WCDMA AIR INTERFACE

Chapter 1
WCDMA Wireless Technology
Objectives of Chapter 1, WCDMA
Technology
After this chapter the participants will be able to:

1. Explain the fundamental principles of cellular WCDMA technology.

2. Explain and compare TDMA and WCDMA multiple access methods.
3. Explain on an overview level, the WCDMA transmitter architecture.
4. Explain the data protection coding methods: CRC Coding, FEC Coding,
Viterbi decoding, block interleaving, turbo codes.
5. Explain the use of channelization and scrambling codes.
6. Explain the modulation and filtering in a WCDMA system.

. Figure 1- 1
WCDMA Air Interface

▪ World Administrative Radio Conference (WARC) of the ITU

(International Telecommunications Union) in 1992 chose
frequencies around 2 GHz as available for use by third generation
mobile systems.
▪ Within the ITU these third generation systems are called
International Mobile Telephony 2000 (IMT-2000).
▪ Within IMT-2000, several different air interfaces are defined for
third generation systems based on either CDMA or TDMA
technology.
▪ The same air interface,WCDMA, is to be used in Europe and Asia,
including Japan and Korea using the frequency bands around 2
GHz.
▪ In North America that spectrum has already been allocated for
operators using second generation systems and no new spectrum
is available for IMT-2000. Thus third generation services must be
implemented within the existing bands.

. Figure 1- 2
WCDMA Air Interface

▪ As well as WCDMA the other air interfaces that can be used are
EDGE and cdma2000.
▪ EDGE (Enhanced Data Rates for GSM Evolution) can provide bit
rates up to 500 kbps within a GSM carrier spacing of 200kHz.
▪ Cdma2000 can be used as an upgrade for the existing IS-95
operators
▪ Spectrum allocation in Europe, Japan and Korea is 1920 - 1980
MHz uplink and 2110 - 2170 MHz downlink for Frequency Division
Duplexing. 1900 - 1920 MHz and 2020 - 2025 MHz for Time
Division Duplexing.
▪ Frequency Division Duplex use different frequency bands for
uplink and downlink while Time Division Duplex use the same
frequency for both uplink and downlink.

. Figure 1- 3
From 2G to 3G
2G
Circuit-Switched Voice

Circuit-Switched Data

Short Message Service (SMS)

3G
Circuit-Switched AMR coded voice

Circuit-Switched data

Streaming

Packet Data

Multiservice: AMR coded voice + Packet data

. Figure 1- 4
 RABs
Conversational/speech RAB 12.2 kbps Circuit switched

64 kbps Circuit switched

CS RABs Conversational 64 kbps CS RAB

Streaming 57.6 kbps RAB 57.6 kbps Circuit switched

Maximum Bitrate 16/64

PS Streaming RAB Guaranteed Bitrate 8/54

PS RABs
Variable rate Packet Switched
Interactive or background PS RAB RACH/FACH, 64/64, 64/128, 64/384,
64/HS, 384/HS

Combination of Conversational
Multi-RAB Speech and Interactive 64/64

. Figure 1- 5
New P5 RAB combinations
CS Conversational Speech Multi Mode AMR

PS Interactive EUL/HS

Speech 12.2

PS Interactive 64/64

PS Interactive 64/64

PS Interactive 64/64, 64/128

PS Interactive 64/64, 64/128

Speech 12.2 kbps

PS 64/128, 128/64, 64/384, 64/HS, 384/HS

. Figure 1- 6
Multiple Access Approaches
Frequency Time Division Spread Code
Division Multiple Spectrum Division
Multiple Access Multiple Multiple
Access Access Access

Multiple
Transmitters

User N
User 3
User 2
User 1
and
User 2

User 3
User 1

Multiple Data
Channels

Frequency Time Frequency

Each User has a unique Each User has a unique Each Transmitter has a unique
frequency time slot Scrambling Code

(1 voice channel per user) Each Data Channel has a unique

Each Data Channel has a unique
position within the time slot
Channelization code
Many users share the same frequency
All users transmit at the same Several users share the
and time
time same frequency

IS-95, cdma2000, WCDMA

IS-136, GSM, PDC
AMPS, NMT, TACS

. Figure 1- 7
The TDMA Transmitter
The Multiplexer allows various data channels
Sync. to share the same timeslot.
Bits
The timeslot selector allows multiple
transmitters to share the same carrier
Control/ Error frequency, by assigning a unique timeslot to
Signaling Protection each transmitter.
Data
Timeslot
Selector

