A low-complexity semi-blind joint channel estimation and data detection scheme is proposed for space-time shift keying (STSK) based multiple-input multiple-output systems. The minimum number of STSK training blocks, which is related to the number of transmitter antennas, is first utilized to provide a rough initial least square channel estimate (LSCE). Then low-complexity single-stream maximum likelihood (ML) data detection is carried out based on the initial LSCE and the detected data are employed to refine the decision-directed LSCE. It is demonstrated that a few iterations are sufficient to approach the optimal ML detection performance obtained with the perfect channel state information.
Published in:
Signal Processing Letters, IEEE
(Volume:17
,
Issue:
12
)
Date of Publication: Dec. 2010