8/2/2019 Ee4105 Wcdma New(1)

1/105

1

EE4105-Cellular CommunicationSystems Design

( Cellular Design Part II)

Wideband-CDMA

Lecturer : Dr. Peter Chong (Email: ehjchong@ntu.edu.sg)

School of EEE

Week # 7-13

8/2/2019 Ee4105 Wcdma New(1)

2/105

2

References

W-CDMA and cdma2000 for 3G MobileNetworks, by M. R. Karim and Mohsen

Saraf, McGraw-Hill, 2002

IS-95 CDMA and CDMA 2000

(Cellular/PCS Systems Implementation),byV K Garg, Prentice Hall PTR, 2000

8/2/2019 Ee4105 Wcdma New(1)

3/105

3

Topics To be Covered

Principles of Wideband CDMA(WCDMA): CDMA, DS CDMA, Capacity,Speech coders, Channel coders, Digital

modulations, spreading ,Walsh codes,Mutipath diversity etc.

CdmaOne and cdma2000

Network Planning and Design Beyond 3G (4G) and future technologies

8/2/2019 Ee4105 Wcdma New(1)

4/105

4

IMT-2000

IMT-2000 has defined four 3G systems Only one UWC-136 is based on TDMA

All the other three are based on direct-sequence code

division multiple access (DS-CDMA)

The three systems are, Universal Mobile

Telecommunications System (UMTS) W-CDMA

Frequency Division Duplex (FDD), UMTS W-CDMA

Time Division Duplex (TDD) and cdma2000 Therefore, it is important to understand principles of W-

CDMA technology

8/2/2019 Ee4105 Wcdma New(1)

5/105

5

Multiple Access (MA)

UMTS W-CDMA FDD uses nominal bandwidth of 5 MHz

UMTS W-CDMA TDD also uses CDMA and 5 MHzbandwidth but now the frequency band is time shared for

forward and reverse direction.

cdma2000 is multicarrier, DS-CDMA FDD system, LikecdmaOne , its first phase will use 1.25 MHz and second

phase may have three carriers to fit in 5 MHz.

The fourth system will use TDMA, where each physical

channel is divided into number of fixed, synchronized timeslots. Each user is assigned one or more time slots.

8/2/2019 Ee4105 Wcdma New(1)

6/105

6

Spread Spectrum MA

In spread spectrum multiple access scheme, all users cantransmit over entire bandwidth using a pseudorandom (PN)

code that is unique for each user.

PN codes are random codes and can be generated by using

multistage shift register

There are so many spread spectrum techniques but we will

focus on direct sequence (DS) as they are used in UMTS

W-CDMA and cdma2000. Each user is assigned a PN code

8/2/2019 Ee4105 Wcdma New(1)

7/105

7

CDMA System

Each user is assigned a specific random PN code which

can be easily generated using shift registers. The clockrate of PN sequence is known as chip rate.

They have specific two level code which is used for

synchronizing the receiver.

8/2/2019 Ee4105 Wcdma New(1)

8/105

8

CDMA Receiver

PN codes are used to identify users Suppose three users are using the different PN codes to

transmit their information.

Modulation and demodulation actions are explained on the

next two pages.

Multiplying the code sequence once spreads the signal at

the transmitter.

Multiplying the signal two time de-spreads the signal at the receiver.

At the receiver desired signal will be de-spread while the

interference signal coming from other user will be further

spread and thereby rejected.

8/2/2019 Ee4105 Wcdma New(1)

9/105

9

8/2/2019 Ee4105 Wcdma New(1)

10/105

10

Make sure that

you understandthe decoding

process clearly

8/2/2019 Ee4105 Wcdma New(1)

11/105

11

UplinkCapacity of a CDMA System

Consider a single cell withnumber of mobiles (each

with unique PN code)

Prx= Received Powerat BS

Eb= Energy per bit

Bc= Chip Rate

fdata= information bit rate

I = Interference powerat BS

No= Interferencepower per bitgainProcessing

)1......(

,

=

==

=

=

=

p

pdata

c

o

b

co

dataoo

b

datab

G

GI

P

fI

BP

N

E

soB

INfN

P

NE

fPE

8/2/2019 Ee4105 Wcdma New(1)

12/105

12

Capacity Continued

For a given bit error rate(BER) , Eb/No is fixed

Larger the processing gain

larger the allowable

interference for a givenBERor using smaller tx power.

If there are N transmitters

using same power andchip rate, then

I=(N-1)P

Nofvaluelargefor

)2.........(1

1

o

b

p

o

b

p

o

b

p

NE

G

NE

G

N

NE

G

PIN

+=

==

8/2/2019 Ee4105 Wcdma New(1)

13/105

13

Implication of capacity formula

N can be increased by increasing Gp or reducing Eb/No Previous equation is only valid in an ideal situation, forexample capacity will reduce if power control is notperfect.

All the cells are using same frequency , so in a multi-cellenvironment interference will increase by 60-85%

System is interference limited so the capacity can beincreased by reducing interference, which can beachieved in many ways. For example, a 3-sector antenna

will increase the capacity by a factor of about 2-3.

8/2/2019 Ee4105 Wcdma New(1)

14/105

14

Formula continued

Human conversation ischaracterized by talkbursts followed by silence.If the transmitter is turnedoff during silence periods,interference can bereduced resulting inincrease in capacity.

The actual capacityformula can be modifiedto include these effects

factoractivityvoice

factorchannel-CofactorcorrectioncontrolPower

)3(..........)1(

1

+

+=

o

b

p

NE

GN

8/2/2019 Ee4105 Wcdma New(1)

15/105

15

Typical values

Power control correction factor, , range between 0.5-1.0 Voice activity factor, range between 0.4 - 0.6

Effect of co-channel interference from other cells in the

system , , range between 0.5-0.9. A typical value for 3-

sector cell is 0.85.

Example: =1 (perfect power control), =0.4, =.85 for a

three sector cell, data rate = 9.6 kbps, and chip rate=1.2288

Mc/s. The required Eb/No= 7 dB.

8/2/2019 Ee4105 Wcdma New(1)

16/105

16

Example

Eb / No = 100.7

= 5.01 Gp = 1.228 X 10

6 /9600=128

N=1+(128/5.01)(1/1.85)(1/0.4)=35

This is known as the sectorized pole capacity

Notice that the capacity can be increased by simplyreducing Eb / Nobut that will result in increase in BER forall users.

One way to avoid this is to minimize Eb

/ No

by usingefficient modulation scheme.

For example BFSK requires Eb / No = 12.6 dB at BER= 10-

5 whereas BPSK or QPSK require only 9.6 dB

8/2/2019 Ee4105 Wcdma New(1)

17/105

17

Error performance

BER increases as the SIRsignal to interference ratio is

minimized, it is necessary to

use an error correcting code.

The convolutional code is

generally used in CDMA and

W-CDMA systems to achieve

coding gain of 4-6 dB.

