3G/4G/5G Cellular Systems

P.2
Transmission of Information
• In the digital age, we commonly transmit signals
through three types of media: electrical wires,
optical fibers, and wireless/radio
• In this Chapter, using radio/wireless
communications and cellular phones as examples,
we will cover topics including:
– Cellular structure
– Frequency translation (already covered in
Chapter 1)
– Methods for multiple access: FDMA, TDMA,
CDMA, OFDMA

P.3
 Transmission Through Electrical Wires
• Electrical wires are typically made of copper, a good conductor of
electricity. They are used everywhere for short-distance
transmission of both analog and digital signals:
– Telephone wire loop from your home to the telephone network (<10km)
– Ethernet cable from your PC to the router (< 100m)
– Co-axial cable for your Cable TV service (~1 km)
• Telephone loops take one pair, or two wires, for communications in
both directions. Ethernet typically takes two pairs. Gigabit Ethernet
takes more pairs. Most are based on unshielded twisted pair (UTP)
which has two wires twisted together so that interference
from nearby electromagnetic signals can be cancelled.
• Co-axial cable is suitable for high bandwidth (many TV channels)
broadcast (one-to-many) transmission. It is also costly and bulky.

unshielded twisted pair Co-axial cable has one wire

has two wires twisted inside of the second wire which
together is in form of a ring and serves as
a shield
P.4
Transmission Through Optical Fibers
• Optical transmission systems came in the 1970’s.
Prof. Charles Kao claimed jokingly that optical
fibers will remain the best media for long-distance
communications for the next 1,000 years (he also
reminded us that we should not accept his claim). Strands of fibers

• Today, the backbone of the telephone network

and the Internet is all based on single-mode
fibers, which enables very high-rate digital data
transmission over very long distances.
• In today’s DWDM (Dense Wavelength Division
Multiplexing) systems, each strand of fiber allows
the transmission of tens to a hundred wavelength
channels, each channel carrying 10 Gbps of data
(up to 40 Gbps since 2009). Optical signals can
be transmitted for hundreds of kilometer before A DWDM
Transmission System
we need to convert them back to electrical signal
P.5
for processing.
 Optical Fibers and Lasers
• Single-mode fibers have a diameter of 125 m (1/8 mm) including
the cladding, with a core diameter of 8-10 m where light
propagates. Made out of silica glass, it is much cheaper than
copper. But single-mode fiber communications require lasers that
can produce very pure light, and expensive modules for coupling
the light signal into the fiber!
• Systems today typically use near-infrared light at wavelengths of
around 1.5 m (1,500 nm), or frequency of 2 x 1014 Hz. Because
different frequencies travel at slightly different speeds down the
fiber, the light from the laser source must be so pure that its
spectrum has a width of no more than a few megahertz. Otherwise,
the signal will become spread out, limiting data rate or
transmission distance.

Propagation of light signal in optical fiber

A laser diode Typical wavelength for optical
Transmission systems P.6
 Radio/Wireless Transmission
• Since the early 1900’s, scientists/engineers have been making use
of wireless propagation of radio waves which are electromagnetic
wave in the frequency range of 10 KHz – 300 GHz to transmit
information. Practical systems use frequencies between 300 KHz
to 30 GHz:
• AM radio broadcast, citizen band radio
• Long-distance telephone communications in early days
• FM radio and television broadcast
• Satellite, marine and remote communications
• Cordless phones
• Mobile/cellular phones
• WiFi, Bluetooth, RFID
• 300 GHz is the practical upper frequency limit from electrical
circuits/processes. 10 KHz is the practical lower limit for radio
propagation in the earth environment.
P.7
Spectrum Allocation
• Many different wireless communication systems share the same
medium (which is free space) for transmission of their signals. For
their signals to be kept separated, these systems use different
frequency bands allocated by the government, based on many
technical, economic and political considerations. Most countries
are signatories of and bounded by the international radio treaty.
• Different bands have different transmission characteristics in the
earth environment - reflection by ionosphere, water absorption,
scattering loss, etc. It means some bands are good for indoor,
some for long distances, bands below 500 KHz only good for
marine environment, etc.
• Frequency spectrum is a valuable government asset. Companies
have paid large sums to license spectrum usage.
• There are some unlicensed bands scattered throughout the
spectrum for use by citizens (e.g., ~27/28 MHz typically used for
remote control toys). P.8
Spectrum Allocation
in the United States
• In Chapter 1, we presented the
simplified US frequency allocation
1 GHz
chart (shown at the right). The
detailed chart (shown below) can
be easily found from FCC (Federal
Commission on Communications)

