Answers are grounded in this paper and cite their sources.
Evaluates improved performance of adaptive wavelength-beam control in underwater networks, highlighting the method's potential to manage interference.
UOWC provides high data rate transfers, but due to multiple users being in close proximity, performance suffers from beam overlap, crosstalk induced by dispersion, turbulence and congestion at access points. The authors propose a proactive method of managing interference as dynamic graphs and controlling interference through joint optimization of wavelength, beam divergence, transmission power and access point association. The authors name their proposed method Physics-Informed Graph Attention with Adaptive Beam and Color control (PIGAT-ABC). The authors incorporate channel characteristics into a graph-attention model which predicts interference patterns and will provide interference information to an optimization process that spans layers (cross-layer). The authors use a physics-based loss function to limit predicted signal-to-interference-plus-noise ratios (SINR) within realistic limits for the three types of water (clear, coastal, turbid) found in oceanic and coastal systems. The authors create a reproducible Monte Carlo simulation model using an analytical model develop to test the performance of their proposed solution. The authors show an improved performance over a baseline with fixed resources in a dense network (8 users with 20 dB SNR) whereby the proposed system outperformed the benchmark by 23.1 % reduction in Bit Error Rate (BER) and 31.3 % increase in spectral efficiency. The authors also demonstrate improved fairness and energy efficiency, and in support of a network centric method of avoiding interference.
No takes yet. Share an insight, caveat, or question.
Srivastava et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: