Channel Estimation Algorithm Based on 2-D DFT with Multiple Pilot Patterns in OFDM Systems

葛启宏,陆建华,梅顺良
DOI: https://doi.org/10.3321/j.issn:1000-0054.2004.06.027
2004-01-01
Abstract:A noise-resilient channel estimation algorithm was developed for orthogonal frequency division multiplexing (OFDM) systems for comb, block, grid, and diamond pilot patterns. The effect of channel noise was mitigated with pre-correction of the pilot slippage, two-dimensional discrete Fourier transforms and filtering in the transformed domain. Zero padding and two-dimensional inverse discrete Fourier transforms were used to interpolate the channel transfer function, with a modified cosine window used to reduce the spectral leakage due to the interpolation. The simulation results show that the system performance with the proposed algorithm is significantly improved compared with the traditional Least Squares estimate, having a 2 dB better signal-to-noise ratio per bit. The algorithm can be used to improve wide-band OFDM systems.
What problem does this paper attempt to address?