1 MHz

10 KHz
P.9
 Frequency Allocation in Hong Kong
• In Hong Kong, frequency band assignment is governed by the Office of the
Communications Authority (OFCA).

P.10
Evolution of 1G to 5G
1G: Analogue switching using Frequency Modulation (FM) for
vice only. Access technique: Frequency Division Multiple
Access (FDMA). (Chapter1: Lecture 7)

P.11
 Digital Cellular Systems (2G)
• 2nd generation (2G) cellular systems are all digital, meaning that
speech signals are transmitted as binary bit streams
• The most popular 2G standard is the Global System for Mobile
Communications (GSM = Groupe Speciale Mobile). Standard
GSM 900 uses carrier frequencies from 890 MHz to 960 MHz.
Different frequencies may be used - such as DCN 1800 in Europe
and PCS 1900 in the US.
• In GSM, speech data is compressed to rates < 14.4 kbps for
efficiency (source coding), and protected by ½ rate error correcting
code (channel coding) to enhance quality
• Use of TDMA (Time Division Multiple Access) together with FDMA
• Being digital, 2G systems use encryption to keep your conversation
and identity private, and support many additional services such as
text messaging, voice mail, caller ID (since 1997 in HK), etc.
P.12
Digital Cellular Systems (2.5G – 4G)
2.5G: Built on 2G systems to provide non-voice data services (e.g.,
web access) at rate of 40 kbps – 60 kbps. Been available for a
few years in HK as General Packet Radio Service (GPRS).

3G: Since 2004 in HK. Use of CDMA (Code Division Multiple

Access, more later) to enhance the capacity and quality for
voice, and higher bit-rate for data services - 2 Mbps
theoretically and ~384 kbps practically. A 3G phone usually
allows video-calls, built-in web browser, PDA functions, etc.

3.5G: Used in HK as HSPA (High Speed Packet Access).

Enhancement of 3G technology to provide data rate up to 20
Mbps (practical 2 Mbps).

4G/LTE: Use of many new technologies, with OFDMA (Orthogonal

Frequency Division Multiple Access) replacing CDMA. Peak
upload (uplink) and download (downlink) rates are 500 Mbps
and 1 Gbps, respectively.
P.13
Digital Cellular Systems (5G and beyond)

5G: It has higher download speeds (up to 10 Gbit/s), lower latency

(<5ms, allow real-time remote control) and enhanced capacity (up to
1,000x more capacity than 4G). The key enabling technologies used
in 5G networks include higher frequency, OFDMA, massive MIMO
(Multiple-Input Multiple-Output), beamforming, etc.

P.14
CORE1240 Electronic & Information Technology

Lecture 19 - Cellular Network Basics

- Cellular Telephony: cellular principle, network
architecture, handoff and roaming, capacity

P.15
Introduction to Cellular Phones
• A cell phone is an extremely sophisticated two-way radio
communication device. With the cell phone you can
communicate anytime anywhere!
• Uses frequencies typically between 800 MHz to 5.9 GHz
• It is called cellular phone because the coverage area is
divided into cells (geographical zones), each served by a
base station which communicate radio signals between
your handset and the telephone network.
• Current cellular systems being used are mostly fourth
generation (4G) systems.