Thus , the capacity of CDMA

system can be increased byusing channel coding

8/2/2019 Ee4105 Wcdma New(1)

18/105

18

3G Transmitter

A simplified diagram showing transmit functions of amulticarrier cdma2000 base transceiver station is shown onthe next page.

The incoming data stream is encoded using CRC code and

convolutional code (constraint length 9 and rate 1/3) . Depending on the data rate , the output of the code may

have to be repeated a few times.

The output of the symbol repetition block is applied to an

interleaver that spreads out burst errors A long PN code that is unique for each user scrambles the

output of interleaver.

8/2/2019 Ee4105 Wcdma New(1)

19/105

19

8/2/2019 Ee4105 Wcdma New(1)

20/105

20

3G Transceiver

Scrambled sequence is applied to a demultiplexer where itis broken into N subsequences, N is the number of CDMAcarriers.

Each of these subsequences is transmitted over a separate

CDMA carrier as shown in the diagram. The chip rate used is N X 1.2288 Mc/s, the value of N may

be 1,3,6,9,0r 12. However, standard currently specify N=1and 3 only.

Subsequences A1, A2, are multiplex with power controlbits, converted in to parallel form and then split into I andQ form. Each bit is mapped into BPSK symbol.

8/2/2019 Ee4105 Wcdma New(1)

21/105

21

3G Transceiver continued

The symbols of the I and Q

branches are multiplied by a

gain factor and spread by a

Walsh code , say WA which

different for each carrier.

Walsh codes , are sequence of+1,-1 s and are orthogonal

codes

The I and Q symbols after

Walsh spreading are added inQuadrature to form complex

symbols.

The complex symbols are again

spread by complex PN code

SI+j SQ where SI and SQ are

the cell specific I-channel and

Q-channel pilot PN sequence,

respectively. The I and Q components of the

output from complex spreading

are passed through a pulse

shaping filter and modulate thedesired CDMA carrier as

shown in the block diagram.

8/2/2019 Ee4105 Wcdma New(1)

22/105

22

Speech Encoding

Various speech coders used in different mobile communications are

listed in the table on the next page.

UMTS uses Adaptive Multirate (AMR) coding based on the principles

of Algebraic Code Excited Linear Prediction (ACELP).

ACELP belong to the Vocoder class of encoders, unlike a waveform

quantizer, model the vocal tract as a time varying digital filter such that

when it is excited with an appropriate input, the output is a desired

speech signal.

The filter coefficients are determined by analyzing the speech input.

The topic of speech coders is quite involved, we just need to be aware

of various coders used in mobile communications.

8/2/2019 Ee4105 Wcdma New(1)

23/105

23

Speech Encoders

8/2/2019 Ee4105 Wcdma New(1)

24/105

24

Channel Coding

In W-CDMA, user data and voice are encoded byconvolutional codes

In UMTS, both convolutional codes and Turbo codes are

suggested.

The convolutional codes used are of rate and 1/3 and

constraint length of 9.

Two structures are specified in octal form as 561(101 110

001) and 753(111 101 011) 557,663 and 711 , encoder diagrams are shown on the next

page.

8/2/2019 Ee4105 Wcdma New(1)

25/105

25

Convolutional Codes

8/2/2019 Ee4105 Wcdma New(1)

26/105

26

Turbo Code

It consists of interleaver and two identical encoders. Above is

a rate 1/3 encoder.

8/2/2019 Ee4105 Wcdma New(1)

27/105

27

Digital Modulation

The simplest modulation scheme is BPSK, where 0represent 0 degree phase and 1 is represented by 180

degree phase

In digital cellular systems QPSK is used where , each

phase is represented by two bits , namely, (00) denote 0degree, (01) denote 90 degree , (11) denote 180 degree and

(10) denote 270 degree

With QPSK null to null bandwidth is Rb , example for 10

Kbps it will be 10 KHz. Using raised cosine filter

bandwidth can be controlled to say .75Rb.

8/2/2019 Ee4105 Wcdma New(1)

28/105

28

Spreading

In UMTS and cdma2000 , data is spread twice in succession, first with

channelization codes and later with scrambling codes

Channelization codes are orthogonal Walsh codes which are inherently

more tolerant of interference caused by multiple users

Scrambling codes are not necessarily orthogonal and are constructed

using PN codes

Channeliztion codes in UMTS W-CDMA and cdma2000 are variable

length Walsh codes, also known as Orthogonal variable spreading

factor (OVSF) codes.

Variable factor in UMTS may vary from 4-256 on uplink and 4-512 on

downlink channels. In cdma2000 , it varies from 4-128.

In IS-95 , 64 fixed length Walsh code is used on forward link

8/2/2019 Ee4105 Wcdma New(1)

29/105

29

Walsh Codes

Walsh code of 2n

can beeasily generated

recursively by using

Hadamard Matrix

Check all the rows areorthogonal to each other

First two rows are,1010

and 1010

Check whether they are

orthogonal?

[ ]

==

==

==

1001

1100

1010

1111

10

11

1

42

22

12

2

1

0

HH

HH

HH

8/2/2019 Ee4105 Wcdma New(1)

30/105

30

Scrambling Codes

PN codes are the basic building block for these codes. These codes are

generated by a shift register where some selected outputs are modulo 2

added and fed back to the input.

The underlying theory is well developed and will not be covered.

But few things can be pointed out. The out put sequence is periodic but

its bit pattern is random, thus it is termed as pseudo random code or

PN code in short.

It satisfies randomness properties

It has two level auto correlation function which can be used to

synchronize the code.

The code in the receiver can be shifted till the autocorrelation function

changes its state to indicate synchronization.

8/2/2019 Ee4105 Wcdma New(1)

31/105

31

Receiver Structure

A QPSK receiver structure is shown on the next page.

In coherent reception, carrier is first generated at the receiver and is

known as carrier recovery

The output of the demodulator is low pass filtered and applied to the

input of a matched filter.

Matched filter (integrate and dump filter) is able to maximize SNR at

the out put

QPSK can be considered as two BPSK signal operating at two

orthogonal carriers.

The output of the matched filter is despread by multiplying it with I-

channel and Q- channel scrambling code

8/2/2019 Ee4105 Wcdma New(1)

32/105

32

QPSK Receiver

8/2/2019 Ee4105 Wcdma New(1)

33/105

33

Mobile Radio ChannelsA land mobile radio channel is characterized by out-of-sight

communication to/from a moving terminal. Wave propagation in the multi-

path channel depends on the actual environment, including factors such as

antenna height,profile of the buildings,roads,and terrain. Therefore, we

must describe mobile radio channels in a statistical way.The received signal

power Pr is expressed as

Pr= Pt G LC

Where Pt is the transmitted power, G denotes the antenna gain, and LCrepresents the propagation loss in the channel.

Wave propagation in a mobile radio channel is characterized by threeaspects: path loss, shadowing , and fast fading.

LC = LP LS LF

8/2/2019 Ee4105 Wcdma New(1)

34/105

34

Multi-Path Diversity in CDMA

Wide bandwidth signals (CDMA) offer some advantages which are not

available in Narrowband system.

