Efficient Collaborative (Viral) Communication in OFDM Based WLANs

1
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs

• Aggelos Bletsas
• MIT Media Laboratory
• aggelos_at_media.mit.edu
• http//web.media.mit.edu/aggelos
• ISART 2003

2
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
The vision Collaborative Radio
1) multi-user diversity radio
2) ad-hoc phased arrays
3) multiple-input multiple-output systems
(MIMO)
3
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
OFDM is appropriate for collaborative radio

• 1) Software defined radio
• baseband signal through IFFT/FFTs
• information recovery through over-sampling

• 2) zero ISI regardless the network topology
  (guard interval)
• possibility for combined transmissions/recept
  ions

4
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
OFDM Properties and Viral Communication
(analytically)

• Software defined baseband signal through
  FFT/IFFTs

• Information recovery through over-sampling

5
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
OFDM Properties and Viral Communication
(analytically)

• Guard interval ensures zero ISI regardless the
  network topology
• (possibility for combined transmission/recept
  ion)

• and a last property
• Convolution with channel impulse response
  multiplication of the original symbol
  with the frequency response.
• More analytically

6
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
7
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs

• Baseband representation at each sub-carrier
• based on analogue Amplifying and Forwarding
  Laneman 2000

• Intermediate node oversamples at first half and
  forwards at second half.
• Receiver oversamples and receives direct at
  first half and
• relayed and direct at second half.

8
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Optimal ML receiver results in Maximum Ratio
Combining (MRC).

• Noise modeled as complex Gaussian with
  variance Ni/2 per dimension.
• Propagation coefficients according to complex
  Gaussian distribution

MRC
9
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Applying the results in our case
MRC receiver
Probability of error
10
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Comparison with direct (point-to-point) BPSK
ML Receiver
11
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Error Probability vs. SNR
E1E2E/2
12
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Error Probability vs. SNR
E1E2E/2
13
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs
Tx Energy ratio (point-to-point/viral) for given
error probability.
E/(E1E2) vs. Pg
v4
14
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs

• Collaboration is possible in OFDM based
  networks.
• Collaboration results in substantial energy
  savings (one order of magnitude or more).
• Energy savings can be realized at the same
  channel (better spectrum allocation - higher
  capacity?)
• Need for Network channel estimation and
• Network Time/Frequency Synchronization.
• (complexity increases challenging problem)

15
Efficient Collaborative (Viral) Communication in
OFDM Based WLANs

• Visit our web site
• http//www.media.mit.edu/aggelos