A New Technique for Sidelobe Suppression

1
A New Technique for Sidelobe Suppression in OFDM
Systems
Sinja Brandes
German Aerospace Center (DLR) Institute of
Communications and Navigation Oberpfaffenhofen,
Germany
COST 289, 7th MCM, Oberpfaffenhofen, Germany 7
March, 2005
2
Overview

• Problem Definition and Techniques for Sidelobe
  Suppression
• Principle of Cancellation Carriers
• Simulation Results
• Comparison With Existing Methods
• Application OFDM Overlay Systems
• Summary and Outlook

3
Spectrum of an OFDM Signal
significant out-of-band radiation
4
Techniques for Sidelobe Suppression

• Pulse shaping
• e.g. raised-cosine pulse instead of
  rectangular pulse
• ? intersymbol interference (ISI)
• ? (I)FFT cant be applied for modulation/demodula
  tion
• Windowing of transmission signal in frequency
  domain
• ? expansion of signal in time domain
• ? intersymbol interference (ISI)
• Guard bands at the borders of the OFDM spectrum,
  e.g. DVB-T
• ? waste of scarce spectral ressources, DVB-T
  ca. 16

5
New Idea Cancellation Carriers
cancellation carrier
cancellation carrier

• cancellation carriers are not
• used for data transmission
• cancellation carriers carry
• complex weighting factors
• weighting factors are
• determined such that the
• sidelobes of the transmission
• signal are minimized

data carriers
sidelobes that should be suppressed
sidelobes that should be suppressed
6
Optimization Constrained Least Squares
s Vector of sampled signal in optimization range
g Vector of weighting factors
C Matrix with non-weighted sampled cancellation
carriers in the columns
Constraint Limit power of cancellation carriers

• Degrees of freedom
• Position of cancellation carriers in the
  spectrum
• Number of cancellation carriers
• Different constraints
• Optimization range

7
Spectrum With Weighted Cancellation Carriers
amplitude
8
Spectrum With and Without Cancellation Carriers
Parameters BPSK, 12 data carriers 2x1
cancellation carrier optimization range 32
sidelobes
ø - 19 dB
9
Spectrum With and Without Cancellation Carriers
Parameters BPSK, 12 data carriers 2x2
cancellation carriers optimization range 32
sidelobes
symbol vector 1 1 1 1 1 1 1 1 1 1 1 1

ø - 40 dB
10
Suppression of Sidelobes

• x-axis
• all possible symbol vectors
• y-axis
• mean power spectral density
• of sidelobes for each symbol vector
• BPSK
• 12 data carriers
• unconstrained optimization
• 1 cancellation carrier
• mean power for CCs
• 19 of total power
• 2 cancellation carriers
• mean power for CCs
• 45 of total power

- 16 dB
- 34 dB
11
Power Ratio of Cancellation Carriers to Total
Power for All Possible Symbol Vectors
power of cancellation carriers 30
unconstrained
mean 0.26 max 0.30
mean 0.45 max 0.92
87 of symbol vectors use maximum amount of
power for cancellation carriers
12
Mean Sidelobe Suppression for Different
Constraints
Parameters BPSK 12 data carriers 2x2
cancellation carriers optimization range 32
sidelobes
- 28 dB
- 26 dB
power of cancellation carriers unconstrained l
imited to 50 of total power limited to 40 of
total power limited to 30 of total
power limited to 20 of total power
- 23 dB
- 20 dB
- 34 dB
13
Bit Error Rate Performance
Parameters 2x2 cancellation carriers, 12 data
carriers, BPSK no coding, AWGNRayleigh-fadin
g channel, Zero ForcingHard Decision
Amount of power spent for CCs Maximal SNR loss (symbol vectors with max. power) Real SNR loss
unconstrained ? (0) 3.5 dB
50 3 dB (50) 2.4 dB
40 2.22 dB (67) 1.9 dB
30 1.55 dB (83) 1.4 dB
20 0.96 dB (91) 0.89 dB
10 0.45 dB (98) 0.44 dB
5 0.22 dB (100) 0.22 dB
.
14
Application OFDM Overlay Systems
subcarriers used by licenced systems
free subcarriers used by OFDM overlay system

• challenges
• co-existence of both systems
• avoid interference towards
• licenced system

task suppress sidelobes
15
Application OFDM Overlay Systems
subcarriers used by licenced systems
free subcarriers used by OFDM overlay system
Parameters BPSK 13 (58) data carriers 4x2
cancellation carriers optimization range all
displayed sidelobes
joint optimization of all cancellation carriers
16
Comparison With Existing Methods

• Advantages
• (I)FFT can still be applied for
  modulation/demodulation
• no additional ISI
• smaller guard bands (some guard carriers can be
  used for data transmission)
• Possible drawbacks
• slight loss in BER performance
• computational complexity of least squares
  optimization

17
Summary and Outlook

• Principle of cancellation carriers
• cancellation carriers are not used for data
  transmission, but carry complex weighting factors
• weighting factors are determined such as to
  minimize sidelobes of the transmission signal
• significant reduction of sidelobes
• promising approach for sidelobe suppression
  in
• overlay systems
• Further investigations
• optimization of minimization algorithm and
  parameters
• implementation in overlay systems
• influence on PAPR