Wireless specifics
A Wireless Communication System
Antenna
2
Technologies for Cell phones to Handle
Multiple Users
Unique feature for voice communications:
FDMA (Frequency Division Multiple Access) in
1G cellular phone technology
TDMA (Time Division Multiple Access, e.g.,
GSM—Global Service for Mobile
communications and IS-54, both use TDMA and
FDMA)
CDMA (Code Division Multiple Access, e.g., IS-
95, WCDMA, CDMA2000)
3
Technologies for Cell Phones to Handle
Up and Down Links
TDD (Time Division Duplex): forward
(down link) and reverse (up link) channels
use the same frequency band but
alternating time slots
FDD (Frequency Division Duplex): forward
and reverse channels use different carrier
frequencies
4
Technology used by WiFi, etc. to handle
multiple users
It is a time division method.
Unlike the TDMA in cell phones, in WiFi no
specific time slots are assigned to users.
Instead, it is basically a first-come-first-served
policy—users form a queue waiting for their
turns to use the connection. It’s very much like
the rule used in a bank or a computer network.
It is good for data transmission, but not that ideal
for voice transmission.
5
Techniques used by all
Spread Spectrum Transmission
FHSS (Frequency Hopping Spread
Spectrum)
DSSS (Direct Sequence Spread
Spectrum)
OFDM (Orthogonal Frequency Division
Multiplexing)
6
What is Spread Spectrum Transmission
The traditional transmitted signal has a
bandwidth of the same order as the information
signal at the baseband. For example, the
bandwidth of a voice signal is about 4 kHz (the
baseband). After modulation, as it being
transmitted it still occupies several thousand Hz
but at a much higher frequency.
The spread spectrum signal occupies a much
larger bandwidth.
7
Why spread spectrum
It is robust against frequency selective fading in
urban and indoor environments.
It is robust against interference emitted by
machines, microwave ovens, etc.
It can provide additional security.
It can provide greater operational flexibility and
system capacity, as in CDMA.
It is required by regulation to use spread
spectrum in unlicensed ISM (Industrial, Scientific
and Medical) bands.
8
Spread spectrum methods
Frequency hopping spread spectrum (FHSS)
The transmitter constantly, often randomly, shifts the
center frequency of the transmitted signal.
Only the machine that knows the hopping pattern can
receive the signal.
Direct sequence spread spectrum (DSSS)
Each transmitted bit is spread into N smaller pulses
(chips) before transmission.
Only the machine that knows the pattern of the spread
can retrieve the signal.
9
FHSS
Invented by Austrian-born movie star Hedy Lamarr to
protect guided torpedoes from jamming
The shifts in frequency (hops) occur according to a
random pattern that is known only to the transmitter and
the receiver. (Actually it is pseudo random: It is
generated by an algorithm, so it is not really random; but
to a person who does not know the pattern, it looks
random.)
If the center frequency moves among 100 different
frequencies, the required transmission bandwidth is at
least 100 times as large as the original transmission
bandwidth.
10
Example of FHSS
11
FHSS and GSM (p.115, Example 3.14)
If the channel coincides with a deep
frequency selective fading or when the
cochannel interfence from another cell
using the same frequency is excessive,
the distortion in the received voice signal
will be large. A slow frequency hopping of
217.6 hops per second can be used in
GSM to tackle these problems.
12
FHSS in 802.11 or WiFi (p. 115, Example
3.15)
Uses 78 hopping channels each separated
by 1 MHz. These frequencies are divided
into three patterns of 26 hops each
corresponding to channel numbers (0, 3,
6, 9, …, 75), (1, 4, 7, 10, …, 76), (2, 5, 8,
11, …, 77). These choices are available
for three different systems to coexist
without any hop collision.
2.5 hops per second
13
FHSS in Bluetooth (p.129, Example 3.24)
Uses a fast frequency hopping (1,600 hops
per second) over 79 MHz of bandwidth.
That is, it hops over 79 channels each
separated by 1 MHz.
14
DSSS
The transmitter spreads one bit, say a one or a zero (you
either have a one or a zero in the digital world), into
many (N) smaller chips (they are a sequence of zeros
and ones) according to a code known to the transmitter
and the receiver.
The receiver, using the code and a correlator, put the
spread chips together to get the original bit.
The bandwidth of the original signal will be N times wider
after the spreading because the chip rate is N times
faster than the bit rate. Therefore the signal will be more
robust against interference and fading.
The code used for spreading and de-spreading can be
secret, if only the transmitter and the receiver know it,
thus providing a security measure.