Due to delay spread in time, coherence bandwidth is to characterize the channel in frequency.

If the signal BW is smaller than the coherence BW of the channel, we have flat

fading , that is different frequency components of the signal have identical

statistics If the bandwidth of the signal is large as in W-CDMA, compared to coherence

BW of the channel, now the different frequency components are statistically

independent , we call this as frequency selective fading

Figure on the next page shows the effect of channel BW on fading

Fade depth is defined as the amount by which signal falls below its average

value with probability of 0.1

Multi-path components delayed by more than chip period can be considered as

multiple copies, which can be combined in a diversity receiver

8/2/2019 Ee4105 Wcdma New(1)

35/105

35

If the channel BW is 30

KHz as in the narrow

band system, fade depth

will be -10 dB withprobability .1

IS-95 , BW is 1.25 MHz,

fade depth will be -8.75

dB with probability .1

For W-CDMA , BW is 5

MHz, fade depth will be

-5.75 dB with

probability 0.1

8/2/2019 Ee4105 Wcdma New(1)

36/105

36

Rake Receiver

Studies of power delay profiles of urban and dense areas around 900

MHz indicate that most of the energy of the received signal is due to

the reflected rays with delays in excess of 0.75 s.

In W-CDMA chip period is quite small (0.2604 for a chip rate of 3.84

Mc/s) compared to the delay spread, multipath components with delays

of more than one chip period may have significant energy. Thus, the various multipath components can be used as different

branches (Fingers) of a diversity receiver.

This is the basis of Rake receiver

Functional block diagram of a rake receiver for W-CDMA is shown onthe next page

8/2/2019 Ee4105 Wcdma New(1)

37/105

37

Various combing schemes like,Maximum ratio combing(MRC) , Equal gain combining (EG) etc. can be used

8/2/2019 Ee4105 Wcdma New(1)

38/105

38

Exploitation of Multipath

The maximum amount of multipath delay that can be exploited ina

Rake receiver is usually limited and is determined by the power delayprofile (urban area typical value 0.25-2.5 s).

For UMTS W-CDMA , where chip rate is 3.84 Mc/s , the delay isabout 1-10 chips.

IS-95, chip rate is 1.2288 Mc/s, delay must be one chip long to providemultipath diversity, so the difference in path lengths must be 244meters.

In W-CDMA with chip rate of 3.84 Mc/s , the path difference must beabout 78 meters

The multipath diversity employed in a rake receiver leads toimprovement in performance

MRC has the best performance

8/2/2019 Ee4105 Wcdma New(1)

39/105

39

Multiuser Detector

Consider the uplink transmission in UMTS, each channel is first spread

using channelization code and then scrambled with a user specific PNcode.

Because channelization codes are orthogonal and thus more resistant tomultiuser interference.

The scrambling codes on the other hand are generally non-orthogonal It is not a problem in Synchronous system such as IS-95

In contrast, because W-CDMA is an asynchronous system, thesedelays are random ( as shown on the next page), may be comparable tothe bit period.

The cross-correlation between the received signals from multiple usersis no longer negligible.

Multiuser detection attempts to overcome this problem.

8/2/2019 Ee4105 Wcdma New(1)

40/105

40

8/2/2019 Ee4105 Wcdma New(1)

41/105

41

Smart Antennas

Extending the range or coverage area in a desired direction

with beamforming

Increasing the system capacity in areas with dense traffic

(hot spots)

Creating nulls to reduce interference

Tracking individual mobile stations using separate, narrow

beam in their direction

Reducing the multipath Possible benefits of using smart antennas in 3G systems

have been studied

8/2/2019 Ee4105 Wcdma New(1)

42/105

42

cdmaOne and cdma2000

We will now intrroduce cdma2000/IS-95

3G standards is to allow graceful evolution of current 2G

wireless networks

cdma2000 is an evolution of the present North American

CDMA system called cdmaOne based on IS-95 standards

The frequency allocation for cellular and PCS is shown on

the next page

50 MHz and 120 MHz for cellular and PCS respectively First let us give brief description of cdmaOne

8/2/2019 Ee4105 Wcdma New(1)

43/105

43

Spectrum Allocation

30KHz spacing for cellular system and 50 KHz for PCS. Thus there are 1200

FDD channels in a PCS. For satisfactory operation, CDMA carriers areseparated by at least 25 channels or 1.25 MHz. Thus the nominal BW of a

CDMA system is 1.25 MHz.

8/2/2019 Ee4105 Wcdma New(1)

44/105

44

Physical Channels

In the uplink, there are two physical channels-the accesschannel and traffic channel.

Access channel are used for signaling messages like callorigination request ,a page response, an order message etc.

A system may have one or more access channels, eachassociated with a paging channel (downlink).

A traffic channel carries user traffic , such as speech ordata, and may be used during a call to send signaling

messages such as a handoff completion or a pilot strengthmeasurement message , report power measurements to BS,and so on.

8/2/2019 Ee4105 Wcdma New(1)

45/105

45

Physical Channels Continued

In the down link, there are four classes of physical channels- a pilot

channel, a sync channel , up to 7 paging channels, and up to 55 trafficchannels per sector.

For an active forward CDMA channel, there is a pilot channel thatcontinuously sends a carrier modulated by an all zero Walsh code sothat mobile can synchronize to a base station. This signal can also beused as reference in coherent demodulation and timing recovery at amobile station.

It can also be used to measure signal power and can also be used forhandoff

The sync channel transmits information that enables mobile stationswithin the coverage area to acquire frame synchronization afterachieving pilot sync

8/2/2019 Ee4105 Wcdma New(1)

46/105

46

Physical Channels continued

A paging channel carries system overhead information,

such as system parameters, access parameters, a CDMA

channel list , a neighbor list and so on. The information

rate on the sync channel is 1200 b/s.

The purpose of a traffic channel is to send the user data aswell as signaling messages to a mobile station during a

call. The information rate may be 8.6 kb/s, 4.0kb/s, 2.0

kb/s and 0.8 kb/s

We will now discuss the transmit functions of cdmaOne

system.

8/2/2019 Ee4105 Wcdma New(1)

47/105

47

Reverse Channel-TransmitterDiagram

8/2/2019 Ee4105 Wcdma New(1)

48/105

48

Transmit Functions The data streams which may originate at different rates are arranged in 20

ms frames.

The output of the interleaver is passed through an orthogonal modulator,where 6 bits are converted to 64 bit modulation using Walsh matrix.

(28.8*64/6=307.2kc/s)

The output modulator feeds to the data burst randomizer that allows only

one copy to be transmitted 20 ms frame is divided into 16 blocks or power control groups as they are

called.

The transmitter is gated on during only some of these groups depending on

the date rate.

8/2/2019 Ee4105 Wcdma New(1)

49/105

49

Data Randomizer

For example if 9.6 kbps is used, the transmitter must be gated on all

the time. If 4.8 kbps is used the transmitter should be gated only half

the time. Gating is randomly selected as shown in the above diagram(b), shaded groups show the gating for 4.8 kbps.

