Baseband Transceiver Design for the DVB-Terrestrial Standard

(1)

NTU Confidential

Baseband Transceiver Design for the DV

B-Terrestrial Standard

Advisor : Tzi-Dar Chiueh

Student : Yi-Ju Chen

(2)

2

NTU Confidential

Outline

• Review of DVB-T SystemReview of DVB-T System

• Transmitter Block DiagramTransmitter Block Diagram

• Channel ModelChannel Model

– Static ChannelStatic Channel

– Dynamic ChannelDynamic Channel

• Receiver ArchitectureReceiver Architecture

– Coarse Boundary DetectionCoarse Boundary Detection

– Integer CFO EstimationInteger CFO Estimation

– WLS Fine CFO EstimationWLS Fine CFO Estimation

• Future WorkFuture Work

(3)

3

NTU Confidential

What is DVB-T ?

• DVB-T stands for Digital Video Broadcasting – Terr estrial

– Wireless video

• DVB-T uses COFDM technique • DVB-T Transmit Block Diagram

(4)

4

NTU Confidential

System Parameters

• Center frequency: 480MHz + ｉ *6MHz

• Region: 480MHz ~ 806MHz  Channel 14 ~ Channel 69

6MHz Channel

Transmission mode 8k mode 2k mode Number of carriers K 6817 1705

Duration TFFT 1194.667 us 298.6667 us

Carrier spacing 1 / TFFT 0.837054 kHz 3.348214 kHz

Bandwidth 5.71 MHz

Modulation QPSK , 16QAM , 64QAM Code rate 1 / 2 , 2 / 3 , 3 / 4, 5 / 6 , 7/8 Guard interval ratio 1 / 4 , 1 / 8 , 1 / 16 , 1 / 32

(5)

NTU Confidential

Channel Model

Static Channel

Dynamic Channel

(6)

6 NTU Confidential

Channel Model (1/4)

Static Channel

• Fixed Reception

– where a directional receiving antenna mounted at roof level is used

– A receiving antenna height of 10m above ground level is considered to be

representative

• Portable Reception

– Portable receiver with attach ed or built-in antenna

– Absence of receiving antenn a gain and directivity

– Generally lower reception he ight

(7)

7 NTU Confidential

Channel Model (2/4)

Static Channel

(8)

8 NTU Confidential

Channel Model (3/4)

Dynamic Channel

• Typical Urban Reception(TU6)

• Typical Rural Area Reception ( RA6 )

Derived from COST 207 project (GSM transmission) [2]

Tap

number Delay (us) Power (dB)

Doppler spectru m 1 0 -3 Rayleigh 2 0.2 0 Rayleigh 3 0.5 -2 Rayleigh 4 1.6 -6 Rayleigh 5 2.3 -8 Rayleigh 6 5.0 -10 Rayleigh Tap numbe r Delay (us) Power (dB) Doppler spectru m 1 0 0 Rice 2 0.1 -4 Rayleigh 3 0.2 -8 Rayleigh 4 0.3 -12 Rayleigh 5 0.4 -16 Rayleigh 6 0.5 -20 Rayleigh [2]

(9)

9 NTU Confidential

Channel Model (4/4)

Dynamic Channel

• SFN( Single Frequency Network) Channel [2]

Tap number Delay(us) Power Doppler spectru

m

Frequency ratio

1 0 0 Pure Doppler -1

(10)

1 0

NTU Confidential

BaseBand Channel Model

(11)

NTU Confidential

Receiver Architecture

(12)

1 2

NTU Confidential

Receiver Block Diagram

(13)

1 3

NTU Confidential

Coarse Symbol Boundary Detection (1/5)

GI-1 * 0 (d) r(d-n) r(d-n-N) n Correlation Sum   





DVB-T System doesn’t have preamble to do symbol boundary detection, but we can utilize the cyclic prefix to implement it.

(14)

1 4

NTU Confidential

Coarse Symbol Boundary Detection

(2/5)

• For 1 tab Channel profile the result of Correlation Sum is as follows, • The correlation sum appears

triangle and changes slowly

• If we transmit the preamble to do auto correlation the correlation sum will be an ideal delta function • The peak is interfered by noise

(15)

1 5

NTU Confidential

Coarse Symbol Boundary Detection

(3/5)