15
DSSS in 802.11
The code (called Baker code) used in 802.11 to spread
the data bits is given by [1, 1, 1, -1, -1, -1, 1, -1, -1, 1, -1].
So one bit of data is spread to become 11 chips.
The Baker code is not a secret code, so it’s not used for
security. It’s used to spread the bandwidth.
16
DSSS in 802.11 (p. 117, Fig. 3.23)
17
More about DSSS
The bandwidth of the transmitted DSSS signal is
N times as large as that of the original signal.
CDMA uses DSSS. Each user is given a unique
code that other users don’s know. Although a
user can receive the signals sent to other users
by the transmitter, in the de-spreading process
only the signal sent to that user can be detected.
The interference generated by other users is
very small.
18
OFDM (Orthogonal Frequency
Division Multiplexing)
What is OFDM
Assume we need to send data at a speed of R
symbols/sec.
We break the data sequence into N (an integer,
say, 48) sub-sequences. The data rate of each
sub-sequence will be R/N, much slower than the
original sequence.
N carriers are used, each having a different
frequency and each sending one sub-sequence.
At the receiver end, the N sub-sequences are
put together to get the original data sequence.
20
Advantages of OFDM
Robust against multipath interference
because:
In each sub-sequence the symbols are N times
longer than the original symbols.
The longer the symbol, the weaker the
multipath interference (the signals representing
the same symbol but coming from multiple
paths will be close enough compared with the
width of the symbol so they don’t interfere with
each other).
21
Advantages of OFDM
Robust against frequency selective fading
To battle the frequency selective fading (signals
at certain frequencies might be much weaker
than that of other frequencies), error-control
coding can be used in each subchannel.
If the signal for a particular subchannel(s) is
weak, the transmitting power of that subchannel
can be increased to compensate for the fading.
High spectral efficiency (high bit rate to
bandwidth ratio)
22
Drawbacks of OFDM
Complexity: you need to put together N
signals to rebuild the original one.
Sensitive to Doppler Shift. When receiver
is moving at a high speed, the received
frequency will shift, too, and that can
cause problem for OFDM.
High peak-to-average-power ratio (PAPR)
and thus lower efficiency.
23
Example: OFDM in 802.11a (p.109)
64 subchannels are used, among which
48 are used for data transmission, the
remaining 16 are for other purposes.
The symbol rate of each channel is 250
kilo symbols per second (ksps).
The actual data rate for the user is 48X250
ksps = 12 Msps.
The overall bandwidth is 20 MHz.
24
SOFDMA (Scalable Orthogonal
Frequency Division Multiple Access)
Used in 802.16e (WiMax for mobile users)
The same OFDM technique will be used,
but each user may only get a part of the
spectrum, depending on the application
the user is running. TDMA is also used.
In 802.16-2004 (WiMax for fixed users)
OFDM is used and a user get all the
available spectrum. Users are separated
by TDMA.
25
Assigning sub-channels
26
Diversity
Time diversity
Frequency diversity
Space diversity
27
Time diversity: DSSS and RAKE receiver
Using the signals from different paths to get one
stronger signal. Signals from different paths
arrive at the receiver at different time instances.
Longer paths create longer delays.
Due to multiple paths, each bit sent by the
transmitter can create several peaks at the
output of the correlator (Fig. 3.25, p. 122).
The RAKE receiver is designed to put the peaks
together. (p. 124, Fig. 3.26)
28
Time diversity (contd.)
Multipath reception in CDMA
Chip rate: 1.25 Mcps, symbol rate: 4,800 Sps
Can resolve multipath components 1/1.25 Mcps
= 800 ms apart.
A multipath spread of up to 1/4800 bps = 2.08
ms cannot cause ISI.
29
Frequency diversity
Frequency selective fading (p. 128, Fig.
3.30)
Frequency hopping and IEEE 802.11
Frequency hopping and GSM
30
Space diversity
Four methods to take advantage of space
diversity (p. 132, Fig. 3.32)
Trisectored antennas for CDMA
31
Coding techniques
Error control coding
Coding for spread spectrum (CDMA)
Orthogonal codes
32
Voice-oriented and data-oriented
networks
Voice-oriented networks use the so-called fixed-
assignment methods. Each user is assigned a
slot of time, a portion of frequency band, or a
specific code for the entire length of the
conversation.
Data-oriented networks use random-access
methods. Users share the same medium (air or
wire). Since data arrive at random instances, the
medium will be assigned to each user in a
random fashion.
33
Comparison of two methods
Fixed assignment ensures constant
connection, which is needed for voice
communication, but can have low
utilization rate.
Random access method is more suitable
for data communication, because data
arrive in bursts.
34