8/2/2019 Ee4105 Wcdma New(1)

50/105

50

After data Randomization

The output of the randomization is spread by a long code which derived

as follows

The 42 bit shift register sequence is ANDed with a long code mask that

is constructed with the permuted electronic serial number (ESN) of the

mobile number. Thus the long code mask and hence the output of the

long code generator are unique for each user. Randomizer output 307.2 kbps is expanded to 1.2288 Mc/s, hence each

bit is spread by a factor of 4.

The resulting output is divided into two sequences , I and Q sequences,

which are spread by zero-offset, I and Q pilot PN sequences of period215-1 (chips).

Offset-QPSK (OQPSK ) more suitable for non-linear amplifiers

8/2/2019 Ee4105 Wcdma New(1)

51/105

51

Continued

The transmit functions of reverse access channel are slightly different

There is only one data rate, namely 4.4 kbps, codes are repeated just

once and no burst randomizer is used.

Each spread code is spread by a long code, which is derived as the

same way as for traffic channel except for a different long code mask

that includes the access channel number, the paging channel number,the base station identification number and so on.

8/2/2019 Ee4105 Wcdma New(1)

52/105

52

Forward Channel Functions

cdmaOne uses a system wide reference time scale that is based upon GPS

synchronized with a universal coordinated time. Each base station derives itstime base from this reference time scale. The functional diagram of the basestation transmitter is shown on the next page.

The Pilot channel carries an all zero pattern and is spread by all zero WalshFunction 0 (W0).

The sync channel is used to transmit a synchronizing sequence at 1.2 kbps.The data is encoded with a rate constraint length 9 convolutional code. Theoutput is repeated once to obtain 4.8 kbps. It is passed through a blockinterleaver and spread by W32

The paging channel data at 9.6 kbps or 4.8kbps, is encoded, symbol repeated,interleaved, scrambled, and spread by a Walsh function (W1-7)

Scrambling is achieved by 42 bit mask that includes 3 bit paging channelnumber.

8/2/2019 Ee4105 Wcdma New(1)

53/105

53

IS-95

ForwardChannels

8/2/2019 Ee4105 Wcdma New(1)

54/105

54

Traffic Channels

Data rates on forward traffic channel may be 8.6,4.0,2.0,or 0.8 kbps.

If it is 8.6 kbps, 12 bits of frame quality CRC (Cyclic redundancycheck) are added . For 4 kbps , 8 bit CRC are added, for other rates noCRC are added. To reset the encoder a sequence of zeros is appendedto each frame.

The resulting output is encoded, interleaved, repeated on a symbol bysymbol basis, and scrambled in the same way as a paging channel.

42 bit mask is constructed with the 32 bit ESN of the particular user.

Once every 1.25 ms, a power control bit with a duration of two codesymbols is transmitted over a forward traffic channel to indicate to the

mobile whether it should increase or decrease its power level. If it is 0,the power should be increased otherwise it is to be decreased.

Power control bit is inserted by puncturing

8/2/2019 Ee4105 Wcdma New(1)

55/105

55

Traffic Channels continued

The I- and Q-channels are spread by two pilot PN sequences of period

215-1 (chips) with an offset with respect to a reference PN code.

This offset is unique for each base station, and is expressed in terms of

chip rate and is given by 64n chips where 0 n 511. Thus cdmaOne

is a synchronous system where each base station is uniquely identified

by offset index n . Each forward channel type- pilot, paging, sync, and traffic channels-is

separated at a mobile station by means of the Walsh codes.

All the codes are same, however, each paging or forward traffic

channel is associated with a unique code mask. Thus, mobile stationscan separate traffic channels by de-spreading the received signal with a

Walsh code (W8-W31, W33-W63) and the user specific long code

8/2/2019 Ee4105 Wcdma New(1)

56/105

56

Power Control

Near far problem as shown in the

figure Stronger signal may swamp out the

weaker signal as they are operating

at the same carrier frequency.

To overcome this problem, BS

measures the signal from mobile

station , and if it is above a

threshold , it sends command to

that mobile to reduce its power

level.

Similarly, if the signal is below a

threshold, the mobile may be asked

to increase mobile power

8/2/2019 Ee4105 Wcdma New(1)

57/105

57

Power Control Contd.

The power control commands are sent at

800 b/s by puncturing code symbols on atraffic channel once every 1.25 ms( 16

times per 20 ms frame)

As shown, power control commands can be

used to combat fadingAnother side benefit of the power control is

that mobile station can operate at an

optimum power level thereby achieving

longer battery life

8/2/2019 Ee4105 Wcdma New(1)

58/105

58

Uplink and Downlink Power Control

Uplink power is needed to achieve satisfactory SIR (signal to

interference ratio) from mobile.

The closed loop power control is based on the direct measurements of

the desired signal

Open loop power control refers to the indirect measurement like,

mobile may change the power based on the received signal power.Both open and closed loop power controls are used on reverse link

Power control is also used on a downlink channel so that each mobile

receives satisfactory SIR. The algorithms are usually closed loop

where each mobile measures the received signal on the forwardchannel and based on the measurements, instruct the BS to adjust the

power

8/2/2019 Ee4105 Wcdma New(1)

59/105

59

Handoff in IS-95

The process of switching from one BS to another is called a handoff.

In the AMPS or TDMA system such as IS-136 or GSM, each cell in a clusteris assigned different frequency. If the mobile moves into an adjacent cell, its

Tx and Rx must operate at different frequencies. This is called hard handoff

(break before make).

Things are different in CDMA system.

If two CDMA systems are using different frequencies, clearly there will be

hard handoff. If, the two adjacent cells are operating at same frequency, it uses

soft handoff (make before break) by combining forward traffic channels.

An intracell or intersector handoff is called a softer handoff.

IS-95 supports three types of handoffs: soft and softer handoff( at samefrequency), Hard handoff (two different CDMA carrier frequency), and

CDMA to analog cellular system handoff

8/2/2019 Ee4105 Wcdma New(1)

60/105

60

Soft Handoff Procedure

Mobile maintains four sets of pilots

Active set- all the pilots that are currently being used

Candidate set- all the pilots that have been determined by

the mobile as sufficiently strong that their forward traffic

channels can be used by this mobile Neighbor set-all the pilots in the serving area not included

in the last two sets, and may be used as probable

candidates for a handoff.

Remaining set- Pilots outside of the previous three sets

8/2/2019 Ee4105 Wcdma New(1)

61/105

61

Procedure Continued

Various steps of a

typical soft

handoff is shown

in the figure

8/2/2019 Ee4105 Wcdma New(1)

62/105

62

Cdma2000

Traffic types-cdma2000 like other 3G technologies, is expected to

support following type of traffic. The data rates may vary from 9.6kbps to 2 Mbps.

