DVB Detector for Cognitive Radio

1
DVB Detector for Cognitive Radio

• Loo Peng Goh, Z. Lei, and Francois Chin
• Presenter Manjeet Singh
• Institute for Infocomm Research, Singapore

2
Outline

• Introduction
• Spectrum Utilization gt Cognitive Radio
• Cyclostationary of DVB-T Signal
• Detection Techniques
• Complexity Analysis
• Detection Results of DVB-T under AWGN
• Conclusions Future Work

3
Introduction

• Societies Needs
• Increased Mobility
• More dependent on information technology
• MORE SPECTRUM!!!
• FCC Frequency Chart
• Running out of Spectrum
• Actual Measurement at Berkeley CA
• Allocated Spectrum is vastly Underutilized

4
Spectrum Utilization

• In order to utilize these spectrum white
• spaces
• Cognitive Radio
• TV Band has been considered by FCC to be
  pioneered for CR usage
• IEEE 802.22 Wireless Regional Area Networks
  (WRAN) is established to utilize the idle
  spectral bands of TV channels.

5
Cyclostationary of DVB-T Signal

• Conventionally, the most commonly and simplest
  approach for detection is based on radiometry
  Energy Detector (ED)
• However, ED can be highly susceptible to unknown
  changing noise level or noise uncertainty
• William Gardner Periodicity or cyclostationary
  of modulated signals can be exploited for
  detection gt Single Cycle Detector (SCD)
• Although SCD is robust against noise ambiguity,
  it is much more complex than ED
• For real time implementation,4 feature detectors
  are scrutinized Blind Twin Peak Detector (BTPD)
  is proposed.

6
Cyclostationary of DVB-T Signal

• A process x(t) is said to be cyclostationary in
  wide sense if its mean autocorrelation are
  periodic with some period T
• mx(t T) mx(t)
• Rx(t T t/2, t T - t/2 ) Rx(t t/2, t -
  t/2 )
• Rx(t t/2, t - t/2 ) can be represented as a
  Fourier series

7
Cyclostationary of DVB-T Signal

• The Fourier coefficients can be calculated as
• which is referred as Cyclic Autocorrelation
  Function (CAF)
• Fourier transform of CAF is defined as Cyclic
  Spectrum Function (CSF)
• a is the cyclic frequency

8
Cyclostationary of DVB-T Signal

• Relationship between CAF AF
• Relationship between CSF PSD

? a0
9
Cyclostationary of DVB-T Signal

• DVB-T Specification for 2k mode 8 Mhz channel

10
Cyclostationary of DVB-T Signal

• OFDM Signal
• where
• f0 is the frequency of the carrier
• ?f is the frequency offset between carriers
• TU is the useful duration of the OFDM symbol
• TS is the duration of the OFDM symbol
  duration
• TG is the guard duration of the OFDM symbol
• CAF for OFDM Signal

11
Cyclostationary of DVB-T Signal
12
Cyclostationary of DVB-T Signal
13
Detection Techniques

• Noise present at a t 0 only
• Discrete CAF surfaces for a kfs
• Peak at t TU (maximum when a 0)
• When a 0,
• G( f )G( an f ) G( f )G( -f ) G2( f )
• ? where a 0, t TU

14
Detection Techniques

• Different Cases
• Energy Detector (ED)
• Single Cycle Detector (SCD)
• Proposed Detectors (PD)
• PD1 Blind Slingle Peak Detector (BSPD)
• PD2 Blind Twin Peak Detector (BTPD)
• PD3 Twin Peak Detector (TPD)
• PD4 Selective Single Cycle Detector (SSCD)

15
Detection Techniques
16
Complexity Analysis
17
Complexity Analysis
18
Detection Results of DVB-T Signal under AWGN
19
Detection Results of DVB-T Signal under AWGN
20
Conclusion

• Single-Cycle Detector
• Best Performance but Most Complexity
• Energy Detector
• Less Complexity but prone to Noise Uncertainty
• Proposed Detectors
• PD2(Blind Twin Peak Detector) has best among the
  4 PD and 2nd in the complexity ranking.
• Hence, BTPD is proposed for WRAN

21
Future Work

• Study detector performance in a frequency
  selective fading channels
• THANK YOU.