P.16
 The Cellular Principle
• Mobile telephony is based on the cellular principle which is to
divide the coverage areas into smaller neighborhoods called cells.

• Each cell is served by a base station which consists of a set of

antennas and control equipment that handles all the wireless
signals in the cell. Users communicate with the nearest base
station that their cell phones detect.

• Many users in different cells can talk to their base stations at the
same time, greatly increasing the service capacity of the network.
The coverage area of a cell (determined by the maximum distance
for a mobile to receive the signal from the base station) is limited,
thus only relatively low-power transmission is needed (so cell
phones can be compact).

• When a user moves from one cell to another, the network needs to
automatically hand off the call from one base station to another.
P.17
 Cell and Network Architecture
• Cell: In a simple scenario, the base station is placed at the center
of the cell and equipped with omni-directional (“all” direction)
transmitters and receivers. In this situation, the cell area can be
represented by a circle or a hexagon and may have a radius of
100’s of m (or shorter) to 10’s of km.
– Current systems further divide each cell into multiple (e.g. 3)
sectors (like slicing a pizza into 3 equal slices).
• In each cell, multiple users communicate with the base station at
the same time using multiple access techniques: FDMA, TDMA,
CDMA, and OFDMA in 4G.
• The base-stations are connected to Mobile Telephone
Switching Offices (MTSO) which interconnect with the landline
telephone network (known as the Public Switched Telephone
Network, PSTN) so that a mobile phone can talk to a fixed line
phone or another mobile phone.
• There are databases in the network to keep track of the locations
of users and their IDs, security codes, etc.
P.18
 A schematic cellular network
Mobile telephone Base stations
switching office (antenna, transceiver
+ control equipment)
Public switched
telephone network

Central
Office

wire-lines
(copper, optical fiber,
Hexagonal Cells
and sometimes The actual cell geometry is
A fixed line wireless) affected by geographical region,
telephone building, base station antenna
design, etc. The “hexagonal”
cells are just a schematic to
model the cellular network.
P.19
Base stations on HKUST campus

P.20
 Inside a Telephone Central Office – MTSO or PSTN
Cable
Ductwork
Switching
Equipment Wiring Cross-
aisle with Connection Panel
front and
back
clearance
standard

Audio and
visual alarm

Line
Interfaces P.21
Inter-cell Interference and Frequency Reuse Plan
• In 1G and 2G systems, adjacent cells must not use the same
frequency channels or their signals will interfere with each other.
The allocation of frequencies to different cells is called frequency
reuse plan. In 3G, adjacent cells can use the same frequency
because signals are separated by CDMA codes (lecture 20)
• In the simple plan below, cells are organized into clusters of 3 and
each cell uses 1/3 of the frequency channels. No adjacent cells use
the same frequency and each cluster reuses all frequencies.

Uses channels 1, 4, 7, …

Uses channels 2, 5, 8, …

Uses channels 3, 6, 9, … A cluster of 3

cells reuses
Cell clusters Adjacent cells using all frequencies
covering entire area different frequencies
P.22
 Different Cluster Sizes
• With a cluster size of 3, there may not be
sufficient separation between cells using
the same frequency
2
• Engineers also use common cluster sizes 7 3
of 4, 7, and 12 , etc 1
6 4
• The larger the cluster size, the further 5
apart are cells using the same
frequencies, and hence the smaller the
interference, but the number of
frequencies available to each cell
becomes smaller!
• A smaller cluster size allows higher Cluster size of 7
frequency reuse
• Frequency planning is not an issue in 3G
P.23
Can we increase the network capacity by raising the transmitter
power?

• Increasing the transmitter power increases the cell size

(i.e. the range or coverage), but not the available number
of channels which is determined by the number of
frequency channels and codes, etc.
• So, increasing the transmit power would only reduce the
number of cells and the network capacity instead of
increasing the capacity!
• To increase the network capacity, we should use smaller
and more cells over the same geographic area. But more
cells also means more base stations and higher cost!

