A Distributed Optimization Approach for the Adaptation of Underwater Acoustic Communication Protocols

We propose a novel multi-agent approach for auto-adjusting OFDM parameters for underwater acoustic communication that utilizes distributed optimization to perform a collaborative choice. The algorithm enhances overall communication performance among all agents, and makes its decision based on environmental information that is first actively collected from each agent at the beginning of their mission, and then communicated via opportune statistics of such sampled information. The proposed method does not rely on link feedback from receivers; while based on distributed optimization (and thus requiring data transmission among the agents), the approach does not introduce any overhead during data transmission and can be used as a separate process at any preferred moment prior to the data transmission. We present numerical comparisons based on simulation results to demonstrate the dependence of the effectiveness of the proposed approach with respect to different marine conditions that may be encountered in field missions, and the dependence of its efficiency on which optimization algorithm is chosen. The overall results indicate that for a various set of conditions the approach may lead to a more effective usage of the underwater acoustic channel.