Optimal Multi-Robot Communication-Aware Trajectory Planning by Constraining the Fiedler Value

TL;DR

A novel approach that builds on a general optimisation framework where the changes in robots positions are used as decision variable, and linear constraints on the trajectories of the robots are introduced to ensure communication performance and collision avoidance, and the Fiedler value is adopted as communication performance metric.

Abstract

The paper present a novel approach for the solution of the Multi-Robot Communication-Aware Trajectory Planning, which builds on a general optimisation framework where the changes in robots positions are used as decision variable, and linear constraints on the trajectories of the robots are introduced to ensure communication performance and collision avoidance. The Fiedler value is adopted as communication performance metric. The validity of the method in computing both feasible and optimal trajectories for the robots is demonstrated both in simulation and experimentally. Results show that the constraint on the Fiedler value ensures that the robot network fulfils its objective while maintaining communication connectivity at all times. Further, the paper shows that the introduction of approximations for the constraints enables a significant improvement in the computational time of the solution, which remain very close to the optimal solution.

Figures (11)

• Figure 1: An example of an ad-hoc network where five drones are communicating wirelessly with each other and a ground station antenna. Three of the drones are outside the communication range of the antenna (blue dashed line), and therefore have to rely on the other drones relaying their information.
• Figure 2: Link quality with increasing values of $\alpha$, in order of red, green, and blue.
• Figure 3: Voronoi partition for 5 points in $\mathbb{R}^2$.
• Figure 4: Positions of robots from simulation of inspection task where $N=10$ robots are deployed to inspect $L=4$ inspection points. The red cross indicates the position of the base-station.
• Figure 5: Fiedler over time from simulation of inspection task where $N=10$ robots are deployed to inspect $L=4$ inspection points.