2023_programme: Comparison of Channel Codes based on Optimal Channel Replay
- Session: 19. Underwater Communications and Networking
Organiser(s): Charalampos Tsimenidis, Paul Mitchell and Konstantinos Pelekanakis
- Lecture: Comparison of Channel Codes based on Optimal Channel Replay
Paper ID: 2091
Author(s): Pelekanakis Konstantinos
Presenter: Pelekanakis Konstantinos
Abstract: The low reliability and throughput of underwater acoustic (UWA) communications requires bit rate adaptation at the physical layer based on in-situ channel conditions. Therefore, it is important to rank the performance of Forward Error Correction (FEC) codes for various channel conditions. Such studies are typically conducted in simulation, however, the lack of UWA channel model standards makes it difficult to compare candidate codes coming from different researchers under the same simulation environments. Furthermore, numerical simulators tend to produce overly optimistic results as complex acoustic phenomena are not fully captured. To achieve a realistic comparison under the same channels, i.e., multipath and noise, the authors extend the dither constellation technique for the Decision Feedback Equalizer (DFE) to include arbitrary-rate FEC codes combined with M-ary Phase-shift keying (PSK) modulation. Audio signals based on single-carrier, 2/4/8-PSK modulation occupying the 3830-6170~Hz band were collected in a linear hydrophone array during the Nordic Recognised Environmental Picture 2021 (NREP’21) sea expedition off Svalbard Islands. A unique characteristic of the environment was the moving sea surface ice floes between the transmitter and the array generating complex multipath conditions. Six FEC codes combined with dithered 2-PSK constellations, including convolutional, turbo, Low-Density Parity Check (LDPC), Bose–Chaudhuri–Hocquenghem (BCH), Turbo Product Code (TPC), and polar, are compared by reusing the recorded PSK signals in post-processing. These six codes are applied to a 128-bit packet and a 1024-bit packet at code rates of 0.2 and 0.5, respectively. The results show that polar codes yield the lowest Packet Error Rate (PER) performance for both packet sizes and a varying number of input hydrophones at the DFE. In addition, the PER superiority of the polar codes is demonstrated when a turbo equalization principle is applied. In contrast, the code with the lowest Bit Error Rate (BER) depends on the code rate and the number of input hydrophones.
- Corresponding author: Dr Konstantinos Pelekanakis
Affiliation: NATO STO CMRE
Country: Italy
e-mail:
