Journal of Systems Engineering and Electronics ›› 2010, Vol. 21 ›› Issue (1): 148-153.doi: 10.3969/j.issn.1004-4132.2010.01.024

• SOFTWARE ALGORITHM AND SIMULATION • Previous Articles     Next Articles

Bayesian sequential state estimation for MIMO-OFDM systems

Maoge Xu1,∗ and Yaoliang Song2   

  1. 1. Southwest Electronic Technology Research Institute, Chengdu 610036, P. R. China;
    2. School of Electronic Engineering and Optoelectronic Technology, Nanjing University of Science and Technology, Nanjing 210094, P. R. China
  • Online:2010-02-26 Published:2010-01-03


For the estimation of MIMO frequency selective channel, to mitigate the curse of dimensionality, a novel particle filtering scheme combined with time delay domain processing is proposed. In order to extract the time delay domain channel impulse response from the observed signal, the least-squares (LS) and minimum mean squared error (MMSE) criteria are discussed and the comparable performance of LS with MMSE for samplespaced channel is revealed. Incorporated the dynamical channel model, gradient particle filtering is further introduced to improve the estimation performance. The robustness of the channel estimator for underestimated Doppler frequency and the effectiveness of the new estimation scheme are illustrated through simulation at last.