Filtering RF
Error Data
Vocoder Transmit + Out
Protection Multiplexer
Gating RF
Modulation

User Data Error

Channel 1 Protection

User Data Error

Channel N Protection

. Figure 1- 8
The WCDMA Transmitter
Channelization Codes provide unique
Channelization identification of each data channel
code 1 Scrambling
Code

Sync.
Scrambling Codes (SC) provide
Bits
unique identification
Channelization Scrambling
code 2 Code
of each transmitter

Control/ Error
Signaling Protection
Data
Channelization Scrambling
code 3 Code

Error Linear Filtering RF

Vocoder Protection + Out
Summation
RF
Channelization Scrambling
Modulation
code 4 Code

User Data Error

Channel 1 Protection

Channelization Scrambling User 1

code N Code User 2
User 3

User Data Error ...

Channel N Protection Frequency

. Figure 1- 9
 WCDMA Transmitter (downlink)
Pre-coded data Data
(bits) Symbols Chips
Modulation
Symbols
Channelization scrambling Code
Code 1 1
Data CRC Coding FEC Inter- I I
Channel 1:2
Coding leaving Demux D/A
1
Q Q

I Pulse I


Shaping
Filter
RF Out
I/Q Modulator

Pulse
Channelization scrambling Code Q Shaping Q
Code n 1 Filter
Data CRC Coding FEC Inter- I I
Channel 1:2
Coding leaving D/A
N Demux
Q Q

Modulation Mapping

Allows for error Allows for Improves error Provides 2x Maps binary Gives a unique Gives a unique Contains Allows both signals
detection in the error correction higher data bits to real identity to each identity to this transmitted from I and Q branch to
receiver correction in in the receiver rate value symbols data stream transmitter frequency share the same RF
the receiver spectrum bandwidth
(WCDMA, 0 → +1
cdma2000
downlink) 1 → -1

. Figure 1- 10
Voice Coding

▪ Example: Two ways to hear the sax player

Record the sax player onto a CD... ... and play back the CD

20 MB per song

Write down the notes he plays... ... and have a friend play the same notes

20 kB per song

. Figure 1- 11
Voice Coding
▪ Vocoding
Human Voice:

‘ss’, ‘ff’, ‘sh’ … ~20% of time

‘ah’, ‘v’, ‘mm’ , … ~80% of time
Voice Re-Synthesis at the Receiver

H(s)
White Noise Generator

Noise Speech
Vocoder parameters Output

 H(s)
Pulse Generator
Filter poles
Pitch
correspond to
parameters
resonances of the
vocal tract

Transmitted Parameters
8~12 kb/s typical,
vs.
64 kbps for log-PCM
32 kbps for ADPCM

. Figure 1- 12
ACELP/AMR Voice Coding
A/D

Linear
Predictive Voice, Tone
Coding Activity
(LPC) Detectors

Speech
Generator • Mode Indication bits
(+)
• Comfort Noise
(-)
Codebook Filter • Tone Emulation

• DTX Indication

Codebook Prediction
Index Error Perceptual Error
Analysis Weighting

Vocoder
MUX
Benefits of Activity Detection: Output Bits
1)

2)

. Figure 1- 13
AMR (Adaptive Multi-rate)

▪ The speech service in UMTS will employ the Adaptive Multi-rate

technique.
▪ This is a single integrated codec with eight source rates: 12.2, 10.2,
7.95, 7.40, 6.70, 5.90, 5.15 and 4.75 kbps. To facilitate
interoperability with existing cellular networks some of the modes are
the same as in existing networks. In P3 only the 12.2 kbps is working.

. Figure 1- 14
Digital Cellular Error Correction
Analog Cellular
Transmitted Signal Received Signal + Noise

Digital Cellular
Transmitted Signal Received Signal + Noise

. Figure 1- 15
Digital Cellular Error Correction
▪ Example: Mailing a letter
– Extra (redundant) symbols in address help correct lost symbols
– ZIP codes used to detect errors in the address

With minimal data... With redundant data...

