Fundamentals of statistical signal processing: estimation theory
Fundamentals of statistical signal processing: estimation theory
Hi-index | 0.00 |
In this paper channel estimation in the context of OFDM systems is investigated. In particular, a technique to improve the channel estimation in the current packet, by averaging out the channel estimates performed during previous packets, is proposed. However, when frame synchronization offset and/or sampling phase offset is present, the estimated channels differ by linear phase terms which have to be estimated and compensated. Therefore, to estimate the phase differences, a maximum likelihood estimator is suggested and a simplification based on the Taylor series, which noticeably reduces the computational load, is proposed. The performance of the algorithm is tested in the OFDM-based HomePlug AV system using the OPERA power-line channel models. Numerical results, in terms of BER versus signal-to-noise ratio, show that the simplified estimator based on the Taylor series achieves almost the same performance of the maximum likelihood estimator.