Robust Communication-Aware Jamming Detection and Avoidance for Distributed Multi-Agent Systems

MATLAB implementations of Formation Control Algorithm.

#multi-agent systems   #communication-aware   #formation control
#particle swarm optimization   #path planning   #jamming area

Swarm robotics has emerged as an innovative field, enabling groups of agents to collectively tackle complex challenges. In this pursuit, Multi-Agent Systems (MAS) offer unparalleled potential for navigating through intricate environments. In this paper, an innovative control strategy is proposed to enable a swarm of agents to navigate through a complex environment, which includes regions affected by electromagnetic jamming. The swarm is unaware of the existence or location of these regions, known as jamming areas. The novel control strategy is proposed as a combination of a gradient and movement controller. The gradient controller achieves the desired swarm formation that maximizes the communication quality between agents. The movement controller uses Particle Swarm Optimization (PSO) and a path planning algorithm to move towards a destination while avoiding jamming areas. Various path planning algorithms, such as A*, Greedy Best First Search, Breadth First Search, Dijkstra, Jump Point Search, and Theta*, are compared in simulations. The results highlight the discovery of an optimal path planning algorithm that facilitates efficient and robust navigation for the MAS, while also upholding high communication quality.

Paper	Presentation

Deadlock Problem

Using Communication-aware Formation Controller + PSO in a complex jamming environment

t=0s	t=4s

Deadlock Resolution

Use Communication-aware Formation Controller + PSO + Path Planning Algorithm

Resolution Workflow

Final Simulation

Vid_Grid.mp4	Vid_Sim.mp4
Path Planning Grid Map	Swarm Trajectory

Simulation Evaluations

A*	Breadth First Search	Dijkstra
Greedy Best First Search	Jump Point Search	Theta*