DX-Ocean collects and analyzes the status and quality of LTE/5G mobile and VSAT satellite links and manages them together with voyage data.
Mobile networks play a larger role in coastal waters and satellite links further offshore, and communication quality also varies with weather, route and vessel position.
DX-Ocean collects the status of every communication channel in one place and builds up a quality history, providing a basis for analyzing and validating path selection policies.

In the Korea Autonomous Surface Ship (KASS) project of the Ministry of Oceans and Fisheries and KIMST, the DX-Ocean multi-network maritime gateway was installed on the autonomous test ship. On the long route sailed by the autonomous test ship between Korea (Incheon and Busan), China (Shanghai), Vietnam and Thailand (July 2024 to December 2025), it ran commercial mobile and VSAT satellite networks together, proving its performance on a real route across coastal and open waters.
Rack layout (top to bottom)The Multi-RAT multi-network maritime gateway collects the status of each link, including LTE/5G and VSAT, and serves as the hub that connects onboard systems with shore systems. Beyond simply linking different networks, it accumulates per-link quality data to support decisions on path selection and communication policies.

Communication status and voyage information collected on board are first preserved on the vessel and synchronized again once the central connection is restored. The system is configured so that communication status and data history can be reviewed through the same operating workflow regardless of equipment differences.
The maritime broadband terminal continuously checks the connection status and signal quality of the LTE modem and reports them to the gateway.
It structures key indicators such as RSRP, RSRQ, RSSI and SINR, together with serving cell and network information, into data that can be used to assess and analyze communication quality.
It supports network configuration so that onboard equipment can use the LTE link as an external communication path, and it provides status information in a form other systems can use, allowing integration with the gateway and management functions.

These are research results from LTE and VSAT quality data actually measured over 12 voyages of a container ship, analyzed by time and voyage segment. They reveal coverage gaps, degradation patterns and link characteristics, and are used to validate path selection policies and switchover criteria.
Measured for nine months on routes between Incheon, Busan, Shanghai, Ho Chi Minh and Laem Chabang

Cellular and satellite terminals were connected to the multi-network gateway to log both links together

Records meeting the received-power criterion ranged from 77.1% (Incheon → Busan) to 17.3% (Shanghai → Ho Chi Minh), about a 4.5-fold gap
Criterion: LTE RSRP ≥ −100 dBm or WCDMA RSCP ≥ −85 dBm. Based on 470,920 valid cellular-state records while sailing.
26.5% of responsive records fell between 6 and 7 dB, so raising the threshold from 6 to 7 dB increased the store-and-forward share from 6.7% to 25.7%
Cellular assignment stays at 28.8% at every threshold
Records go to cellular when it meets the received-power criterion, otherwise to VSAT when it responds with SNR at or above the threshold, and otherwise to store-and-forward. Retrospective analysis of 319,363 joint-analysis records; no live traffic was redirected.
※ The map and setup diagram above are taken from the reference at the bottom of the page, and the two charts adapt results from the same paper
Communication status collection, history management, onboard preservation and synchronization are implemented. Path selection that considers latency, loss, available capacity and service importance is a technology under research and validation and is not presented as a completed commercial automatic switchover function.
Koo, H., Ryoo, C., & Song, J. (2026). Repeated-voyage measurement of cellular–GEO satellite complementarity and buffering implications for maritime IoT backhaul. Applied Sciences, 16(18), 9361. https://doi.org/10.3390/app16189361 (CC BY 4.0)