A Software-Defined Opto-Acoustic Network Architecture for Internet of Underwater Things

Abdulkadir Celik, Nasir Saeed, Basem Shihada, Tareq Y. Al-Naffouri, Mohamed-Slim Alouini

This paper, published in late 2019, outlines a vision for the future of underwater connectivity. Here is a summary of the proposed architecture and its implications for the "Internet of Underwater Things."

TL;DR

This paper proposes a hybrid "opto-acoustic" network architecture for the Internet of Underwater Things (IoUT). By integrating Software-Defined Networking (SDN) and Network Function Virtualization (NFV), the authors provide a framework to combine the long-range reliability of acoustic waves with the high-speed performance of optical wireless communication, creating a flexible and programmable underwater internet.

What Problem Does This Paper Address?

The "aquatic medium" is notoriously hostile to data. Terrestrial IoT relies on Radio Frequency (RF) waves, but in water, RF is absorbed almost immediately, limiting range to a few meters.

Existing underwater solutions generally fall into two camps, both with major drawbacks:

  1. Acoustic Communication (UAC): Can travel kilometers but is painfully slow (Kbps) and suffers from high latency due to the slow speed of sound in water.
  2. Optical Wireless Communication (OWC): Offers blazing speeds (Gbps) and low latency but is limited to short ranges (50-100m) and requires precise "line-of-sight" alignment between nodes.

Furthermore, most current underwater networks are "ossified"—they are hard-coded for a single purpose (like monitoring a specific oil rig). They cannot easily adapt to new tasks or changing water conditions (like turbidity or currents).

Methods

The authors propose a Software-Defined Opto-Acoustic Network (UOAN). The core innovation is the application of two "cloud-native" concepts to the ocean floor:

  • Software-Defined Underwater Networking (SDUN): This decouples the "brains" (control plane) from the "brawn" (data plane). Instead of each sensor deciding how to route data, a centralized controller (potentially housed in an underwater data center) manages the whole network.
  • Network Function Virtualization (NFV): This allows network tasks (like firewalls or specific routing protocols) to run as software "slices." This means one physical network can be "sliced" to serve different applications—one slice for low-power environmental sensing and another for high-speed video from an Autonomous Underwater Vehicle (AUV).

The architecture is organized into four layers:

  1. Perception Layer: The physical sensors and AUVs.
  2. Network Layer: The links (acoustic and optical) and the hardware like Optical Base Stations (OBS).
  3. Middleware Layer: The "translator" that masks the complexity of the hardware from the applications.
  4. Application Layer: Where the actual data processing for disaster prevention or offshore exploration happens.

Datasets

As a conceptual architecture and vision paper, it does not use a standard training dataset. Instead, the authors evaluate their ideas using:

  • Mathematical Modeling: Utilizing the Beer-Lambert law to model underwater light propagation.
  • Simulations: A simulated environment of 50 nodes randomly distributed over a 500 $m^3$ volume to test connectivity and end-to-end (E2E) data rates.

Key Insights and Ideas

  • Complementary Hybridization: The authors argue that we shouldn't choose between acoustic and optical. Instead, use acoustic for "control traffic" (slow but reliable) and optical for "data traffic" (fast but finicky).
  • The "Pointing" Solution: To solve the alignment problem of lasers, the authors suggest using geodesic polyhedral nodes—multi-faceted spheres with transceivers on every face to provide omnidirectional optical coverage.
  • Underwater Data Centers (UDC): The paper highlights projects like Microsoft’s Project Natick, suggesting that sinking data centers into the ocean provides the necessary "centralized brains" and cooling for SDUNs.
  • Cross-Layer Orchestration: By using SDN, the network can sense a change in water turbidity and automatically tell nodes to switch from optical to acoustic or adjust their transmission power.

Key Results

The simulation results (visualized in Figure 3 of the paper) demonstrate the trade-offs of the hybrid approach:

  • Connectivity vs. Rate: Acoustic links provide nearly 100% connectivity in sparse networks but at very low rates. Optical links provide Gbps speeds but only when the "divergence angle" (the spread of the light beam) is perfectly tuned.
  • Localization Accuracy: Hybridization significantly improves the ability to track underwater objects. The Root Mean Square Error (RMSE) of localization is much lower in the "Hybrid-Pure" scenario compared to using acoustic or optical alone.
  • Resilience: The SDN/NFV approach allows for "Fault Recovery." If a node fails, the central controller can instantly reroute traffic through a different path or switch communication modes to maintain the link.

Limitations and Future Work

The authors acknowledge several "daunting challenges" that still need to be solved:

  • Energy Scarcity: Replacing batteries at the bottom of the ocean is nearly impossible. Future work needs to focus on energy harvesting (like microbial fuel cells or solar-powered AUVs).
  • Hardware Costs: Waterproofing and protecting equipment from corrosion and "bio-fouling" (sea life growing on sensors) remains expensive.
  • Machine Learning Integration: The authors suggest that ML will be essential for predicting the "turbulent and turbid" nature of water to optimize links in real-time.
  • Deployment Optimization: Deciding exactly where to place base stations and controllers to maximize coverage while minimizing cost is a complex mathematical problem that remains open.

Summary updated

Abstract

In this article, we envision a hybrid opto-acoustic network design for IoUT. The software defined underwater network (SDUN) is presented as an enabler of hybridizing benefits of optic and acoustic systems and adapting IoUT nodes to the challenging and dynamically changing underwater environment. We explain inextricably interwoven relations among functionalities of different SDUN layers and analyze their impacts on key network attributes. The NFV concept is then introduced to realize application-specific cross-layer protocol suites through an NFV management and orchestration system. We finally discuss how SDUN and NFV can slice available network resources as per the diverging service demands of different underwater applications. Such a revolutionary architectural paradigm shift is not only a cure for chronic underwater networking problems but also a way of smoothly integrating IoUT and IoT ecosystems.
BibTeX
@article{celik2020software,
  author = {Abdulkadir Celik and Nasir Saeed and Basem Shihada and Tareq Y. Al-Naffouri and Mohamed-Slim Alouini},
  title = {{A Software-Defined Opto-Acoustic Network Architecture for Internet of Underwater Things}},
  journal = {IEEE Communications Magazine},
  year = {2020},
  url = {https://openreview.net/forum?id=xi_dnJeVN0},
}

Loading related content...

Figure