P.24
 Macro, Micro and Pico cells
• So, what we find is that in heavy-traffic areas such as urban areas, cells
are subdivided into micro cells and pico cells (with coverage of 100 m or
smaller) to maximize frequency reuse.
• In the country side we use larger (macro) cells with coverage of ~10 – 20
km (also consuming more power from your cell phone; hence your battery
drains faster!)
Micro cell (urban area)
Pico cell e.g. Central, Mongkok,
(e.g. airport, Causeway Bay
shopping mall,
HKUST campus)

Macro cell (countryside)

e.g. Sai Kung

P.25
Hand-off – from cell to cell
• In cellular, the mobile unit may move around from cell to cell.
• The process of switching a user from one cell to another while the call is still in
progress is known as a hand-over or hand-off.
• Base stations perform calculations to determine when a user is crossing the cell
boundary, and to which cell (and base station) the user should be handed off to.
The process only takes fraction of a second (done by computer chips!).
• If the user moves too fast (say riding on a high-speed train or on a high way), the
call may be dropped!
• For early 1G analog cellular systems the speed limit is only ~100 km/hr. 3G
systems are designed to support a speed limit of ~ 200 km/hr. (a car moving in
a highway ~ 100 km/hr; high-speed rail ~ 400 km/hr; airplane ~ 800 km/hr.) So
we can hardly make even 3G phone calls when riding on a high-speed rail!
• If the handoff is transferred to a cell served by a different MTSO, the process is
more complicated.
• Hand-off also imposes a lower practical limit to cell size as we do not want the
system to be doing hand-offs all the time.

P.26
Hand-off Sometimes the
hand-off involves a
As you travel, the different MTSO ->
signal is passed more complex
from cell to cell. entering Cell 3

leaving Cell 2 Cell 4

Cell 3
entering Cell 4

entering Cell 2
Cell 2
MTSOB
leaving Cell 1

MTSOA to other MTSOs

Cell 1
P.27
 Roaming
• Roaming is when a user carries the mobile phone to a different
city/country served by different operators (e.g., when you cross the
border from HK to Mainland). The visited network automatically talks to
your home network to verify your service subscription and to
acknowledge your current location so that the home network can forward
incoming calls to you.
• Your handset needs to be compatible to the cellular system and
frequency bands used in the host operator! (e.g. Japan and Korea use
different frequency bands from most other countries)

served by Hong Kong operators served by Mainland operators

Cross the boarder

P.28
 Summary – Lecture 19
• Explained the cellular principle
• Described the schematic of a cellular and PSTN network
• Described the issue of frequency planning/reuse, cell size,
and network capacity
• Described the process of handoff and roaming

P.29
CORE1240 Electronic & Information Technology

Lecture 20 - Multiple Access

Technologies

How do we allow many subscribers to access the

network at the same time?

P.30
Frequency-Division Multiple Access (FDMA)
• Recall the concept of frequency translation by
frequency mixing in Chapter 1
• Frequency-Division Multiple Access (FDMA) is a scheme
that divides the total frequency spectrum bandwidth into
narrow frequency bands so that multiple users can
transmit information at the same time.
• For example, many FM radio stations broadcast their
baseband signals at the same time in different frequency
bands with 200 kHz in bandwidth
75kHz 25kHz
200kHz

88M 88.2M 88.4M 107.8M 108M

P.31
FDMA in 1G Cellular Networks

• 1G cellular networks used

FDMA. Frequency channels
25MHz
were assigned in pairs: one 30
kHz channel for uplink and one
30 kHz channel for downlink.
The mobile phone is assigned
a duplex channel dynamically 45MHz
for each call.
• The duplex channel contains
an uplink and a downlink that
25MHz
are separated by 45 MHz. This
is known as frequency division
duplex.