Errors are uncorrectable Errors are correctable

EM
Eddie McConnell
5 Main Street
5 Main Street
Littletown
Littletown LT1701

. Figure 1- 16
CRC Coding
▪ Cyclic-Redundancy Check (CRC) Coding
– Identifies corrupted data
– CRC is used for checking BLER (Block Error Ratio) in the outer
loop power control

Transmitter

Original Data CRC Original Data Checksum 12 bits

Generator 1001011010.. 110010110011
244 bits

RF
Transmission Path

Receiver

Received Data Received Checksum If Checksums do not match,

1001010010.. there is an error
110010110011

CRC Re-Generated Checksum

Generator 110010110001

. Figure 1- 17
CRC Algorithms
3GPP TS 25.212¶ 4.2.1.1

▪ CRC Algorithms
– 0, 8, 12, 16, or 24 parity bits (determined by upper layers)

▪ g(CRC24) = D24 + D23 + D6 + D5 + D + 1

▪ g(CRC16) = D16 + D12 + D5 + 1

▪ g(CRC12) = D12 + D11 + D3 + D2 + D + 1

▪ g(CRC8) = D8 + D7 + D4 + D3 + D + 1

. Figure 1- 18
FEC Coding

▪ Error Correction
– How do you correct errors at the receiver?
Send Forward
message Error
many times? Correction!

010010110,
010010110,
010010110,
010010110,
010010110,

•
•
•

Up to 6x data expansion...

But the most powerful results

. Figure 1- 19
FEC Coding Approaches
– Block Codes (Hamming Codes, BCH Codes, Reed-Solomon Codes)
▪ Data is processed into unique Codewords
▪ Each Codeword can be positively identified even if one or
more bits are corrupted
▪ Example: “Little Town” is a code word for “LT”.
– Continuous Codes (Convolutional Codes, Turbo Codes)
▪ Data is processed continuously through FEC generator
▪ Resulting data stream has built-in redundancy that can be
extracted to correct bit errors.
– IS-95, cdma2000, and WCDMA utilize Convolutional Codes for the
services speech and signaling
▪ Powerful error correction
▪ Simple implementation allows low-latency, real-time
processing
– cdma2000 and WCDMA utilize Turbo Codes for all other services
▪ Most powerful error correction
▪ More processing power (MIPS) required for decoding

. Figure 1- 20
FEC Coding

Transmitter

Original Data FEC FEC Encoded data

00011011... Generator 1010011100110110...

RF
Transmission Path

Receiver
Viterbi/ Original Data
Turbo 00011011
Decoder

. Figure 1- 21
FEC Coding: Convolutional Coder
example
X2k

clock
MUX Coder Output

Input Data 1010... D D

X2k+1

R = 1/2 , k=2 Convolutional Coder

• For every input bit, there are two output bits

• The maximum time delay is 2 clock cycles

. Figure 1- 22
 FEC Coding: Convolutional Coder
State Diagram
▪ FEC Coding: Example x2k x2k+1 = Coder Output
X2k = (Dk) XOR (Dk-2)

X2k+1 = (Dk) XOR (Dk-1) XOR (Dk-2)

State [00] 00 State [00]
Clock Current Delayed Outputs
11
Cycle Input Inputs

Dk Dk-1 Dk-2 X2k X2k+1

11
1 0 0 0 0 0 State [01] State [01]
00
2 1 0 0 1 1

3 0 1 0 0 1
01
4 1 0 1 0 0
State [10] State [10]
5 1 1 0 1 0 10

6 1 1 1 0 1
10
7 0 1 1 1 0
State [11] 01 State [11]
8 0 0 1 1 1

STATE

. Figure 1- 23
WCDMA Convolutional Code
Generators
Rate 1/2, k=9 coder: G0 = 5618 , G1 = 7538
3GPP TS 25.212¶ 4.2.3.1

Data
In
D D D D D D D D

2:1 Data
MUX Out

Rate 1/3 , k=9 coder: G0 = 5578 , G1 = 6638 , G2 = 7118

Data
In
D D D D D D D D

3:1 Data
MUX Out

. Figure 1- 24
FEC Coding: Viterbi Decoder
Example:
Viterbi Decoding Process: Received Signal R1,R2 = [0 1] = Branch Metric

1) Calculate Branch Metric for each possible

state transition State [00] T1,T2 = [0 0] 1 State [00]

BM = (|R1 - T1| + |R2 - T2|)2

T1,T2 = [1 1] 1
R1 , R2 = Received data values

T1 , T2 = Transmitted data values

T1,T2 = [1 1]
1
2) Calculate Cumulative Path Metric State [01] State [01]

Path Metric is sum of “N” previous T1,T2 = [0 0] 0

Branch Metrics (N is memory depth
of Viterbi Decoder).

T1,T2 = [0 1] 1
3) Calculate surviving Path
State [10] 4 State [10]
The surviving path is the path
with the lowest Path Metric. T1,T2 = [1 0]

4) Extract the error-corrected Data T1,T2 = [1 0] 4

The error-corrected data sequence

is equal to the first bit of each state State [11] T1,T2 = [0 1] State [11]
code along the surviving path 0

. Figure 1- 25
Viterbi Decoding (No noise)
Transmitted Data: 11 01 00 10

Received Data: 11 01 00 10

[0 0] 4 1 0 1 [0 0]

1
1 4
0

[0 1] 0 [0 1]

0 1

0
1 4

[1 0] 0 0 [1 0]

1 0 Path with lowest path

4 metric has the least
likelihood of error

[1 1] 4 0 [1 1]
1

Output --->> 0 1 0 1 1

. Figure 1- 26
Viterbi Decoding (With noise)
Transmitted Data: 1 1 0 1 0 0 1 0

Received Data: 1.1 0.8 -.3 1.2 0.6 0.5 0.8 0.3

[0 0] 3.61 2.25 1.21 1.15 1.21 2.36 [0 0]

0.81
0.81 0.81 0.81
2.25
0.09

[0 1] .34 3.80 [0 1]

1.21
1.21

.25 2.25

[1 0] .09 1.55 1.96 [1 0]

6.25

0.25

[1 1] 1.80 [1 1]

Output --->> 0 1 0 1 1

. Figure 1- 27
Multipath Fading

▪ One user’s signal reflects off many objects

▪ The received signal contains many time-delayed
replicas

. Figure 1- 28
Multipath Fading

Direct Signal

Reflected Signal

Combined Signal

. Figure 1- 29
Block Interleaving
Transmitter
Interleaving
Matrix
Original Data Samples Interleaved Data Samples
1 2 3 4 5 6 7 8 9 1 2 3 1 4 7 2 5 8 3 6 9
4 5 6
7 8 9

RF
Transmission Path
Amplitude

Receiver
De-
Time Interleaving
Interleaved Data Samples Matrix De-Interleaved Data Samples To Viterbi
1 4 7 2 5 8 3 6 9 1 2 3 1 2 3 4 5 6 7 8 9 decoder
4 5 6
Errors Clustered 7 8 9 Errors Distributed

. Figure 1- 30
 Interleaving
3GPP TS 25.212 ¶ 4.2.5 , 4.2.11

▪ Interleaving

– 1st-Stage Interleaver
▪ Performed prior to service multiplexing

▪ Interleaving depth of 1, 2, 4, or 8 columns. (10,20,40 or 80

ms)

– 2nd-Stage Interleaver

▪ Performed after service multiplexing

▪ Interleaving depth of 30 columns (always 10 ms)

. Figure 1- 31
Turbo Coding
– Outperform Convolutional codes
▪ Requires much more processing power; data packets may be
decoded off-line
▪ Used for CS64, streaming 57.6, Interactive (64/64,64/128,
64/384)
– Interleaving (time diversity) enhances error correction

Turbo Encoder Turbo Decoder

D D Decoded
Data Data
Decoder #1
P1

De-Interleaver
Interleaver
DE-
MUX Interleaver
P1 MUX
Encoder #1

D
Interleaver Encoder #2
P2
Decoder #2
P2

. Figure 1- 32
WCDMA Turbo Code Generator
3GPP TS 25.212¶ 4.2.3.2

Xk

Zk

Data In
Rate = X
Xk D D D

M
U Data Out
X
Turbo 3x input bits
Z’k + 12 Termination bits
Interleaver

D D D

X’k

At end of data block, both switches go “down” to provide 12-bit Trellis Termination:
[ xK+1, zK+1, xK+2, zK+2, xK+3, zK+3, x'K+1, z'K+1, x'K+2, z'K+2, x'K+3, z'K+3 ]

. Figure 1- 33
Rate Matching
3GPP TS 25.212 ¶ 4.2.7

– When coded data rates of services are incompatible,

“Rate Matching” is used to equalize the data rates.