• As for multi-path channel profile, take a 2-path channel for example

• The correlation sum peak becomes ambiguous

• The peak might occur in the boundary which induces ISI 1 * 0 1 * * * * 1 2 1 2 0 1 1 * * * * 1 1 2 2 0 0 1 * * 1 2 0 ( ) ( - ) ( - - ) ( ( - ) ( - - )) ( ( - - ) ( - - - )) ( ( - ) ( - - )) ( ( - - ) ( - - - )) ( ( - ) ( - - - )) ( GI n GI n GI GI n n GI n d r d n r d n N p x d n p x d n p x d n N p x d n N p x d n p x d n N p x d n p x d n N p x d n p x d n N p                                   1 * * 2 1 0 (1' ) (2' ) (1' , 2' ) (2' ( - - ) ( - - )) , 1' ) GI n

AutoCorr st Path AutoCorr nd Path CrossCorr st Path nd Path

x d n p x d n

CrossCorr nd Path st Path

N         

(16)

1 6

NTU Confidential

Coarse Symbol Boundary Detection

(4/5)

• In the SFN Channel Model, there is no peak

(extreme example)

• Add moving average to find the max.

average power position

(17)

1 7

NTU Confidential

Coarse Symbol Boundary Detection

(5/5)

• We can observe that

the resulting boundary may induce ISI because of the delay spread of the channel

• Desired FFT window = Max Boundary Position

(18)

1 8

NTU Confidential

Fractional CFO Estimation

(19)

1 9

NTU Confidential

Fractional CFO Estimation

• If we assume the normalized CFO is 

• Can’t use this formula to calculate Integer part of  GI-1 GI-1 * 2 2 0 0 (d) r(d-n) r(d-n-N) |r(d-n)| j n n Correlation Sum e     



 



    1 * 0 1 * 0 Im 1 arctan 2 _Re GI d n d n N n GI d n d n N n r r Fractional Normalized CFO

r r                           2 2 ( ) ( ) j N N ( ) j r k N r k e r k e       * 2 2 ( ) ( ) | ( ) | j r k N r k  r k e 

(20)

2 0

NTU Confidential

Performance of Different Ns

(21)

2 1

NTU Confidential

Integer CFO Estimation

(22)

2 2

NTU Confidential

Integer CFO Estimation(1/2)

• Integer Frequency Offset (Normalized by Sub

Carrier Spacing) will cause the sub-carrier index shift error

• Using Continual Pilots to find the shift of the

index _* 1, , ˆ _max l k l k m k Pm f Y _ Y     _  _ 





Y : Received Signal in Freq. Domain P_m : [ p₁+m , p₂+m , ……, p_L+m ] Continual Pilots Position shift m l : l ’ th Symbol

(23)

2 3

NTU Confidential

Integer CFO Estimation(2/2)

• Block Diagram

(24)

NTU Confidential

Fine CFO Estimation Loop

WLS Estimation

(25)

2 5

NTU Confidential

Fine CFO Estimation

Joint WLS Estimation(1/2)

[4]

WLSE Block Diagram

Assume the source of CFO and TFO is the same, by the

result of joint WLS Estimation k

k

y : Phase Difference between 2 Symbols of k'th sub-carrier index w : Weighting factor of k ' th sub carrier index

GR : Guard Interval Ratio



(26)

2 6

NTU Confidential

Fine CFO Estimation

Joint WLS Estimation(2/2)

• Simulation Result

Adding extra error=0.04

to see the convergence System Simulation result _{The result of fractional CFO is in the} convergence region already

(27)

2 7

NTU Confidential

Conclusion and Future Work

• Conclusion

– Although there is no preamble, the DVB-T can exchange time for good acquisition performance

– It is the advantage of broadcast system

• Future Work

– Design unfinished block

– Add channel coding and outer decoding

(28)

2 8 NTU Confidential

Reference

• [1]ETSI EN 300 744 V1.4.1(2001-01)

Framing structure, channel coding and modulation for digital terrestri al television

• [2]MOTIVATE report to the 36th DVB-T Meeting (2000-01)

Using DVB-T standard to deliver broadcast Services to mobile receiver

• [3]Joint weighted least squares estimation of frequency and timi ng offset for OFDM systems over fading channels

Pei-Yun Tsai; Hsin-Yu Kang; Tzi-Dar Chiueh;

Vehicular Technology Conference, 2003. VTC 2003-Spring. The 57th IEEE S emiannual , Volume: 4 , April 22-25, 2003

• [4]Design and Implementation of an MC-CDMA Baseband Transceiver

Hsin-Yu Kang; July , 2003

• [5] Frequency synchronization algorithms for OFDM systems suita ble for communication over frequency selective fading channels

Classen, F.; Meyr, H.;

Vehicular Technology Conference, 1994 IEEE 44th , 8-10 June 1994 Page(s): 1655 -1659 vol.3