Traditional Voice and voice over IP (VoIP)

Data services- packet data, Circuit broadband data and SMS

3G systems are intended for indoor and outdoor environments,pedestrian or vehicle applications, and fixed environments such as

wireless local loops. Cells sizes may range from a few tens of meters

(less than 50 m for picocells) to few tens of kilometers (in excess of 35

km for large cells) Bandwidth-A cdma2000 system may operate at different bandwidths

with one or more carriers

8/2/2019 Ee4105 Wcdma New(1)

63/105

63

Bandwidth requirements incdma2000

A cdma2000 system may

operate at different

bandwidths with one or

more carriers

When three carriers areused BW is 5 MHz with

a chip rate (3 X 1.2288

Mc/s = 3.6864 Mc/s)

Wide bandwidth canprovide more resolvable

paths

8/2/2019 Ee4105 Wcdma New(1)

64/105

64

Forward Physical Channels

As in IS-95 , pilot channel continuously transmits a carrier modulated

for initial cell synchronization and coherent demodulation. Thereceiver strength can also be used for handoffs.

A common auxiliary pilot channel has been added to cdma2000 so thatadaptive antennas can be used. It is necessary that the pilot and datasignals travel along the same path for an accurate channel estimate.

A dedicated auxiliary pilot channel is dedicated to a given mobile (or agroup of mobile stations) for the purpose of beam steering using anadaptive array.

A sync channel operates at 1200 bps, transmitting synchronization

messages, so the mobile in the coverage area can acquire framesynchronization after cell acquisition.

Paging channel uses two data rates namely, 9.6 and 4.8kbps.

8/2/2019 Ee4105 Wcdma New(1)

65/105

65

Continued

The fundamental channel is used

for lower data rates: 9.6 kbps andits subrates , grouped as rate set 1,

and 14.4 kb/s and its subrates

grouped as rate set 2. This is

supported in single and multicarrier

cdma2000. Both 20 ms and 5 msframe are permissible.

Supplementary channel 1 and 2 are

designed for higher data rates.

Rates supported are shown in the

table. Frames are usually 20 ms.

8/2/2019 Ee4105 Wcdma New(1)

66/105

66

Reverse Physical Channels

The reverse pilot channel is similar in concept to the forward pilot channel.

Used in conjunction with the reverse dedicated channels, it enables the basestation to acquire initial time synchronization and recover a phase coherent

carrier for rake receiver. It also includes a power control sub channel ( 1 bit is

1.25 ms). The base station can use this bit to adjust power level .

Access channel (9.6 kbps)- Multiple users access this channel using a

mechanism that is very similar to slotted Aloha channel. There may be morethan one access channel each identified by a unique orthogonal code

The dedicated control channel (9.6 or 14.4 kbps)

The fundamental channel- 9.6 kbps and its subrates(4.8,2.4 and 1.2 kbps)

Supplementary channel 1 and 2 are similar to forward link, provide higherrates(1)9.6,19.2,38.4,76.8 and 153.6 kbps and (2) 14.4, 28.8,57.6,115.2, and

230.4 kbps.

8/2/2019 Ee4105 Wcdma New(1)

67/105

67

Forward Transmit Functions

Notice the similarity with IS-95 (see the figure on the next page).

Some of the differences areas follows. Cdma2000 has two traffic channel type- fundamental and secondary, a

number of data rates are supported.

I- and Q- channel symbols are multiplied by gain factors to provide

additional power control. As in IS-95, cells are separated by different pilot PN sequence offsets,

however, now complex spreading is used.

Complex spreading is able to achieve improved power efficiency

8/2/2019 Ee4105 Wcdma New(1)

68/105

68

8/2/2019 Ee4105 Wcdma New(1)

69/105

69

Reverse Transmit Functions

Refer to the diagram on the next page

The fundamental channel is processed in the usual way The output of the interleaver is spread with a Walsh code, mapped into

modulation symbols, and multiplied by gain factors , resulting in a signal labeledAfund.

The supplementary channels 1 and 2 , and control channels are processed in the

same way. The processed outputs are labeled as Asup1, Asup2, Acont, andApilot .

The fundamental channel and supplementary channel 1 are summed togethergiving an output Q. Similarly, the remaining channels are summed separately,giving I as the output.

The I and Q sequences are spread by a complex code of the type SI+ SJ,where SIand SJare user specific obtained from 42- bit long mask code, I- and Q- channelpilot PN sequences, and a Walsh code.

8/2/2019 Ee4105 Wcdma New(1)

70/105

70

cdma2000 reverse channel transmit functions

8/2/2019 Ee4105 Wcdma New(1)

71/105

71

Features of cdma2000

Wider bandwidth and chip rate-allows for much higher data rates(144 kbps-2 Mbps). Many moreresolvable paths

Multicarrier System- Each carrieris orthogonally spread, W-CDMAcan be overlaid on an existing IS-95

Spreading Codes- Similar to IS-95. On the reverse link, Incdma2000, physical channels areseparated by Walsh codes andmobile stations by long codes

Quality of Service- Multimediaservices at variable rates with userspecified QoS

Variable length Walsh Codes-traffic channels in cdma2000 isrequired to support many rates,variable length Walsh codes areneeded

Complex spreading- is needed tomake the signal more suitable fornon-linear amplifiers

Additional pilot channels- areused in cdma2000

New Traffic channels-Fundamental and Supplementary

Packet mode data Services-

Traffic channels, control channelscan be used slotted Aloha scheme

8/2/2019 Ee4105 Wcdma New(1)

72/105

72

Network Planning and Design

Objective is to provide wireless telephony services in a serving area in

the most cost effective manner In an existing system, the objective is to expand and augment its

facilities so as to add new features and capabilities or increase itscapacity in case the system has reached its coverage limit

The design involves determining the number of base stations and their

location that will provide necessary coverage in the serving area, andmeet the grade of service and satisfy growth requirements.

The design also requires the capacity and type of connecting links asthe base stations have to be connected to the mobile switching office

Operators need to generate a set of requirements concerning thedesired system (Analog, GSM, and CDMA and so on), the expectedtraffic and desired service quality.

8/2/2019 Ee4105 Wcdma New(1)

73/105

73

Quality of Service Indicators

Received signal to interference ratio (S/I) and bit error rates are generally usedas the quality of service indicators.

Based on the above requirements, an appropriate propagation model is used tocalculate link budget that gives us the maximum allowable path loss for giventransmitter power so that the sufficient S/I can be maintained to ensure desiredquality.

Maps and the terrain of the serving area are inspected, and assuming

approximate locations of base stations, the signal distribution over that area isthen calculated.

Goal is to provide coverage on the entire serving area with a minimum basestations consistent with the projected traffic growth.

Currently, software tools are available to that takes into account designrequirements and terrain features to predict signal distribution over the serving

area. It may be useful to verify the design by some field tests as actual antenna

heights and location may be different from the simulation.

8/2/2019 Ee4105 Wcdma New(1)

74/105

74

System Requirements

The coverage area- This involves areas to be served, e.g. countries

comprised by the area Terrain and clutter e.g., the average height and density of buildings,

streets, hills, forests, large water bodies if any , highways, population

distribution and so on.