– Rate Matching may be performed by:

▪ Padding with extra bits

▪ Puncturing of bits using a pseudo-random algorithm

– For complete rate matching rules, see 3GPP TS25.212 ¶ 4.2.7

. Figure 1- 34
WCDMA Codes

▪ Channelization codes (also sometimes called orthogonal codes, short codes,

Walsh codes or Spreading codes)

Allows multiple data channels to be sent from each transmitter (cell or UE)

▪ Scrambling Codes (also sometimes called PN codes, Spread Spectrum

Multiple Access Codes, Long codes or Spreading codes):

Allows multiple WCDMA transmitters to share the same Radio Frequency

. Figure 1- 35
 Code Correlation
Case III: Correlation using channelization codes

(a) Same channelization code; (b) Different channelization codes; (c) Same code with non-zero time offset

Input Data +1
+1 -1
x x x
Channelization code
in Transmitter -1 +1 –1 +1 +1 –1 +1 -1 -1 +1 –1 +1 +1 –1 +1 -1 -1 +1 –1 +1 +1 –1 +1 -1
Transmitter
= = =
Transmitted
-1 +1 –1 +1 +1 –1 +1 -1 +1 –1 +1 –1 –1 +1 –1 +1 -1 +1 –1 +1 +1 –1 +1 -1
Sequence

x x x
Channelization Code
-1 +1 –1 +1 +1 –1 +1 -1 +1 +1 +1 +1 +1 +1 +1 +1 -1 -1 +1 –1 +1 +1 –1 +1
used in Receiver
= = =
+1 +1 +1 +1 +1 +1 +1 +1 +1 –1 +1 –1 –1 +1 –1 +1 +1 –1 –1 –1 +1 –1 –1 -1 Receiver

Integrate Integrate Integrate

Integrate
Result 8 0 -4

Divide by
Code Length +1 0 -0.5

. Figure 1- 36
Code Correlation: Key Points

▪ TX, RX use same codes, at the same time offset

▪ Channelization Codes: 100% correlation

▪ TX, RX use different codes

▪ Channelization Codes: 0 % correlation (perfect separation)

▪ TX, RX use same codes, but at different time offsets

▪ Channelization Codes: Unpredictable results (orthogonality

is lost)

. Figure 1- 37
Decoding with Channelization
Codes (CC)
Transmitter

CC 1
Data Channel 1

CC 2
Data Channel 2

Receiver CC 3

Linear RF RF
CC 3 Addition Modulation Demod
Data Channel 3

In this example, the receiver correlates

the composite received signal using
CC 4 Channelization Code 3.
Data Channel 4
The result is a perfect reconstruction of
Data Channel #3, with no interference
from the other data channels.

To realize this perfect cross-correlation

property, it is essential that the
channelization codes be received in
perfect timing relation to each other.

. Figure 1- 38
Sending data using channelization
codes
Channelization code

Transmitted “chips”
Data

User 1 Data: D/A conv. Multiply with channelization Code User 1 channelization coded data:
1 0 1 -1 +1 -1 1 –1 1-1 -1 1-1 1 1-1 1-1 -1 1-1 1

You send one channelization code for every data bit!

If you want to send a digital “0”, you transmit the assigned channelization
code

If you want to send a digital “1”, you transmit the inverted channelization
code

. Figure 1- 39
Channelization coding example -
Transmitter
Data Channel 1 After D/A Mapping Multiply with CC1 After channelization coding
010 +1 –1 +1 (1 1 1 1) (+1+1+1+1)(-1-1-1-1)(+1+1+1+1)

Data Channel 2 After D/A Mapping Multiply with CC2 After channelization coding
001 +1 +1 –1 (1 1-1-1) (+1+1-1-1)(+1+1-1-1)(-1-1+1+1)


Data Channel 3 After D/A Mapping Multiply with CC3 After channelization coding
101 -1 +1 -1 (1–1 1-1) (-1+1-1+1)(+1-1+1-1)(-1+1-1+1)

Data Channel 4 After D/A Mapping Multiply with CC4 After channelization coding
000 +1 +1 +1 (1-1-1 1) (+1-1-1+1)(+1-1-1+1)(+1-1-1+1)

Composite Transmitted Data:

(+2 +2 -2 +2) (+2 -2 -2 -2) (0 0 0 +4)
4-chip
Channelization Code Set
1) 1 1 1 1
2) 1 1 -1 -1
3) 1 –1 1 -1
4) 1 -1 -1 1

