Time-Varying Channel and Intrablock Carrier Frequency Offset Estimation for OFDM Underwater Acoustic Communication

Orthogonal frequency division multiplexing (OFDM) is a prevalent scheme for medium-range high-speed underwater acoustic (UWA) communication; therefore, carrier frequency offset (CFO) estimation and compensation have been widely researched to maintain the orthogonality among subcarriers. However, the CFO is assumed to be a constant, or it changes only among the OFDM blocks for convenience. Few studies have focused on the intrablock CFO, which is a practical case. In this study, a slowly time-varying channel and intrablock CFO sequential estimation for OFDM is proposed based on the relationship between the channel impulse response (CIR) and the CFO. First, the relationship was introduced based on block-type pilot signals filled with chirp signals. Subsequently, the fractional Fourier transform (FrFT) of the chirp signals is adopted to achieve coarse channel estimation (CE). Consequently, a differential evolution (DE) algorithm is applied to refine it iteratively. Therefore, the CFO on the pilot blocks (PBs) is extracted with accurate CE. Finally, the CFO of each subcarrier within the data blocks (DBs) is obtained by performing interpolation in the time domain. The proposed time-varying channel and intrablock CFO estimations are verified by conducting simulation experiments based on the measured environmental parameters of Fuxian Lake. The accuracy of the CE of the proposed method is significantly higher than those of the other compared methods. The accuracies of the delay and amplitude estimates at 20 dB are 100% and 96.31%, respectively. With the presence of the intrablock CFO, the bit error rate (BER) after compensation is $3.20 \times 10<^>{-3}$ at 30 dB.