System-related requirements- Technology type, like whether itshould be CDMA,GSM,W-CDMA, TDMA , cellular etc.

The allocated bandwidth- the number of available channels

The type of antennas to be used for link budget

Maximum cell size The cost objective

8/2/2019 Ee4105 Wcdma New(1)

75/105

75

The Traffic

This should include the

following The number of mobile stations

to be served

The amount of traffic-the

offered load per mobile and theholding time

The geographical distribution of

the traffic if it is not uniform

over the whole area (traffic is

rarely uniform over whole

serving area)

Specification of the traffic types

(such as constant bit rate,variable bit rate, delay-

intolerant data, elastic data) and

traffic descriptors ( e,g.,

maximum tolerable delay

during busy hour)

The probability of the calls

being blocked or the grade of

service

Ratio of the total daily traffic to

the busy hour traffic

8/2/2019 Ee4105 Wcdma New(1)

76/105

76

The Traffic Continued

For satisfactory service, the system should be designed so that the

mobiles receive a sufficiently strong signal inside buildings or vehicles,outside buildings and on highways.

The system should be designed to optimize following parameters.

The signal distribution as received by mobiles and base stations

The S/I ratio at base stations The S/I ratio at mobiles or any combinations of these parameters

However, the usual practice is to design the system such that both

forward and reverse links have a balanced signal distribution

The forward link loss and reverse link loss must be adjusted so as tohave almost or in the same range

8/2/2019 Ee4105 Wcdma New(1)

77/105

77

Network Design

The first step is that the service provider has the appropriate license for

the spectrum for the amount of traffic and the call-blockingprobability. The system must be designed to carry peak traffic, the

traffic during the busy-hour

The traffic is determined by call arrival rate and the holding time of

each call The unit of the traffic is theErlang, which is defined as the traffic that

a circuit can carry if it utilized 100% of the time during a busy hour.

The holding time varies depending on the application, telephone

conversations during a busy hour lies in the range of 60-80 seconds.

The probability that a call is blocked depends on the traffic channels

available (circuits) and the total amount of the traffic coming into the

network (the offered load), and is given by the well known Erlang B

formulation.

8/2/2019 Ee4105 Wcdma New(1)

78/105

78

Network Design Continued

Call blocking probabilities for various values of the offered load and

circuits are available as tables ( appendix) and graphs, where it isassumed that calls arrive at the system randomly with Poissondistribution and that blocked calls are cleared.

Let us now show with the help of an example how to determine therequired bandwidth

Example: Suppose we want to design a cellular system for 50,000subscriber. On the average, each subscriber makes about two callsduring a busy hour with average holding time of a call to be twominutes. Let us assume that the serving area will have 14 , 3 sectorcells and that the traffic is uniformly distributed over the entire

serving area. If the call-blocking probability is to be 1%, calculatethe bandwidth required to provide the service.

8/2/2019 Ee4105 Wcdma New(1)

79/105

79

BW for an analog system

Consider an analog system

The total traffic during the busy hour =Number of subscribers X Number ofcalls/hour X holding time in hours = 50,000 X 2 X (2/60) = 3,333 Erlangs

The traffic per sector of a cell = 3,333 / (No. cells X No. sectors) = 3,333/ (14

X 3) = 79.36 Erlangs

The number of channels or circuits per sector required to support this traffic

for a call blocking probability of 1 % ( from the appendix) = 95. Hence 95 X 3

= 285 channels are needed per cell.

For an analog system, each channel uses 30 kHz, so the total bandwidth per

cell amounts to ( 285 X 30 kHz) = 8.55 MHz.

Because 7 cell clusters are used , the total spectrum required = 8.55 MHz X7 = 59.85 MHz , which well beyond the spectrum allocated by the FCC to

cellular systems to allow 40 MH.

8/2/2019 Ee4105 Wcdma New(1)

80/105

80

Effect of call-blocking Probability

Let us increase the call-blocking probability to 5% , and calculate the spectrum

requirements. For 79.36 Erlangs , and 5% blocking the number of channels from the table are

between 84-85 , we choose 84.

Channels per cell = 84 X 3= 252

BW for analog system =252 X 30 kHz = 7.56 MHz

So for a seven cell cluster = 7.56 X 7 = 52.92 MHz

Hence it can be seen that BW is not substantially reduced as compared to the 1

% case (59.85 MHz).

To be able to meet the spectrum requirements many more cells must be added

Let us calculate the spectrum requirement if 70 cells are used instead of 14 ,and the blocking of 1%.

8/2/2019 Ee4105 Wcdma New(1)

81/105

81

Increase the number of cells

Let us calculate the spectrum requirement for 70 cells.

The traffic per sector of a cell = 3,333 / (No. cells X No. sectors) =3,333/ (70 X 3) = 15.87 Erlangs

Number of circuits for 1% blocking = 25

Circuits per cell = 25X 3 = 75

Spectrum per cell =75 X 30 kHz = 2.25 MHz

Total spectrum for 7 cells ( if 7 cell cluster is used) = 2.25X7= 15.75

MHz

The spectrum requirement may be acceptable but the system becomes

very complex as 70 cells are needed.

8/2/2019 Ee4105 Wcdma New(1)

82/105

82

CDMA System Let us now consider CDMA system. As in the analog case, the number of

channels per sector for a 1 % blocking = 95. The number of users per sector

of a CDMA cell is given by equation (3) discussed earlier

bitrateon/informatibandwidth,B/RgainocessingPr

factoractivityvoicefactor,channel-Colimit)upper(gives

caseidealanfor1factorcorrectioncontrolPower

/,)1(

1

bp ===

==

==

=++=

G

densitynoiseenergyBitNE

NE

GN

o

b

o

b

p

Chip rate

8/2/2019 Ee4105 Wcdma New(1)

83/105

83

Example- CDMA

Equation (3) gives an upper limit on the number of users because it

assumes ideal power control of the mobiles in this cell is perfect andthat the interference to mobiles in other sectors is zero.

Second, the mobiles in the other cells are not controlled by this cell, so

the interference they cause varies randomly. However, because there

are so many mobiles, it is possible to consider an average value, which

is represented by (let us assume this value to be 0.85)

And = 0.4= Voice activity

A satisfactory value of Eb/No= 7 dB= 10^0.7= 5.1

Gp = (95-1) X (5.1) X (1.85) X (0.4) = 348.5 If the bit rate is 14.4 kb/s, then B = 5.0184 Mcps ~7-8 MHz

Which well within the allowable realm of the allowable spectrum

8/2/2019 Ee4105 Wcdma New(1)

84/105

84

DownlinkLink Budget Calculation The link budget calculation is fundamental to the design of cellular systems.

Let us illustrate this with the help of an example.

Example: Determine the transmitter output power PBTS of a base stationtransceiver that can provide 30 dB SNR at the baseband in an urban coveragearea at the Rx. The system is assumed to be analog FM with following parameters.