. Figure 1- 40
 Channelization decoding example
– Receiver 4-chip
Channelization Code Set
1) 1 1 1 1
2) 1 1 -1 -1
Composite Received Data: 3) 1 –1 1 -1
(+2 +2 -2 +2)(+2 -2 -2 -2)(0 0 0 +4) 4) 1 -1 -1 1
“Correlation”

Multiply with CC1 Integrate & Result: Map A→D

(+1 +1 +1 +1) Normalize 1 -1 1 010

Multiply with CC2 Integrate & Result: Map A→D

(+1 +1 -1 -1) Normalize 1 1 -1 001

Multiply with CC3 Integrate & Result: Map A→D

(+1 -1 +1 -1) Normalize -1 1 -1 101

Multiply with CC4 Integrate & Result: Map A→D

(+1 -1 -1 +1) Normalize 111 000

Integrate: Sum four consecutive values after multiplication with CC.

Normalize: Multiply by [ 1 / code length]

. Figure 1- 41
Channelization Codes

Downlink: Channelization Codes used to distinguish data channels

coming from each cell

CC3, CC4
CC1, CC2

CC5, CC6, CC7

Uplink: Channelization Codes used to distinguish data channels

coming from each User Equipment, UE

CC1, CC2
CC1 , CC2, CC3

CC1, CC2, CC3, CC4

. Figure 1- 42
Generation of Channelization
Codes
1

11 1-1

1111 11-1-1 1-11-1 1-1-11

11111111 1111-1-1-1-1 11-1-111-1-1 11-1-1-1-111 1-11-11-11-1 1-11-1-11-11 1-1-111-1-11 1-1-11-111-1

11-1-111-1-111-1-1 11-1-1
Digital/Analog Mapping

logic 0  analog +1
logic 1  analog - 1

. Figure 1- 43
Usage of the Channelization Code
Tree - (DL example)
1 Chip Rate = 3.840 Mcps

11 1-1

1111 11-1-1 1-11-1 1-1-11

11111111 1111-1-1-1-1 11-1-111-1-1 11-1-1-1-111 1-11-11-11-1 1-11-1-11-11 1-1-111-1-11 1-1-11-111-1

480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s

. Figure 1- 44
Usage of the Channelization Code
Tree – (DL example)
1
Chip Rate = 3.840 Mcps
User with 4x Bit Rate

11 1.92 Msymb/s 1-1

1111 11-1-1 1-11-1 1-1-11

11111111 1111-1-1-1-1 11-1-111-1-1 11-1-1-1-111 1-11-11-11-1 1-11-1-11-11 1-1-111-1-11 1-1-11-111-1

480 ksymbol/s 480 ksymbol/s 480 ksymbol/s 480 ksymbol/s

= Unusable Code Space

. Figure 1- 45
Channelization Codes: Summary
Code
Division
Multiple
Access ▪ WCDMA allows multiple data streams to
be sent on the same RF carrier
Data 1
– Perfect isolation between data
Data 2 streams
Data 3 – Timing between data streams must be
... exact
– Maximum number of data channels =
Frequency
Channelization code length
Each Data Stream has ▪ The longer the code, the
a unique
Channelization Code
slower the data rate

Many users share the same ▪ WCDMA advantages are limited in

frequency and time practice
– Multipath, small timing errors, and
IS-95, cdma2000,
WCDMA motion-related effects diminish the
usable code space

. Figure 1- 46
The WCDMA Cocktail Party
What do YOU hear...
• If you only speak Japanese?
• If you only speak English?
• If you only speak Italian?
• If you only speak Japanese, but the Japanese-speaking
person is all the way across the room?
• If you only speak Japanese, but the Spanish-speaking
person is talking very loudly?

. Figure 1- 47
Spread Spectrum Multiple Access

SC 1
Transmitter 1

RF
Modulation

SC2 Receiver
Transmitter 2
SC3
RF
Modulation
RF
Demod

SC3
Transmitter 3

RF In this example, the receiver correlates

Modulation
the composite received signal using
Scrambling Code (SC) 3.

SC4 The result is the recovered transmission

Transmitter 4 from Transmitter #3, plus some spread
spectrum interference from transmitters
RF #1, #2, and #4
Modulation

. Figure 1- 48
Why is it called “Spread
Spectrum”?