P.32
FDMA in 1G Cellular Networks

• 1G cellular networks used

FDMA. Frequency channels 25MHz
were assigned in pairs: one 30
kHz channel for uplink and one
30 kHz channel for downlink.
The mobile phone is assigned
a duplex channel dynamically 45MHz
for each call.
• The duplex channel contains
an uplink and a downlink that
25MHz
are separated by 45 MHz. This
is known as frequency division
duplex.

P.33
Example: analog cellular systems (1G mobile)

• We illustrate how to determine the capacity of a cell using a

simple 1G example.
• Typically the government allocate spectrum license in pairs
as follow:
Uplink frequencies: 824 MHz – 849 MHz (25 MHz bandwidth)
Downlink frequencies: 869 MHz – 894 MHz (25 MHz bandwidth)
Channel size: 30 kHz (modulated voice + guard band)
Control & signaling: Uses multiple channels at 9.6 kbps

(Each uplink and downlink frequency pair separated by 45 MHz.)

P.34
Example: number of available voice channels

• The 25 MHz bandwidth (for uplink or downlink) with 30 kHz per

channel gives
25 M / 30 k ~ 832 channels (uplink & downlink in pair)

• Assume the 832 duplex channels are shared by two service providers
(operators A, B) to avoid monopoly. Each operator gets
832 / 2 = 416 channels

• Assume an operator uses a frequency reuse factor of 4. Then 416/4 =

104 frequency channels are available for each cell.

• Assume 21 channels are needed for control and signaling (instruction

signals exchanged between the handset and the base station), the
number of simultaneous voice calls in 1G cellular allowed in each cell
is:
104 – 21 = 83 channels (i.e. the 84th simultaneous caller in this
cell will receive a network busy signal
and cannot access the network!)

P.35
Advanced Multiple Access Methods: TDMA & CDMA

• FDMA is the earliest method for multiple access. FDMA is

basically the only available method for analog systems.
• Digital communications (2G and beyond) open up other
possibilities:
- GSM networks (2G) use Time Division Multiple Access
(TDMA) in addition to FDMA
- CDMA networks (3G) use Code Division Multiple-Access
(CDMA)
- 4G networks use Orthogonal Frequency-Division Multiple
Access (OFDMA). Goes back to FDMA but in a highly
advanced form! Users use many finely spaced frequency
channels to transmit data in parallel. We will not describe
OFDMA in this course.
P.36
Time-division multiple access (TDMA)
• Multiple conversations are sent in the same frequency
channel but at different time slots. TDMA is made possible
because of digital signal processing.
• 2G systems such as GSM still use multiple frequency
channels, but use TDMA within individual frequency
channels to transmit multiple voice channels within each
frequency channel. So, they use both FDMA and TDMA.
With fewer frequency channels, transceiver (transmitter +
receiver) cost can be reduced (fewer frequency mixers for
modulation / demodulation).

So, why don’t we just use one frequency channel with large bandwidth and
high data rate? The reason is that the efficiency may drop because there
must be guard times provided for each TDM channel as users cannot be
perfectly synchronized (users at different distances from the base station
have different time delays sending signals to the base station). P.37
 TDMA in GSM

• GSM is the most popular 2G standard in the world.

• The allocated up and down stream spectrum (most commonly 25 MHz
each) is first divided into disjoint frequency channels of 200 kHz using
FDMA. Each frequency channel supports a bit rate of 270.85 kbps.
• GSM further divides each frequency channel into 8 time slots, so that up
to 8 users can take turn to access the network in each frequency
channel. This is TDMA.
• Each time channel has a theoretical data rate of 270.85 kbps / 8 =
33.85 kbps. But in practice, only 14.4 kbps is used to carry voice data
because of guard times, other overheads, and as many error correction
bits as data bits (1/2 rate error correction coding)!
• In 1G, each user (1-way) takes a 30 kHz channel. In GSM, there are 8
users per 200 kHz (i.e. 25 kHz per user for 1-way link). Spectrum
efficiency in GSM is therefore 16% better compared with 1G.