Carrier frequency 900 MHz

BS height 50 m

User equipment (UE or mobile) height 1.5 m Distance, d, between UE and BS 2.0 km

BS antenna gain (GBTS) 9 dB

UE antenna gain (GUE) 3 dB

UE receiver noise figure , NFR 5 dB

RF (radio frequency) bandwidth 30 kHz Bandwidth of Speech 3 kHz

8/2/2019 Ee4105 Wcdma New(1)

85/105

85

Using Hata-Okumura model, the path loss PL in a typical urban area at900 MHz with respect to a reference point at a distance of 1 km from

the transmitter antenna is given by

PL = 123.33 + 33.77 log r dB, r 1 km

So for our example,

PL = 123.33 + 33.77 log (2.0) = 133.5 dB

50 m

1.5 m

Referring to the figure on the last page, the input (carrier) power to the

8/2/2019 Ee4105 Wcdma New(1)

86/105

86

mobile receiver isPr = PBTS + GBTS PL + GUE= PBTS +9.0- 133.5 +3.0 = PBTS - 121.5

The receiver noise floor is given by 10 log (KTB), where K is the

Boltzman constant, T is the absolute temperature and is taken to be 290

degrees Kelvin, B is the RF bandwidth (BW).

Here, KT = 1.38 X 10-20 X 290 = 4 X 10-18 mW/Hz

So the noise floor is 10 log (KT) = -174 dBm/Hz

Because the noise figure is 5 dB, the receiver noise density is given by

-174+5.0 = -169 dBm/Hz

So,

Noise power = -169 + 10 log (B)

CNR = PBTS - 121.5 (-169 + 10 log (B) ) = PBTS + 47.5 10 log (30000)

dBm

dB

l l d d

8/2/2019 Ee4105 Wcdma New(1)

87/105

87

Example concluded

The baseband SNR of an analog FM system depends on the average

value of the carrier-to-noise ratio (CNR) at the input of the receiverand the RF channel bandwidth. The average CNR for any desired SNR

at the baseband can be calculated by averaging the fading signal at the

receiver input over the fading distribution. It can be shown that to

achieve 30 dB SNR at baseband with an RF bandwidth of 30 kHz , a

CNR of 33 dB is needed at 100 km/h, so

33= PBTS + 47.5 10 log (30000)

PBTS = 33- 47.5 + 10 log (30000)= 30.27 dBm = 1.06 W

In these calculations no diversity has been assumed

CDMA E l

8/2/2019 Ee4105 Wcdma New(1)

88/105

88

CDMA Example

It has shown that the local mean of the signal level varies randomly

with a log-normal distribution with a standard deviation of 8- 12 dB. In the CDMA system, soft handoffs provide 2-3 dB gain due to

diversity

Therefore, log-normal fade margin and the soft handoff must be taken

in to account in the link budget. It is also necessary to include a receiver interference margin in order to

avoid overly optimistic estimate of the path loss

This margin depends on the cell loading, which indicate the percentage

of the maximum number of users that the cell has been designed ,clearly the greater the loading, the larger the interference margin

should be

I f M i

8/2/2019 Ee4105 Wcdma New(1)

89/105

89

Interference Margin

budgetlinkin theusedbeshoulddB3ofmarginfadeain thisSo

dB35.0-1

1log10marginreceiverthesoand

,5.0system,theusingareusers15averageon,If(users)30iscellsectorthreeaofcellpercapacityaifexample,anAs

factor.loadingtheisWhere

-1

1log10MarginceInterferenReceiver

formulafollowingtheusetoismarginthisestimateway toOne

f

f

f

==

=

=

l

l

l

interference

F t F di M i

8/2/2019 Ee4105 Wcdma New(1)

90/105

90

Fast Fading Margin

The fast power control can be effectively used to overcome fading for

slow moving vehicles ( 10 km/h ). At higher speeds, say, 120 km/h or higher, the number of fades per

second is significantly higher, fade duration is lower than for vehicle

speeds of, say 10 km/h.

As a result, the fast power control can not compensate for fading athigher speeds

To compensate for inaccuracies in power control algorithms , a fast

fading margin of 2 to 5 dB should included in the link budget forlow

speed vehicles.

Li k b d t ti d

8/2/2019 Ee4105 Wcdma New(1)

91/105

92

Link budget continued

The input to the base station receiver,

Pin= 24 Body loss penetration loss- Path loss + BS receiver antennagain- cable loss= 24- 2 8 PL+ 15-1 = 28 - PL dBm

Noise interference density at the BS receiver = -174 + receiver noisefigure + interference margin = -174 +5 +3 = -166 dBm/Hz

The input to the receiver must provide a Eb/No= 7 dB and 8 dB log-

normal fade margin and must support 14.4 kb/s So, the required input signal = - 166 +7+8+10log(14400) = -109.4

dBm

The soft handoff gain is 2 dB , so the required input = -109.4 2= -

111.4 Therefore, Pin 28 - PL - 111.9 or PL 139.4 dB = maximum path

loss

C ll Si

8/2/2019 Ee4105 Wcdma New(1)

92/105

93

Cell Size

If a propagation model is known, the path loss can be used todetermine the cell size. For example, if the base station height is 50 m,the mobile station antenna height is 1.5 m and the carrier frequency900 MHz, the following the Hata-Okumura model for a large city , the

propagation loss is given by

L = 123.33 + 33.77 log r

So, for a maximum path loss of 139.4 dB , r = 2.99 km. In other words,the maximum cell radius is 2.99 km. This value can then be used todetermine the number of cells required to provide the desire coveragein a serving area. The signal strength will be different at different

points within a cell depending on the terrain and clutter. However,

since necessary margins have been included in the design, the signalstrength everywhere in the cell will be within the prescribed limit.

Uplink W CDMA li k b d t

8/2/2019 Ee4105 Wcdma New(1)

93/105

94

UplinkW-CDMA link budget

Supports a number of services at various data rates. One of them is

delay tolerant interactive data service (web browsing) or a file transferat 384 kbps or more in urban or suburban environments for pedstrians

Let us consider a W-CDMA application involving non-real time data

transfer at 256 kbps in an urban area at low vehicle speeds.

A low Eb/No can be used for this service than for speech or real-timemulti- media applications

Even though the vehicle speed is low , a fast fading margin of about 3

dB is generally used.

