wcdma radio link performance

8/19/2019 WCDMA Radio link performance

http://slidepdf.com/reader/full/wcdma-radio-link-performance 1/8

1 ©NOKIA WCDMA_rlperf.PPT/ 1.2.2002 Kari Heiska, TLI361-WCDMA radioverkkosuunnittelu - luentosarja Jyvaskylän yliopistossa

WCDMA Radio Link Performance1.2.2002 Kari Heiska


Contents• General

• Definitions

• Multipath channel conditions

• Link-level Simulations

• Physical Layer Measurements




General

• Link performance indicators are used in radio network planning anddimensioning as an input data

• Realistic values needed

• For example one important link performance indicators is E b /N 0setpoint (=bit-energy per noise density for which the frame error rateover the radio link exceeds the predefined value)

• Link performance indicators depend on: service, radio propagationchannel, mobile speed, diversity

• These can be dependent on environment and cell deployment

• Very difficult to obtain from real cases standards needed

• Produced with link level simulations / measurements


Definitions• Radio link

• In this case term radio link means the physical layer connectionbetween one transmitter and one receiver (transmitted bit toreceived bit)

• Radio link modeltypically includes:

• channel coding/decoding, interleaving, rate matching• spreading/despreading• power control• radio channel (fading channel, gaussian noise modeling other user's interference)• RAKE receiver, channel estimation, SIR-estimation• detection

but typically does not include:• handovers• multiple users, multiple services• RF (carrier frequency, automatic gain control (AGC), etc.)• code acquisition and tracking

• All these items can be modeled, as well but typically this is not thecase




Definitions• BLER

• Block Error Rate. Long term average calculated for the transport blocks (TB).

• e.g. 100000 blocks with 1 error Black (at least one bit error in the block)BLER=1e-5

• BER

• Bit Error Rate after decoding. The channel BER (Raw BER) is a allways higher thanBER

• Bit Rate, R

• Rate of user information bits =bit rate before L1 processing (CRC, coding, etc) =output from L2 (MAC)

• Eb/N0and Orthogonality• Originally ’bit-energy over noise spectral density’

• Eb/N0relates to respective quality target (e.g. BER/FER)

• In uplink theE b /N 0 is deined as: Eb/N0=(prx/R)/(I/W)=W/R·prx/I, where prx is thereceived constant power, R is the user bit-rate, I is the received interfernecepower and the W is the bandwidth

• Target of the power control is to keep Eb/N0 constant


Definitions• Orthogonality

• Short codes are orthogonal when the codes are parallel

• In the case of single path channel Interference from the own basestation is small (I own =P rx (1- α )=0 , α =1

• In the case of multipath channel the orthogonality, codes are notany more parallel. Interference from the own base station isI own =P rx (1- α ),whereα is the orthogonality of the channelα<1

• The orthogonality is highly varying and only mean value is used inthe radio network planning

two tap channel (α =0.5)

single tap

channel (α =1)




Definitions

E c /I 0

• Received chip energy relative to total power spectral density. Inuplink E c /I 0= E b /N 0 /(W/R) and in downlink I 0 is the total receivedpower (wideband power) without any effects of orthogonality.CPICH E c /I 0 is used for example when comparing different linkstrengths in handover algorithm

Ec

Ioth

I 0=(I own+I oth+N)/W, [Watts/Hz]

Iown

E c=Prx /W, [Watts/Hz]


Definitions

• E c /I or

• E c /I or is the transmitted energy per chip relative to the total transmitted powerspectral density. This is the same as the received energy per chip relative to thetotal (wideband) received power spectral density

• E c /I or =(Ptx,channel /W)/(Ptx /W)= Ptx,channel /Ptx for example

• Average Power Rise

• When the MS speed is low, the fast power control is able to compensate the fast

fading (low E b /N 0 values for a certain BLER)• This increases the average transmitted power !!

• The average power rise is measured from simulations as the difference betweenthe average transmitted power and the average received power when the averagechannel gain is 1

• Increases the average interference to adjacent cells. Has to be taken into accountin dimensioning

• In downlink the average power rise is effecting the total increase in BStransmission power and has to be taken into account in downlink E b /N 0 values




Definitions• Power control headroom

• In the cell border the fast power control starts to hit the maximum Tx power (21dBm / 24 dBm). This means that the BLER starts to increase. The margin isrequired in order to ensure good enough link quality at the cell border, as well. This margin is called power control headroom

• Macro Diversity Combining Gain (MDC gain)

• =the reduction of Eb/N0 per link in soft(er) handover compared to the single linkcase.