P.38
 TDMA – different time slots to different users
• In a GSM frequency channel, “time” is partitioned into frames with
duration of 4.6 ms each
• Each frame is divided into 8 time slots of 4.6ms/8 = ~0.57ms
• In downstream direction, the base station transmits to 8 mobiles in
round-robin fashion, for 0.57ms (including guard times) at a time.

TDMA – Downstream (Station-to-Mobile)

10100010 …

8 time slots
= 1 “frame”
of 4.6 ms

P.39
 TDMA – Synchronization
• In the upstream direction, different mobiles are at different
distances from the base station and are moving around. It is
difficult to perfectly synchronize their transmissions. So there is a
guard time of 0.03 ms (9 km of propagation) in each time slot to
prevent collision of signals from different mobiles

TDMA – Upstream (Mobile to Station) Signal Arrival at Base Station

Time Slot
0 1 2 3 4 5 6 7 0 1 2 3 4 5

0.03 ms
0.57 ms
Signals from different users arriving with different delays from Time
time slot boundary (guard time exaggerated for illustration)
P.40
 TDMA Summary
TDMA in GSM
User A
• In TDMA, different users communicate (t1, t9, t17, …)
with the base station in different
assigned time slots User B
(t2, t10, t18, …)
• “Time Division” is very common in User C
digital communication networks: (t3, t11, t19, …)

– “T1” transmission systems in the T1 transmission system

telephone network backbone group
together 24 voice channels and transmit 125 us frame,
8 bits from each channel at a time; 24 channels, 8 bits per channel
– routers take turn to transmit data packets
Internet Router
in the Internet, etc.

…..and many other examples

P.41
Code-division multiple access (CDMA)

Other than time (TDMA) and frequency (FDMA), what else

can be used to allow multiple access?

We can code different messages differently so that multiple

coded messages can be simultaneously transmitted on the
same frequency channel at the same time. This is known
as Code-Division Multiple Access (CDMA), and it offers
many advantages.

All 3G and some 2G mobile phone systems are based

on CDMA.

P.42
 Spread each bit with a code
• CDMA is a form of Spread Spectrum technology which was first
developed during the World War II (invented by a Hollywood actress)
to avoid detection and jamming of communication signals; a
classified technology until 1980s.

• CDMA works by “spreading the signal” by multiplying a slow data

rate information stream by a much higher data rate code. There are
different variants of spread spectrum.

• Direct sequence spread spectrum is realized by multiplying the data

bit stream by a Pseudo random Noise code (PN code) that runs at m
times the data rate. To detect what is being transmitted, a receiver
must know the PN code. If the receiver knows the PN code, the
receiver can pick out the transmitted signal even if it is very weak. If
the receiver does not have the PN code, everything just appears like
random noise.

P.43
 Realization of Direct Sequence CDMA
• Direct sequence spread spectrum is realized by multiplying the
data bit stream by a Pseudo random Noise code (PN code) that
runs at m times the data rate.

 Each bit of the PN code is called a chip.

 The code rate is called the chip rate.
 An information bit rate of d bps will result in a chip rate of md cps.

• In 3G cellular systems m ranges from 4 to 512

• For Global Positioning System (GPS) satellites, m equals 1024.
The PN code allows our GPS receiver to distinguish signals from
different GPS satellites, and the large value of m enables the
receiver to detect the satellite signals even as they are extremely
weak!

