A Beacon Based Solution for Autonomous UUVs GNSS-Denied Stealthy Navigation

Technical Summary: A Beacon Based Solution for Autonomous UUVs

Problem Statement
Autonomous Unmanned Underwater Vehicles (UUVs) are increasingly required for military and civilian covert operations in coastal areas where surfacing is prohibited or dangerous. In these scenarios, Global Navigation Satellite Systems (GNSS) are unavailable due to the rapid attenuation of high-frequency electromagnetic waves underwater. Consequently, UUVs cannot rely on standard satellite positioning, and surfacing to acquire a fix compromises stealth by exposing the vehicle to visual, radar, and other surveillance systems. Furthermore, the underwater environment presents dynamic challenges, including variable currents, limited communication bandwidth, and sensor limitations (e.g., the inability to use Doppler Velocity Loggers without bottom detection). Existing acoustic localization methods, such as Long Baseline (LBL) or Single Beacon systems, often face issues with deployment complexity, range limitations, or a lack of unique position solutions. The core challenge addressed is enabling a fleet of UUVs to navigate precisely from the continental shelf to a shore-based goal without GNSS or immediate surface support, while maintaining concealment.

Methodology
The proposed solution employs a two-stage approach: the deployment of a synthetic landmark network followed by hierarchical planning for fleet navigation.

1. Synthetic Landmark Network (Beacon Deployment):
   Instead of relying on pre-existing infrastructure, the system utilizes a constellation of acoustic beacons deployed by aerial or surface drones prior to the UUV mission. These beacons can be submerged, mid-water, or surface-floating. To optimize their placement, the authors implement a Constrained Lloyd's algorithm. This algorithm iteratively distributes $N$ beacons within a defined water volume ($V_{tot}$) to maximize information gathering. It minimizes the average absolute deviation between the volume of water where a beacon signal can be received and the volume where it cannot, effectively splitting the mission area into $N$ equal-volume Voronoi cells. The centroid of each cell determines the optimal beacon deployment position.

2. Hierarchical Task Network (HTN) Planning:
   Once beacons are deployed, they guide the UUV fleet. The navigation plan is synthesized using an HTN planner. Unlike classical planning which focuses on state transitions, HTN planning decomposes complex abstract tasks (e.g., "navigate to goal") into sequences of primitive actions (e.g., "navigate course," "sense beacon," "broadcast message").

   • Planning Model: The domain is modeled using the Hierarchical Task Network paradigm. The planner generates a sequence of actions to guide UUVs from beacon to beacon.
   • Execution and Replanning: The system operates in a closed-loop fashion. An execution monitor continuously gathers sensor data (acoustic signals, inertial navigation, etc.) to update the fleet's state. If the current state deviates from the expected state (e.g., a beacon signal is not detected due to currents or interference), a replanning episode is triggered. This allows the system to adapt to environmental perturbations and maintain trajectory accuracy.
   • Temporal Constraints: While the current implementation uses an untimed HDDL 1.0 model, the authors note the intention to evolve toward HDDL 2.1 to incorporate temporal constraints necessary for synchronized fleet coordination.

3. Software Architecture:
   The framework is implemented as a suite of QGIS plugins.

   • One plugin handles the beacon deployment algorithm.
   • TraceQGIS serves as the visualization and execution interface. It loads HTN planning models (in HDDL language) and configuration files (YAML), visualizing entities (UUVs, beacons), communication links, and movement traces on bathymetric maps.

Key Contributions

• Beacon Deployment Algorithm: The application of a Constrained Lloyd's algorithm to optimize the placement of acoustic beacons for underwater localization, ensuring maximal coverage and signal availability within a specific mission volume.
• HTN-Based Navigation Framework: The adaptation of Hierarchical Task Network planning for underwater GNSS-denied navigation. This approach allows for the decomposition of complex missions into adaptive sequences of actions, handling uncertainty through continuous monitoring and replanning.
• Integrated QGIS Implementation: The development of a practical software toolchain (TraceQGIS) that integrates beacon planning, HTN model loading, and real-time mission visualization, demonstrating the feasibility of the approach in a geospatial context.
• Stealthy Navigation Strategy: A methodology that avoids the need for UUVs to surface for GNSS fixes, thereby preserving operational stealth in restricted or dangerous zones.

Results
The paper presents a simulation-based validation of the approach using a scenario off the coast of Capo Testa, Sardinia.

• Scenario: A fleet of UUVs with uncertain initial positions navigates toward a goal using five deployed beacons.
• Execution: The simulation demonstrates UUV 1 navigating toward a beacon, detecting its acoustic signature, circling to refine its position, and broadcasting a message to the fleet. Other UUVs then adjust their paths based on this broadcast.
• Visualization: The TraceQGIS interface successfully visualized the mission layers, including bathymetric data, beacon positions, UUV trajectories, and communication events, confirming the system's ability to represent the mission state and the HTN execution flow.

Significance and Claims
The authors position this work as a necessary advancement for autonomous operations in GNSS-denied underwater environments, a domain where public research is currently scarce compared to aerial or terrestrial robotics. The significance of the approach lies in its ability to:

1. Enhance Robustness: By utilizing a hierarchical planner with continuous monitoring, the system can dynamically adapt to environmental uncertainties (currents, signal loss) rather than relying on a static path.
2. Enable Covert Operations: The reliance on submerged or floating beacons and acoustic signals allows UUVs to operate without surfacing, maintaining concealment from surveillance systems.
3. Provide a Scalable Framework: The use of HTN planning offers a structured way to manage complex missions, allowing for the decomposition of tasks and the generation of alternative action sequences based on the fleet's real-time state.

The paper modestly acknowledges current limitations, noting that the initial model assumes independent UUV navigation and does not fully account for the complex constraints of inter-UUV communication or the unknown target position problem (which could be addressed via trilateration in future work). The authors emphasize that the development of a temporal HTN planner and the integration of communication constraints represent critical future steps to fully realize the potential of this fleet-based navigation system.