Comparison of error correction codes via optimal channel replay of high north underwater acoustic channels

This paper aims to enhance the reliability of underwater acoustic (UWA) communications by evaluating the effectiveness of various Forward Error Correction (FEC) codes. In the absence of standardized UWA physical layer simulators, we rely on in-situ data to assess the performance of six FEC codes: convolutional, turbo, Low-Density Parity Check (LDPC), Bose– Chaudhuri–Hocquenghem (BCH), Turbo Product Code (TPC), and polar. We examine two common payload sizes: 128-bit packets at a 1/5 encoding rate and 1024-bit packets at a 1/2 rate. Our comparative analysis leverages previously recorded single-carrier Phase-Shift Keying (PSK) signals, originally not intended for FEC evaluation. The signals were transmitted over various underwater links in a High North environment, including areas of sea surface ice floes, northwest of the Svalbard Islands. Using these signals, we formulate an optimal channel replay framework for FEC comparison through constellation dithering for M-ary PSK (M-PSK) modulation. This framework involves a Channel Estimate-Based-Decision Feedback Equalizer (CEB-DFE) with a varying number of input channels (hydrophones) at the receiver end. We focus on the Packet Error Rate (PER), maintaining consistent channel conditions for all codes under consideration. The polar code demonstrates the best performance by achieving the lowest PER for all tested packet lengths and varying number of hydrophones. Moreover, upgrading the equalizer with a turbo feature enables the polar code to achieve up to a 45% further reduction in PER in certain scenarios.