Exploiting subcarrier correlation for blind channel estimation in block preceded OFDM systems

This paper proposes a second-order statistics based blind channel estimation method for BP-OFDM systems. To enable the proposed method, a slightly modification needs to be made for the modulation scheme by exploiting the frequency diversity. It is established that the channel identifiability is guaranteed regardless of channel nulls location. The remaining sign ambiguity can be easily resolved using various blind or semi-blind algorithms. Compared with the existing subspace approach, the proposed method does not need high complexity eigen decomposition, and adopts an inverse fast Fourier transform to improve the computation efficiency. Simulation results have demonstrated the promising performance of the proposed method in static and slowly time varying channels.