
Around 70 percent of the earth's surface is covered by water. However, very little is known about what lies below sea level. Underwater sensors, gliders and autonomous submersible vehicles can help us learn more about the oceans by recording environmental data such as salinity, currents and temperature, or by mapping the seabed. The potential areas of applications are vast. But how does the data make its way to the water's surface? Fraunhofer FKIE has been working to find answers to this question for over ten years.
The challenge posed by underwater communication is the water itself. Radio and light waves are heavily absorbed by the water. They disappear at a maximum depth of 60 meters. So how can data still be transmitted in the dark depths of the oceans? Nature itself has the answer: Whales and other marine mammals communicate via sound. The deeper the sound, the farther it travels, meaning that it can be heard hundreds of kilometers away. Sound not only achieves greater ranges underwater, it also travels around five times faster than in the air.
Nevertheless, there are some disadvantages. Various noises such as passing ships, rain or waves on the surface of the water can disrupt communication. However, the greatest challenge lies in the low data rates that can be achieved with waterborne sound. While it is now possible to transmit several gigabits per second in the radio range, only a few kilobits per second can be transmitted underwater. The greater the desired range, the lower the selected frequency and the associated bandwidth must be. A frequency band of four to eight kilohertz with a net data rate of less than 1,000 bits per second is suitable for a range of ten kilometers.
The aim, however, is not only to connect individual devices. The aim is to create an entire network, comparable to the internet. To achieve this, Fraunhofer FKIE is working with partners to develop completely new concepts, as standard IP networks are not suited to narrow-band channels. This raises questions such as what information is absolutely essential, and how can data be pre-processed and optimally compressed?
These questions should not only be considered from a theoretical standpoint. Fraunhofer FKIE also implements the solutions on real hardware and tests them out during practical sea trials. Only then can the concepts be proven to work in practice. The goal is to develop a prototype that has been evaluated and is fully functional under real operating conditions.
First of all, the researchers examined which existing technologies are suited to the underwater communication environment. These are typically adapted to the specific requirements. However, it soon became apparent that researchers would need to take completely new approaches in the case of extremely low data rates and high latencies. To this end, they carried out a requirements analysis: Who wants to communicate what, when and who to? Fraunhofer FKIE has investigated this question as part of a European consortium of navy, industry and research institutes and, together with WTD 71, has specified an application language under the name “Generic Underwater Application Language” (GUWAL). GUWAL thus forms the basis of a common standard that devices can use to exchange data underwater. GUWAL defines the protocol for requesting and transmitting sensor data as well as exchanging commands with autonomous diving robots. Data compression efficiency plays a key role here in order to meet the requirements of the narrow-band underwater channel. As a result, GUWAL only requires a packet length of 128 bits.
The researchers have also developed a network protocol based on GUWAL. It transmits the short message packets between the participants while they are under water. As the range is linked to the data rate, not all network participants can directly exchange data with each other. In large networks that extend over a hundred kilometers or more, the data must be forwarded from participant to participant in the direction of the destination. This process is known as “multi hopping” and is a key technology used in the underwater environment. However, this also requires that routes to the destination be found automatically and adapted to the constantly changing conditions. Fraunhofer FKIE has specifically tailored the GUWMANET network protocol to these requirements.
Before processes can be tested under real conditions during sea trials using complex and cost-intensive methods, they must first be simulated and emulated in the laboratory. Fraunhofer FKIE has collaborated with research institutes and universities to develop a simulation environment. In this environment, GUWMANET was tested in a large number of simulations before it was subsequently trialed at sea.
As part of the "Robust Acoustic Communications in Underwater Networks" (RACUN) project conducted by the European Defense Agency (EDA), the five participating nations of Germany, the Netherlands, Italy, Norway and Sweden developed various underwater protocols and tested them against each other between 2010 and 2014. Submersible vehicles from ATLAS Elektronik and modems from the Hamburg-based company develogic were also deployed. The GUWMANET network protocol developed by Fraunhofer FKIE and WTD 71 was used as a point of reference for evaluating the other processes. GUWMANET delivered the best results by far in both the simulations and the final sea trial, with packet transmission rates of around 90 percent. For this reason, it was chosen as the standard for the follow-up project “Smart Adaptive Long- and Short-range Acoustic network” (SALSA) and is currently in further development together with the European partners.
The SALSA project has developed new strategies for enabling the network protocol to switch automatically between different profiles. Some profiles are particularly robust and suited to difficult conditions with high noise levels, while others are less robust but deliver higher data rates. The processes were already trialed during a port test in Kiel in 2020 and in Lake Vättern in Sweden in 2021 and were also used in NATO exercises REP(MUS) 21, 22, 23 and 24. The final trial took place in the Oslofjord in Norway in April 2022. GUWMANET was also operated for the first time on the UT3000 underwater telephone from ELAC Sonar, which is used, among other things, for submarine communication. This made communication with the autonomous submersible vehicles and the sensor nodes on the seabed possible. The network provided regular position and status updates and enabled the operator to intervene if necessary. As part of the current EDF project “SWArm and Teaming operation of manned & unmanned underwater vehicle SHOAL” (SWAT-SHOAL), this technology is undergoing further development in order to command entire swarms of unmanned underwater drones. An international underwater network of this kind, with project partners from the Netherlands, Finland, Norway, Sweden, Germany, Spain, Portugal, Poland, Estonia, Italy, Belgium and Greece, could only be established through the implementation of common standards such as GUWMANET and GUWAL.
In addition to GUWAL and GUWMANET, Fraunhofer FKIE has developed a prototype gateway buoy with the WTD 71. It serves as an intersection between the surface and underwater worlds and makes it possible to connect underwater devices to existing terrestrial networks. The radio technologies used include 4G and 5G networks with ranges of around ten kilometers as well as HF radio for ranges of over 65 kilometers. The gateway buoy was also equipped with technology that enables transmission via the LoRa radio standard. The prototype was deployed in the large-scale NATO exercises REP(MUS) near Lisbon, Portugal, between the years 2021 and 2024. The buoy relayed the underwater packages from GUWAL to a maritime operations center on land. Going forward, other projects of interest will be those in which the buoy can act as an amplifier of a 5G network or in which the buoy can be integrated as an end device in a 5G NTN (Non Terrestrial Networks) infrastructure.
International partners in the SALSA project