Channel Estimation Algorithms for Uplink OFDMA Tiles

OFDM is known to be robust to wireless multipath channels since interferences are alleviated by the frequency-domain signal construction and the use of a cyclic prefix. However, it is still necessary to estimate the channel at each subcarrier position and to equalize the flat fading distortion. In the context of orthogonal frequency division multiple access (OFDMA) where uplink data from multiple users are grouped in a time and frequency structure, the use of conventional channel estimation algorithms based on statistic acquisition is not practical. In this paper, four channel estimation algorithms designed for the high diversity uplink OFDMA tile structure are compared. Simulations results for an averaging algorithm, an interpolation algorithm, a regression algorithm and of a novel method combining both previously mentioned with decision guiding are presented for channel models similar to SUI. The novel two-steps algorithm outperforms other algorithms in terms of bit-error rate. This is observed for a range of coding and modulation configurations and varying the multipath channel conditions