Efficient Long‐Distance Underwater Wireless Communication Using CP‐GS‐KF and SATR‐ANN Techniques in a UWA‐OFDM System

ABSTRACT For long‐distance wireless communication in underwater environments, Underwater Acoustic (UWA) communication serves as a vital solution. However, UWA communication faces challenges like lower data rates and the time‐varying nature of the UWA Channel, including the presence of the Doppler Effect. To overcome these limitations, this paper proposed an Underwater Acoustic‐Orthogonal Frequency Division Multiplexing (UWA‐OFDM) communication system employing a Cosine Probability Distributed Golden Search Optimized Covariance State Induced Kalman Filter (CP‐GS‐KF) and Semi‐Averaged Trend Removed Artificial Neural Network (SATR‐ANN) approach. Here, the data bits undergo convolution coding and interleaving at the transmitter before being modulated using Henkel Function First Kind‐employed Quadrature Phase Shift Keying (HFFK‐QPSK). Cyclic Prefix (CP) addition and upsampling are applied to ensure reliable transmission. The pulse shape is accurately recognized using a Perfectly Positive Auto Correlated‐Raised Cosine Filter (PPA‐RCF), followed by efficient upconversion. The baseband signal is transformed into a digital representation at the receiver end and synchronized for further processing. For decreasing the effects of the Doppler effect and estimating the channel, the CP‐GS‐KF and SATR‐ANN techniques are employed, respectively. Several essential steps, namely downconversion, downsampling, CP elimination, frequency domain conversion, and equalization, are then executed. Lastly, the received signals are demodulated using HFFK‐QPSK and decoded to recover the original data. By achieving a significantly reduced Bit Error Rate (BER) of 0.1986, Symbol Error Rate (SER) of 0.0203 during channel estimation using SATR‐ANN, and removing the Doppler effect using CP‐GS‐KF with a Mean Square Error (MSE) of 0.12% that is lower than existing EKF, the proposed UWA‐OFDM system demonstrates its superiority through extensive simulations and comparisons with prevailing approaches.