Back in 2009, I wrote about advances in underwater wireless communications and the challenges involved in connecting underwater sensors and autonomous vehicles. At the time, much of the discussion centred on underwater acoustic sensor networks, where limited bandwidth, long propagation delays, multipath and high error rates created challenges very different from those encountered in terrestrial wireless networks. Nearly two decades later, many of those fundamental challenges remain.
Earlier this year, I looked again at underwater communications and the challenge of ocean connectivity, including the growing interest in heterogeneous networks combining acoustic, optical and, where practical, radio technologies. There is unlikely to be a single underwater equivalent of Wi-Fi or cellular connectivity. Instead, different technologies occupy different parts of the range, data-rate, latency and power-consumption trade-off.
A new research project from the University of Florida adds another interesting option to this mix. BlueME is a compact magnetoelectric antenna system designed for communication between underwater robots and sensors. Researchers Mehron Talebi, Sultan Mahmud, Adam Khalifa and Md Jahidul Islam have demonstrated the system in both freshwater and saltwater, including an ocean experiment in which signals were detected at distances of more than 700 metres using around 10 watts of transmitter power.
Conventional radio becomes increasingly difficult to use underwater because seawater is electrically conductive. Electromagnetic waves are absorbed rapidly, with attenuation increasing substantially as frequency rises.
Moving to much lower frequencies helps with propagation, but creates another problem. Conventional antennas generally become extremely large relative to practical underwater vehicles and sensor nodes as the wavelength increases.
This is one reason underwater communications has traditionally relied heavily on acoustics.
Sound can travel kilometres through water and, in some circumstances, much further. The downside is that sound travels at only around 1,500 metres per second rather than anything approaching the propagation speed of electromagnetic waves. Acoustic channels also suffer from environmental noise, multipath and Doppler effects, while usable bandwidth can be extremely limited. These issues were already central to underwater networking research when I first wrote about the subject in 2009.
Optical communications occupy another part of the trade-off. Blue-green optical systems can potentially provide very high data rates with low latency, but range is considerably shorter and performance can be affected by turbidity, scattering and alignment.
Future underwater networks are therefore likely to be multimodal, using different technologies depending on the application.
BlueME takes a different approach to creating low-frequency electromagnetic signals.
Instead of relying on a conventional electrically resonant antenna, magnetoelectric antennas exploit mechanical resonance and the interaction between piezoelectric and magnetostrictive materials. This can allow very low-frequency electromagnetic signals to be generated using an antenna structure that is dramatically smaller than a conventional antenna designed for the same frequency.
The BlueME arrays have a primary resonance around 35 to 36 kHz. The researchers refer more generally to VLF/LF magnetoelectric communications, although frequencies above 30 kHz technically fall within the LF rather than VLF band.
For the experiments, multiple magnetoelectric elements were combined into transmitter and receiver arrays and installed in underwater housings suitable for integration with robotic platforms. The researchers mounted the equipment on a BlueBoat autonomous surface vehicle and a BlueROV2 remotely operated vehicle.
Freshwater testing produced detectable signals at the maximum available test distance of 200 metres while consuming around 1 watt. The team then moved to ocean tests off Florida's Gulf Coast.
There, the transmitter was suspended around three metres below the surface while the receiver operated in water approximately 8 to 11 metres deep. Horizontal separation was increased to a maximum of approximately 730 metres, with the system operating at around 10 watts.
Those figures are impressive for such compact low-frequency equipment, but they need some qualification.
The 730-metre result should not be interpreted as demonstrating a 730-metre commercial underwater data modem.
During the range trials, the transmitter performed frequency sweeps, broadly between 31 and 41 kHz, while the receiver measured the resulting signal and signal-to-noise ratio. The researchers subsequently analysed how modulation schemes such as binary frequency-shift keying, or BFSK, might perform using the measured channel data.
In other words, the propagation link has been demonstrated, but practical digital communications at useful payload data rates still require further development.
This distinction is particularly important because magnetoelectric antennas are highly resonant devices. Their available bandwidth is limited and their response becomes nonlinear at higher drive levels. The researchers suggest constant-envelope modulation techniques such as BFSK, MSK or CPFSK as possible approaches and identify tailored modulation schemes as an area for future work.
There is another interesting uncertainty in the range measurements. At longer distances, the received signal did not decay as rapidly as a simple underwater propagation model might suggest. The researchers believe some energy may have propagated along the water-air or water-seabed interfaces.
That could potentially be useful, but it also means the 730-metre shallow-water result cannot simply be extrapolated to a robot operating hundreds or thousands of metres below the surface. Deep-water experiments will be particularly important.
Where BlueME becomes especially interesting is not as a replacement for acoustics or optical communications, but as another possible layer within a heterogeneous underwater network.
A future architecture could connect underwater sensors, autonomous underwater vehicles and remotely operated vehicles through acoustic, optical or magnetoelectric wireless links to a subsea gateway. From there, connectivity could pass through a buoy, surface vessel or offshore platform before continuing over private 4G or 5G, microwave, satellite/NTN or fibre to an edge platform, cloud environment or operations centre.
For AUV (autonomous underwater vehicle) fleets, magnetoelectric links might eventually provide relatively low-latency coordination, status messages, commands or telemetry without requiring vehicles to surface.
Submarine cable operators could potentially use the technology for communicating with inspection robots. Offshore wind farms and oil and gas installations could connect underwater sensors and maintenance systems to gateways on turbines, platforms or substations. Similar architectures could support ports, aquaculture and environmental monitoring.
Subsea data centres are another possible, although more speculative, use case. Their primary connectivity would still be provided by fibre, but magnetoelectric links might support surrounding environmental sensors, inspection robots or maintenance systems.
This also connects with another technology I looked at last year in a post on pushing the boundaries of wireless sensing technologies: underwater acoustic backscatter. Researchers are investigating ways of allowing battery-constrained ocean sensors to communicate by reflecting acoustic signals rather than actively generating them, dramatically reducing their energy requirements.
The wider trend is therefore not towards one dominant underwater radio technology. It is towards an increasingly diverse Internet of Underwater Things, where acoustic, optical, backscatter and electromagnetic systems may each be used where their particular characteristics make sense.
BlueME is still at an early stage. The University of Florida team says it has a complete prototype and has filed a provisional patent, but further funding and development are required. Manufacturing consistency between antenna elements is one issue identified in the research, alongside modulation, deeper-water testing and integration with autonomous systems.
There are also questions beyond the physical layer.
A practical network would need addressing, medium-access control, interference management and mechanisms for scaling beyond a small number of devices. Spectrum use around 35 kHz would need to consider national allocations and possible emissions beyond the underwater environment. Interoperability and standardisation would become increasingly important if magnetoelectric communications were to move beyond proprietary point-to-point systems.
Security would need attention as well. An underwater wireless link connecting critical offshore infrastructure would ultimately require authentication, encryption and protection against deliberate interference just as terrestrial infrastructure does.
None of this diminishes the significance of the BlueME experiments. It simply places them at the right stage of development.
The ocean is therefore not quite the wireless "dead zone" that some headlines suggest. Underwater wireless communications have existed for decades.
The real challenge is that there is still no universal underwater access technology comparable with the radio networks we take for granted on land.
Magnetoelectric antennas probably will not provide that universal solution either. But BlueME suggests they could eventually become another useful piece of the underwater connectivity puzzle.
Comments
Post a Comment