Sparse OFDM

1
Sparse OFDM

• Erica Lynn Karen Guan

2
Objective

• Compress information into an OFDM signal that
  uses only a fraction of the available frequencies
• We desire a bit rate that is similar to standard
  OFDM
• We also desire a high spectral efficiency

3
The Idea

• Map the signal to a length N sparse vector
• Some bits can be represented as the indices of
  the used spaces.
• Other bits are represented as QAM signals in the
  used slots
• This vector is modulated as an OFDM signal by
  sending the vector through an IFFT

4
Bit Rate Comparison
Nsize of FFT k of frequencies utilized
Lbits per symbol
Bits per sparse OFDM vector Bits per ordinary
OFDM vector
5
Bit Rate Comparison
BPSK
Nsize of FFT ksparsitythe number of
frequencies utilized Lbits per symbol Bits per
sparse OFDM vector Bits per ordinary OFDM vector
6
Notes

• Sparse OFDM compares favorably to standard OFDM
  using BPSK.
• Sparse OFDM will dramatically increase the SNR.
  For a given transmission power, you will allocate
  more power per active sub-carrier with sparse
  OFDM than with standard OFDM.
• Sparse OFDM is on average less prone to ICI than
  standard OFDM.
• It is possible that when you factor in these
  benefits of sparse OFDM that it will outperform
  standard OFDM using QAM or even higher order
  constellations.