In this work we evaluate the performance of multistage blind equalization scheme for time-varying MIMO systems. Equalization of the signals is achieved by minimizing a cost function, which is of a linear combination of the constant modulus algorithm (CMA) and the alphabet-matched algorithm (AMA). This has been previously shown to outperform the conventional CMA equalizer in recovery of non-constant modulus signals, such as 16-QAM. The transmitted sources in a MIMO system can be recovered in a multistage fashion using a cascade of CMA-based MISO equalizers. After equalization each captured source is cancelled by estimating its corresponding channel vector and subtracting the contribution of that captured source from the received signal. The modified received signal vector is then used as input to the equalizer at the next stage. A time-varying Gaussian channel model is used to evaluate the effect of time variation on multistage equalization and cancellation.
[1]
Buyurmaiz Baykal,et al.
Blind channel estimation via combining autocorrelation and blind phase estimation
,
2004,
IEEE Transactions on Circuits and Systems I: Regular Papers.
[2]
Michael L. Honig,et al.
Adaptive beamforming in interference networks via bi-directional training
,
2010,
2010 44th Annual Conference on Information Sciences and Systems (CISS).
[3]
A. Cole-Rhodes,et al.
An adaptive blind equalizer with signal separation for a MIMO system transmitting QAM signals
,
2008,
MILCOM 2008 - 2008 IEEE Military Communications Conference.
[4]
John J. Shynk,et al.
The constant modulus array for cochannel signal copy and direction finding
,
1996,
IEEE Trans. Signal Process..
[5]
L. Tong,et al.
Multichannel blind identification: from subspace to maximum likelihood methods
,
1998,
Proc. IEEE.