Power Spectrum Magnitude (dB)

Power Spectrum Magnitude (dB)

Power Spectrum Magnitude (dB)
10 10
80

0
0
60

-10
-10
40
-20

-20
20
-30

-30 0
-40

-40 -20
-50

-50 -60 -40

0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
7
Frequency Frequency Frequency x 10

Nulls @ N*Rc Frf

”Chips”

RF
Filter
Modulator

“Chips”

cos(rf*t)

Scrambling Code Scrambling Code

Generator Key

Chip Clock
Fc >> Fd

. Figure 1- 49
Spread Spectrum Multiple Access
Case II: Two Transmitters at the same frequency
Amplitude

Signal 1

Frequency Scrambling
Code 1

Both signals “mixed” in the

air interface
Amplitude

Signal 2

Frequency
Scrambling
Code 2

AT THE RECEIVER... Spread Spectrum

Processing Gain
Both signals are
received together  Chip rate 
= 
 Bit rate 
Scrambling Signal 1 is reconstructed
Code 1 Signal 2 looks like noise

. Figure 1- 50
Code Correlation: Key Points

▪ TX, RX use same codes, at the same time offset

▪ Scrambling Codes: 100% correlation

▪ TX, RX use different codes

▪ Scrambling Codes: “Low” (noise-like) correlation at
any time offset
Average correlation level proportional to
1/(code length)

▪ TX, RX use same codes, but at different time offsets

▪ Scrambling Codes: “Low” (noise-like) correlation for
any offset > +1 chip

. Figure 1- 51
Scrambling Code properties

– Scrambling Codes may be generated using Linear

Feedback Shift Registers
– Scrambling Codes are repeating, defined-length blocks of
1’s and 0’s
▪ Approximately equal number of 1’s and 0’s
▪ The statistics appear randomly distributed within the
block
– Good Autocorrelation and Cross-Correlation properties
▪ Scrambling Code cross-correlation properties do not
depend on time alignment

. Figure 1- 52
Scrambling Code Generation

▪ Use a linear- Feedback Shift Register

1 2 3 N

D D D D
1010010010001110101...

clock

•  n values are 0 or 1 (determined by the specified “generator polynomial”)

• Maximal-length (m-sequence) has a repetitive cycle of ( 2N - 1 ) bits
• A code of 16 777 215 bits can be replicated using only a 24-bit “key” in Uplink. In
downlink a 18-bit “key” is used

. Figure 1- 53
Scrambling Code planning
Uplink: Scrambling Code used to distinguish each UE

Downlink: Scrambling Code used to distinguish each cell

Cell “1” transmits using SC 1

SC1 SC1

SC3 SC4

Cell “2” transmits using SC 2

SC2 SC2

SC5 SC6

. Figure 1- 54
Scrambling Code planning
example
N

SC1
W E
SC2 SC3 SC4

SC5 SC6 SC7 SC8

SC9 SC10 SC11 SC12 SC13

S
SC14 SC15 SC16 SC17 SC18

SC19 SC20 SC21 SC22 SC23 SC24

SC25 SC26 SC27 SC28 SC29

SC30 SC31 SC32 SC33 SC34

SC35 SC36 SC37 SC38

SC39 SC40 SC41

SC42

. Figure 1- 55
Scrambling Codes: Summary
Spread ▪ Scrambling Code Utilization
Spectrum – Used to distinguish the transmission
Multiple source (Cell or UE) in WCDMA systems
Access ▪ Provides good (but not 100%)
separation between multiple
Tx 1
transmissions in the same
Tx 2 geographic area, on the same
Tx 3 frequency
... – Works regardless of time-of-arrival delays
– Code Planning instead of Frequency
Frequency
Planning
Each Transmitter ▪ Frequency Reuse = 1
has a unique
Scrambling Code
▪ Limitations using Scrambling Codes
Several Transmitters share – Imperfect signal separation
the same frequency
and time
– Not good for transmitting multiple data
streams from one transmitter

. Figure 1- 56
WCDMA (Scrambling and
Channelization Codes)
Spread Code
Spectrum Division
▪ - Scrambling Codes are used:
Multiple Multiple
Access Access ▪ To distinguish between User
Equipments in uplink
User 1 ▪ To distinguish between cells
User 2
– Channelization Codes are used:
User 3
▪ To distinguish between data channels
...
coming from each User Equipment
Frequency ▪ To distinguish between data channels
Scrambling Codes from each cell

and

Channelization Codes

are simultaneously utilized

. Figure 1- 57
 WCDMA (SC + CC)
Pilot, Broadcast
SC1 + CCP + CCB
2 data channels
1 data channels
Voice (voice, control)
(control)
SC1 + CC1 + CC2 Uplink
Conversation SC1 + CC3
Packet Data