Let us calculate the allowable path loss for the parameters given on the

next page

W CDMA l l ti

8/2/2019 Ee4105 Wcdma New(1)

94/105

95

W-CDMA calculations We will consider the link

budget on a reverse link for W-CDMA system

Chip rate = 3.84 Mc/s

The mobile transmitter power =250 mw = 24 dBm

Body loss = 2 dB In-vehicle penetration loss = 8

dB

BS receiver antenna gain = 16

dB Receiver cable loss = 2 dB

Receiver noise figure= 5 dB

Receiver interference margin =3 dB

Information rate = 256 kbps

Eb/No= 6 dB

Soft handoff gain = 2 dB

Fast fading margin = 3 DB

Log-normal fade margin = 8 dB

We are required to calculate themaximum allowable path loss

Suppose the path loss from

mobile station to base station isPL

Path loss and cell size

8/2/2019 Ee4105 Wcdma New(1)

95/105

96

Path loss and cell size The input to the base station receiver,

Pin= 24 penetration loss- Path loss + BS receiver antenna gain- cableloss= 24 8 PL+ 16-2 = 30 - PL dBm

Noise interference density at the BS receiver = -174 + receiver noisefigure + interference margin = -174 +5 +3 = -166 dBm/Hz

The input to the receiver must provide a Eb/No= 6 dB and 8 dB log-

normal fade margin and must support 256 kb/s So, the required input signal = - 166 + 6 +8+10log(256000) = - 97.9

dBm

The soft handoff gain is 2 dB , so the required input = -97.9 2= - 99.9

dBm Therefore, Pin 30 - PL - 99.9 or PL 129.9 dB = maximum path

loss

Cell size

8/2/2019 Ee4105 Wcdma New(1)

96/105

97

Cell size

If a propagation model is known, the path loss can be used to

determine the cell size. For example, if the base station height is 50 m,the mobile station antenna height is 1.5 m and the carrier frequency

900 MHz, the following the Hata-Okumura model for a large city , the

propagation loss is given by

L = 123.33 + 33.77 log r

So, for a maximum path loss of 129.9 dB ,

33.77logr= 129.9-123.33 or log r = 0.1946

r= 10^0.1946=1.5653 km =1.57 km

Notice that if the requirement on the signal-to- noise ratio is relaxed sothat the QoS is slightly lowered for all users, radius of the cell will

increase

Beyond 3G

8/2/2019 Ee4105 Wcdma New(1)

97/105

98

Beyond 3G

The demand for the mobile telephone services has been phenomenal.

Since the introduction in 1981, the annual growth in the mobilesubscribers has been about 40%, whereas the telephone services over

fixed networks has grown at a rate of about 5-7%.

Most of the traffic in present day Mobile telephony consists of voice,

however, the demand for mobile data has gone up steadily, spurred to a

large extent by the availability of Internet-based applications

The data service of the earlier GSM was limited to short messaging

service (SMS) and circuit switched data at a rates up to 9.6 kbps. As

the demand for data services began to grow ETSI developed a standard

for GPRS (General packet radio service), to provide packet mode dataservice at 12-20 kbps per slot.

The growth forecast in data and voice is shown on the next page.

The mobile data traffic is

expected to double the

voice traffic by 2010 and

8/2/2019 Ee4105 Wcdma New(1)

98/105

99

voice traffic by 2010 and

increase by a factor of

about 24 in the year 2015.

2G and 3G systems may

run out of capacity by that

time and it may be

necessary to consider

allocation for new radio

spectrum to wirelesscommunications for next

generation systems

4G

8/2/2019 Ee4105 Wcdma New(1)

99/105

100

4G The source of most of the traffic in the above scenario is expected to

be multimedia services. The data transport will be asymmetrical(mostly)- downlink traffic will be much more than uplink.

In 3G, multimedia services for mobile outdoor applications willoperate at 384 kbps.

Mobile networks may be required to provide multimedia service at 2

Mbps same as the fixed network Other applications such as full-motion video may require rates at 10-20

Mbps

Since the maximum data rate for fixed and indoor applications in 3Ggoes only up to 2 Mbps

This may be another reason to consider 4G. Figure on next page showsvarious systems with their services and applications.

8/2/2019 Ee4105 Wcdma New(1)

100/105

101

Voice

GPRS: up to 170

kbps; 64-115kbps.

SMS, Mobile IP

Applications and features of 4G

8/2/2019 Ee4105 Wcdma New(1)

101/105

102

Applications and features of 4G Possible applications of 4G include multimedia services for mobile

environments (vehicular, aeronautical, satellite and so on) at rates up to2 Mbps, compact disc (CD) quality radio broadcasting , video

surveillance of ones home, full- motion video and home

entertainment at rates up to 20 Mbps for indoor applications , position

locating systems and so on.

The goal is to provide multimedia service to anyone, anywhere,

anytime.

Unrestricted, seamless roaming and global mobility not only for voice,

but also for data services over regional and global networks ( likely to

be all-IP architecture Interoperability between 3G and 4G and between 2G and 4G

Time frame of 4G

8/2/2019 Ee4105 Wcdma New(1)

102/105

103

Time frame of 4G Analog systems were introduced in 1981 in US and Europe

Within 10 years , around 1991, digital systems were deployed 3G systems are targeted for 2001 and 2002

It is reasonable to expect a probable time frame for 4G would be

around 2010-2012

Even though the present systems have not reached saturation, it may benecessary to consider new spectrum allocation

It takes about three to four years for the standards to develop, standards

work might begin in two years

Technologies

8/2/2019 Ee4105 Wcdma New(1)

103/105

104

Technologies Technologies that are likely to play a key role in the development and

eventual success of 4G and , to lesser extent, 3G are

Software radio- In software radio, most of the processing is done withdigital signal processors. For example, baseband data processing (pulseshaping, error coding and so on), modulation, and up-conversion at thetransmitter, channel separation , demodulation , detection , and

baseband data processing at the receiver are all performed in thedigital domain. In addition, DSP is used to characterize the channeland adjust power level as needed, analyze the received signal todetermine the quality ( BER, forward error correction, etc.), reduce orcancel interference from specific sources , implement multipath

diversity and so on. DSP is crucial for wireless system implementation and will continue to

play a major role in future 4G systems

Technologies continued

8/2/2019 Ee4105 Wcdma New(1)

104/105

105

Technologies continued Adaptive antenna arrays- Adaptive arrays may be used for beam forming in a

specific direction so as to provide extended coverage in certain areas

Development of suitable multimode configurable terminals, special keyboards,

and video displays. We might consider wearable PCs with voice activated,

hands-free operation, specially designed keyboards that can be strapped to

wrist , and small LCD displays with magnifying optics attached to, or reflected

onto, the users glasses. QoS- For efficient utilization of bandwidth, the network must implement a

flexible resource management scheme to provide mobile stations with an end-

to-end QoS across all IP- networks.

Reduced power levels- data rates in 4G are much higher than 2G and 3G, this

will result in higher power consumption . The terminals should be designed tooperate at reduced power levels.

Economic Perspective

8/2/2019 Ee4105 Wcdma New(1)

105/105

106

Economic Perspective The discussion so far strictly from technical point of view. The

technology necessary for the 3G and 4G exists or in the process ofdevelopment. There is , however, an economic perspective.

Millions ofdollars were spent on 3G licenses in Western Europe.

3G infrastructure needs considerable amount of capital

Thus to ensure a reasonable rate of return, it is necessary to generatesufficient customer demand for 3G services and continue to use the

(3G) infrastructure for 4G and beyond.

It would be necessary to develop applications that are meaningful and

attractive to customers, at the same time commercially viable from

service providers point of view.