Statistical QoS Provisioning for Underwater Magnetic Induction Communication

Zhichao Li, Jianyu Wang, Wenchi Cheng, Yudong FangarXiv 2025 · DOI · arXiv

Introduction

Underwater communication is a critical frontier for ocean exploration and environmental monitoring. Traditional methods like acoustic waves suffer from low bandwidth and high latency, while electromagnetic (EM) waves are rapidly attenuated in seawater. Magnetic Induction (MI) communication, which utilizes the mutual inductance between coils, offers a promising alternative due to its stability and immunity to water turbulence. However, the narrowband nature of MI leads to significant queuing delays. This paper introduces a statistical Quality of Service (QoS) framework designed to maximize data throughput while strictly adhering to delay and queue-length constraints.

System Model and Problem Formulation

The proposed system considers an underwater MI link where a transmitter (e.g., on a surface vessel) communicates with underwater receivers. A key challenge in this environment is stochastic misalignment; water currents cause random changes in the orientation of the coils, affecting the mutual inductance.

Figure 1: System architecture of the underwater MI communication link, illustrating the protocol layers and the magnetic induction channel.

To address timing requirements, the authors employ Effective Capacity Theory. Unlike Shannon capacity, which focuses on maximum theoretical speed, effective capacity models the maximum arrival rate a channel can support while meeting a specific delay-bound requirement, characterized by a QoS exponent ($\theta$). The goal is to optimize the transmitter's current to maximize this capacity under average power and peak current hardware constraints.

The Optimal Current Control Strategy

Using Karush-Kuhn-Tucker (KKT) conditions, the researchers derived a closed-form expression for the optimal current control strategy. A significant finding is the departure from the traditional "Water-Filling" (WF) approach. While standard WF allocates more power to the best channel states, the proposed Water-Filling with Current Constraint (WFWC) strategy intelligently limits current when the channel is exceptionally good. This "restraint" saves power for weaker states and prevents hardware saturation, ultimately stabilizing the link.

The QoS exponent $\theta$ serves as a critical parameter: as $\theta$ increases, the system prioritizes timely delivery over raw throughput, adapting the power allocation to ensure packets do not exceed delay limits.

Key Results and Performance

The effectiveness of the optimal policy was evaluated through rigorous simulations. The results demonstrate that the proposed strategy significantly outperforms traditional methods like Constant Power, Constant Current, and Channel Inversion.

Figure 4: Comparison of Normalized Effective Capacity across different transmission policies as a function of the QoS exponent θ.

As shown in Figure 4, the "Optimal Policy" consistently maintains the highest effective capacity across a wide range of delay requirements. Notably, the proposed method reduces the probability of a communication outage by approximately 11% compared to standard Water-Filling. The model also proves adaptable to varying transmission distances, maintaining its mathematical robustness even as path loss increases.

Conclusion and Future Directions

This research provides a mathematical roadmap for reliable, time-sensitive underwater MI communication. By balancing throughput with strict delay constraints, the proposed framework enables more robust underwater sensor networks. Future work will look to expand this model to multi-hop mesh networks and incorporate dynamic "drifting" models where node distances change rapidly. Additionally, extending this framework to Multiple-Input Multiple-Output (MIMO) MI systems holds great potential for further increasing underwater data rates.

Summary updated

BibTeX
@misc{li2025statistical,
  author = {Zhichao Li and Jianyu Wang and Wenchi Cheng and Yudong Fang},
  title = {{Statistical QoS Provisioning for Underwater Magnetic Induction Communication}},
  year = {2025},
  eprint = {2502.13386},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2502.13386},
}
Figure