2 data channels 2 data channels

(voice, control) (14 kbps data, control)
SC3 + CC1 + CC2 SC4 + CC1 + CC2

Pilot, Broadcast
SC2 + CCP + CCB

3 data channels 4 data channels

(voice, video, control)
Videoconference
(384 kbps data, voice, video, control)
SC2 + CC1 + CC2 + CC3 SC2 + CC4 + CC5 + CC6 + CC7
with Data
Videoconference

3 data channels
(voice, video, control) 4 data channels
SC5 + CC1 + CC2 + CC3 (384 kbps data, voice, video, control)
SC6 + CC1 + CC2 + CC3 + CC4

. Figure 1- 58
 I/Q Modulation
▪ I/Q (In-phase/Quadrature) Modulation: Definition
– Two data streams are multiplied by a common carrier frequency, but
at phase offsets of 0 degrees (cosine) and 90 degrees (sine)

Data Stream #1 “ Q ”
+1

-1

I cos(wt)
90o
- Q sin(wt)
SUM

cos (wt)

Data Stream #2 “ I ”
+1

-1

. Figure 1- 59
 QPSK Modulation

▪ Graphical representation of an QPSK modulated signal

Q

( I = -1, Q = 1 ) ( I = 1, Q = 1 )

RF Carrier amplitude

RF Carrier phase angle

( I = -1, Q = -1 ) ( I = 1, Q = -1 )

1 Modulation Symbol represents 2 data bits

Modulation efficiency = 2 bits/symbol

. Figure 1- 60
 I/Q Modulation

▪ By multiplying by the sin and cosine at the receiver, the original I and
Q data streams are recovered
Data Stream #1 “ I ”
+1
LPF
-1

Q cos(wt)
- I sin(wt) 90o
DEMOD

cos(wt)

Data Stream #2 “ Q ”
+1
LPF
-1

. Figure 1- 61
Data Filtering

▪ Data Filtering: Why?

20

10

-10

RF Modulator -20

-30

-40

-50

-60
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Frequency

Baseband filtering of data stream is required

to contain RF bandwidth

. Figure 1- 62
Data Filtering
Ordinary Channel Filter: Impulse Response

1.5

0.5

-0.5

-1
0 10 20 30 40 50 60 70 80 90 100

Channel Filter

(Digital Chebyshev, 10-tap, Fc = 0.2Fs)

-5
x 10
6

4 Ringing may interfere

with subsequent bit decisions
2

-2
0 20 40 60 80 100 120 140 160 180 200

. Figure 1- 63
Raised-Cosine Data Filter
Raised Cosine Filter: Equations

T;
0    (1 −  ) / T

H ( ) = T T    
1 − sin    −  ; (1 −  ) / T    (1 +  ) / T
2  2  T  

  (1 +  ) / T
0;
 = 0.1

1  = 0.3

0.9  = 0.5
0.8

0.7

0.6

0.5

 = 0.7
0.4 WCDMA uses alpha = 0.22
 = 0.9
0.3

0.2

0.1

0
0 50 100 150 200 250 300

. Figure 1- 64
Raised-Cosine Data Filter
Raised Cosine Filter: Impulse Response

0.8
 = 0.3  = 0.01
0.6
(Wide filter) (Narrow filter)
0.4

0.2

-0.2

-0.4
t1 t2 t3 t4 t5 t6 t7 t8 t9

Notes:

1) Ringing = 0 at exact time instants where future data points are to be sampled

2) Low ‘alpha’ provides narrowest spectrum; best for reducing adjacent channel interference

3) High ‘alpha’ provides lowest ringing amplitude; best for reducing ISI

4) Theoretically, even filters with very low ‘alpha’ provide zero ringing at future sample points

5) Practically, low-alpha filters create greater ISI when there is timing jitter present

. Figure 1- 65
 Data Filtering: The RRC Filter
▪ Eye Diagram
1.4

1.2

0.8
Raised
0.6
Cosine
0.4 Filter
0.2

-0.2

-0.4
0 50 100 150

1.4

1.2

0.8

0.6
Chebyshev
0.4 Filter
0.2

-0.2

-0.4
0 50 100 150

. Figure 1- 66