P.44
Direct Sequence Spread Spectrum (DSSS) Example

Below, we illustrate DSSS by spreading each data bit by m=4

chips. we assume that a binary “1” is represented by a signal
level of +1 and a binary “0” by a signal level of -1

+1
1 0 1 1
data bit stream t
-1
x 4 chips
1
PN code (1011) 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
t
-1
||
“chip stream” +1 1 0 1 1 0 1 0 0 1 0 1 1 1 0 1 1
transmitted
t
-1

P.45
 CDMA – Decoding via Match Filtering
• To recover the data bits at the receiver, you must know the PN
code and perform a match-filtering of the received signal.
• Match-filtering means you multiply the received signal with the PN
code and sum up (integrate) the result over all m chips in a bit.
• If the correct (a matched) PN code is applied, the output of the
match filter will be an amplified magnitude of either m or –m. If a
wrong PN code is applied, the output will be zero or noise-like.
CDMA Decoder
(Match Filter)
Message bit

Message
CDMA Encoder
(Spreading) Encoded
* stream amplified

signal Communication Correct PN Code

bit stream X
Channel

Null signal

PN Code * or noise-like
?????
m chips per bit Incorrect PN Code

P.46
 Match Filtering with the correct PN code
• Match-filtered by the same PN code, the result for each bit will
be m or –m corresponding to a 1 or 0 bit transmitted:

Encoded chip 1 1 0 1 1 0 1 0 0 10 1 1 1 0 1 1
stream received t
-1
x
4 chips
PN code 1
1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
(1011) t
-1
||
1
Result sums to a 1 0 1 1
t
large magnitude -1
(summation/integration) 1+1+1+1=4 ->(1) 1+1+1+1=4 ->(1) 1+1+1+1=4 ->(1)
(-1)+(-1)+(-1)+(-1) =-4 ->(0)
P.47
 Match-Filtering with a wrong PN code
• If a wrong PN code is used, the output of the match filter
is designed to give nothing.
Example: assume we match filter the encoded signal with the wrong PN code 1101

Encoded chip 1 1 0 1 1 0 1 0 0 10 1 1 1 0 1 1
stream received t
-1
x
4 chips
wrong 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1
PN code t
(1101) -1
||
1
result 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1
sums to 0
t
-1
(summation/integration)1+(-1)+(-1)+1=0 1+(-1)+(-1)+1=0 1+(-1)+(-1)+1=0
(-1)+1+1+(-1)=0
P.48
Combining two CDMA signals
• When two users transmit at the same time, the signal in the
channel is the sum of their individual signals
User A, Message=(10), PNA=1011 -> Encoded = +1-1+1+1, -1+1-1-1
User B, Message=(11), PNB=1101 -> Encoded = +1+1-1+1, +1+1-1+1
1V
Encoded Signal A =
-1V t

1V
Encoded Signal B=
-1V t
Total Signal in Channel =
Encoded Signal A
+ Encoded Signal B 2V
= +2 0 0 +2, 0 +2 -2 0
-2V
t

P.49
 Decoding
Match filtering the total signal with PNA gives:
2V
Total Signal Received
(*assume no attenuation)
-2V
t

Multiply with PNA 1V

+ - + + + - + +
-1V t

2V

= t
= 10
-2V Which is
what user A
Summation => 2+0+0+2=4 0+(-2)+(-2)+0=-4 transmits!
->(1) ->(0)
P.50
 Decoding
Match filtering the total signal with PNB gives:
2V
Total Signal Received
(*assume no attenuation)

-2V
t

1V
Multiply with PNB
-1V t

=
2V
= 11
t Which is
-2V
what user B
transmits!
Summation => 2+0+0+2=4 0+2+2+0=4
->(1) ->(1)
P.51
 Orthogonality of PN Codes
• Why does match filtering an encoded signal with the wrong PN
code yields a null message? This is because the different PN
codes we have seen are orthogonal, meaning that they yield
zero when match-filtered.
For example: PNA* PNB = (+1 -1 +1 +1)*(+1 +1 -1 +1) = +1-1-1+1=0
where we use the symbol * to mean match filtering
(For those who know basic linear algebra, match filtering two PN
codes is really like taking the dot product of two vectors. If the two
vectors are perpendicular to each other – orthogonal – then the dot
product is zero.)
• If PNB is used to encode (multiply) a message bit b, match
filtering with PNA also yields zero
PNA* (bPNB)= b(PNA* PNB)= b x 0 = 0!

