UWB WAVEFORM ESTIMATION AND TIMING SYNCHRONIZATION

1
UWB WAVEFORM ESTIMATION AND TIMING SYNCHRONIZATION

• Gregori Vàzquez, José A. López-Salcedo
• Signal Processing Communications Group
• Department of Signal Theory and Communications
• Technical University of Catalonia (UPC)
• Mòdul D5, c/Jordi Girona 1-3, Campus Nord UPC,
  Barcelona (Spain)
• Emailgregori,jlopez_at_gps.tsc.upc.edu

2
BACKGROUND

• WIDEBAND SPREAD-SPECTRUM COMMUNICATION SYSTEMS
• ESA/NASA systems.
• FREQUENCY HOPPING RECEIVERS
• Classified Project.
• - FAST CODE ACQUISITION SCHEMES BASED ON
  MULTIPLE-DWELL
• - TIME AND FREQUENCY HOPPING METHODS IN SINGLE
  AND MULTIPLE ACCESS
• (more detailed profile under request)

3
BACKGROUND

• 2003 BEST PAPER SENIOR AWARD OF THE IEEE SIGNAL
  PROCESSING SOCIETY in the area of Signal
  Processing for Communications
• DIGITAL TIMING SYNCHRONIZATION TECHNIQUES
• IEEE BEST PAPER AWARD INTERNATIONAL SYMPOSIUM ON
  TURBOCODES.

4
BACKGROUND/UWB

• THE WIRELESS ELECTRONICS CONCEPT
• UWB APPROACHES
• NON-COHERENT DETECTION (no knowledge of the
  waveform)
• COHERENT DETECTION
• PILOT-BASED AND NDA.

5
Index

• Research lines in UWB Signal Processing
• Waveform Estimation
• Waveform-Independent Timing Acquisition

6
Research Lines in UWB Signal Processing

• One of the major issues of UWB systems is the
  dramatic distortion of the transmitted pulse when
  propagating to the receiver.
• The severe multipath channel and the
  direction-dependent distortion of wideband
  antennas make the fast and accurate channel
  estimation a challenging and computationally
  demanding task.
• For this reason, the study of waveform-independent
  techniques becomes crucial. Some of the research
  lines on this topic within the SPCOMM Group are
• Channel estimation
• Non-data aided waveform estimation.
• Semiblind waveform estimation (transmitted
  reference approach).
• Synchronization
• Non-data aided and waveform-independent
  frame-timing acquisition.
• Detection
• Symbol detection in the absence of channel state
  information.
• Information theoretic aspects of UWB
  transmissions.

7
Subspace-Compressed Waveform Estimation

• The problem of waveform estimation is addressed
  under the low-SNR maximum likelihood criterion.
• A solution is proposed which is based on
  exploiting the signal subspace of the received
  signal.
• The result is a robust technique which overcomes
  most of the problems of traditional approaches to
  non-data aided channel estimation.
• The proposed method has been evaluated under the
  IEEE 802.15.3a channel model proposed by Intel
  Foe03.

Foe03 J. Foerster, Channel Modeling
Sub-committee Final Report, IEEE P802.15 Working
Group for Wireless Personal Area Networks
(WPANs), February 2003.
8
Subspace-Compressed Waveform Estimation

• Simulation results (I) Estimate of a random
  realization of the received waveform

• Simulation parameters
• Binary-APPM modulation
• 200 transmitted symbols
• 4 ns. 2nd derivative Gaussian pulse.
• Frame duration 94 ns.
• Frame repetition factor 64.
• PPM time-shift 8 ns.
• Channel model Intel CM1.
• Line-of-sight from 0-4 meters
• Delay spread 72 ns.
• Ef/N0 3dB

9
Subspace-Compressed Waveform Estimation

• Simulation results (II) Cumulative distribution
  of normalized mean square error

• Simulation parameters
• Binary-APPM modulation
• 200 transmitted symbols
• 4 ns. 2nd derivative Gaussian pulse.
• Frame duration 94 ns.
• Frame repetition factor 64.
• PPM time-shift 8 ns.
• Channel model Intel CM1.
• Line-of-sight from 0-4 meters
• Delay spread 72 ns.
• Ef/N0 3dB

10
Waveform-Independent Timing Acquisition

• Timing synchronization is essential for the
  recovery of the transmitted symbols.

• Main drawbacks of existing timing acquisition
  techniques
• An accurate estimate of the channel response is
  required for matched filtering the incoming
  signal.
• A high computational burden is required to obtain
  fast and accurate channel estimates in the
  presence of a direction dependent antenna
  distortions and severe multipath.
• In order to circumvent these limitations, a
  waveform-independent frame-timing recovery is
  appreciated
• No prior channel estimation is required ? a
  shorter start-up time is achieved.
• Simple channel estimation can be
• performed once timing is acquired ? significant
  complexity can be saved.

11
Waveform-Independent Timing Acquisition

• Main advantages of the proposed technique
• It is the optimal frame-timing estimator in the
  low-SNR UML sense.
• It does not depend on neither the rx waveform nor
  the transmitted symbols.
• It outperforms popular frame-timing acquisition
  techniques from the current literature.

• Main disadvantages
• The optimal solution requires a high
  computational burden.
• An efficient implementation is proposed which
  saves up to 75 of the computational load
  required by the optimal solution.
• The efficient implementation has no significat
  degradation in terms of BER.

Yan04 L. Yang and G.B. Giannakis, Blind UWB
Timing with a Dirty Template, Proc. IEEE Intl.
Conf. Acoustics, Speech and Signal Processing
(ICASSP), Montreal (Canada), May 2004.
12
Waveform-Independent Timing Acquisition

• Simulation results Bit error rate and
  Probability of correct acquisition

• Simulation parameters
• Binary-PAM modulation
• 4 ns. 2nd derivative Gaussian pulse.
• Frame duration 46 ns.
• Frame repetition factor 8.
• Channel model Intel CM1.
• Line-of-sight from 0-4 meters
• Delay spread 78 ns
• (causes frame overlapping)