• Uplink

• The gain in received Eb/N0 values is relatively small (good TPC)

• The gain in transmitted powers is significant (cell edge)

• Downlink

• Decreases the needed Tx power per BS. However, more BSs needed. The net-effectcan be quite small

• SHO gain≠ MDC gain

• In SHO gain the link is connected to the best cell and can, in principle, obtainedwith very fast hard handover


Definitions• Little i

• Other-to-own cell interference ratio: little i =total power received from the owncell per total power received from surrounding cells

• Gives the isolation of the cell; i.e. how well the radio cell is isolated from itsadjacent cells

• In microcells i≈0.2: Good isolation because of surrounding buildings goodspectral efficiency =good capacity

• In macrocells i≈0.6-0.8: Worse isolation, cells are overlappingworse spectraldensity

Ec

Ioth

Iown

own

other

I

I i =




Definitions

• Geometry Factor

• Used in downlink performance evaluation

• When the network is interference limited ( I other >P N ): G≈ 1/i

• Typically –3 dB for the cell edge and 20 dB close to the BS

N other

own

P I

I G

+

=

G≈-6dB…-3dB

G≈20dB


Definitions• 3GPP Multipath Channel Models

• For BS and MS performance test in specifications• The main performance indicators (like Eb/N0) are dependent on the propagation

channel conditions• Well defined radio channel models:• For example one 3GPP model:

• MS speed 3 km/h,• tap1: delay 0 and average tap power 0 dB,• tap2: delay 976 ns and average tap power –10 dB

• ITU Multipath Models• Similar models like 3GPP• Models:

• IndoorA, IndoorB• Outdoor-to-indoorA, Outdoor-to-indoorB• VehicularA, Vehicular B

• Reference Measurement Channel and the Link Performance in the Standard• Reference services: 12.2 kbps, 64 kbps, 144 kbps, 384 kbps and 2048 kbps• Uplink (UL) and downlink (DL)• Link performance of above shown channels are in standard• In UL average Eb/N0performance with constant BLER without TCP• In DL average required Ec/Ior for fixed BLER @ selected G

• Common channels not defined in 3GPP rel 99 (latest accepted standard)




Link Level Simulations

• Uplink simulations

Data Transmitter Channel

AWGN noise

Power control

Receiver Evaluation+

SIR BLER

• Outputs• Received power prx

• (Own cell capacity)

• Tx power, ptx average power rise• (Interference to other cells)

• BER/BLER

BLER)accepted(@ /

/

0 W I

R p

N

E rxb=


Link Level Simulations• Downlink Simulations

Data Transmitter Channel

AWGN noise

Power control

Receiver Evaluation+

SIR BLER

CPICH

User 1

User N

+

• Interference from the own cell has to be taken into account

• CPICH used for the channel estimation in DL

• In uplink other users interference can be modelled with AWGN whi isindependent on the MS location

• In DL the own cell users can cause interference which is dependent on the MSlocation

+−==

+−

⋅

GW

R

I

p

G

I

p

R

W

own

tx

own

tx 1)1(

1)1(

1α

ρ ρ

α




Physical Layer Measurements

• The typical measurements defined in the standardization characterizingthe link performance in live network

• UE measurements• CPICH RSCP

• Received power of the CPICH channel measured by the UE (path loss estimation)

• RSSI• Total wideband power

• CPICH Ec/N0• =CPICH RSCP/ RSSI

• Used for Handover measurements• Transport Channel BLER (Block Error Rate)• UE Transmitted power


Physical Layer Measurements

• Network measurements• RSSI (Total received power)• SIR (Signal to interference ratio)

• (RSCP/ISCP)*SF• RSCP =Received power of the DPCCH channel in question• ISCP =Received total power (=interference) of the uplink quadrature channel• SF is the spreading factor of the DPCCH (=256)

• SIRerror• Difference betweeh SIR and SIR target in the closed loop PC

• Transmitted carrier power• Ratio between the total transmitted power and the maximum transmission power of

the base station

• Transmitted code power• Transmitted power (BS) of one code channel

• Transport Channel BER• Physical Channer BER (=Raw BER)

wcdma radio link performance

Documents