P.52
CDMA - Same frequency, same time, different codes
• Adjacent cells can use the same
frequency, eliminating the

User D
frequency planning problem. • CDMA network allocates
• The users can use the base
(code D) station at the adjacent cells different codes to users
• Code A can only decodes
Cell 2 information encoded by
code A
User A
(code A)
• Code A cannot decode
information encoded by
User B code B, C, D… and vice
(code B) versa. Trying to decode
without the correct code
User C will yield either a null
(code C)
message or noise.
Cell 1

P.53
 CDMA – Key Merits for Mobile Network
• Key merits:
1. Neighboring cells can now use the same frequency channel but
individually unique codes to avoid interference! This increases the
capacity significantly and eliminates the frequency planning problem in
network design. It also enables a CDMA phone to simultaneously
connect with two or more base stations, allowing “glitch-less” smooth
handoff.

2. Unlike FDMA and TDMA which have a hard limit on the number of
simultaneous calls, CDMA allows for soft capacity limit, which means that
the number of simultaneous calls can be increased flexibly at penalty of
increased interference. By taking advantage of the fact that not all users
are talking and causing interference all the time, even when they are in a
phone call, overall capacity can be increased.

• The first commercial CDMA system was launched in Hong Kong in 1995.

P.54
 Mobile Networks with CDMA
• In summary, as shown in the right, CDMA allows multiple users to use
the same frequency channel to transmit at the same time using
different codes. The decoder, by using the correct code, greatly amplifies
the intended information signal. Transmissions from other users appear
as a noise-like interference if the PN codes are not strictly orthogonal.

Note : near-far effect

P.55
CDMA is like different languages or different
instruments in an orchestra

• Sound Tracks in CD
– E.g. CD put two sound tracks on the L and R audio
channels
– If you select the stereo audio setting on your CD
player, you can hear two sound tracks simultaneously
– Yet, you can focus (or tune) to one track and treat the
other as noise!!
– CDMA receiver works the same way!!
• Different musical instruments in an orchestra
– A trained ear not only can tell which instruments are
there, but also the melody each instrument is playing!!

P.56
How can CDMA reuse the same frequency
channel in all cells?
Imagine a cocktail party where people are standing in small
ethic groups carrying on conversation within each group in
different languages：
• Groups are like Cells, can be big or small
• No coordination among them
• Even though there are interference (sound from other
groups), can still communicate
• As long as interference level is not too high, can sustain
all conversations
• Interference reduces immediately when there are gaps
between words or sentences

P.57
OFDMA
• OFDM (Orthogonal Frequency Division Multiplexing) is a
multi-carrier transmission scheme in 4G LTE that splits
the overall carrier bandwidth into smaller sub-carriers of
15 kHz each. OFDM is robust and has the ability to
exploit both time and frequency domains.
• Orthogonal frequency-division
multiple access (OFDMA) is a
multi-user version of OFDM
digital modulation scheme.
Multiple access is achieved in
OFDMA by assigning subsets
of subcarriers to individual users.
This allows simultaneous low-
data-rate transmission from
several users.
P.58
Massive MIMO
• Massive MIMO (massive multiple-input multiple-output)
is a type of wireless communications technology in which
base stations are equipped with a very large number of
antenna elements to improve spectral and energy
efficiency.

Active Phased-Array Antenna (APAA)

P.59
 Summary
1G 2G

3G 4G, 5G

NOMA Future?
(Non-orthogonal multiple access)

P.60
 Summary – Lecture 20
• Described FDMA as it is applied in 1G cellular network
• Described how TDMA is used on top of FDMA in 2G GSM
• Discussed the issue of synchronization of uplink data in
TDMA
• Described the basic idea and advantages of CDMA
• Use the Cocktail Party Analogy to understand CDMA
• Described the concept of OFDMA and Massive MIMO

P.61