[Paper Review] Deployment of UAVs for Optimal Multihop Ad-hoc Networks Using Particle Swarm Optimization and Behavior-based Control

Ngan Duong Thi Thuy, Duy Nam Bui+2 more|arXiv (Cornell University)|Dec 26, 2022

UAV Applications and Optimization4 citations

TL;DR

This paper proposes a two-stage UAV deployment strategy for optimal multihop ad-hoc networks in disaster areas using Particle Swarm Optimization (PSO) to determine optimal UAV positions and a behavior-based controller to navigate UAVs safely to those positions. The approach achieves real-time, obstacle-avoiding deployment with average position deviation of 70 m and angular deviation of 0.2 rad, enabling reliable communication links over 1000 m with 4 UAVs in complex environments.

ABSTRACT

This study proposes an approach for establishing an optimal multihop ad-hoc network using multiple unmanned aerial vehicles (UAVs) to provide emergency communication in disaster areas. The approach includes two stages, one uses particle swarm optimization (PSO) to find optimal positions to deploy UAVs, and the other uses a behavior-based controller to navigate the UAVs to their assigned positions without colliding with obstacles in an unknown environment. Several constraints related to the UAVs' sensing and communication ranges have been imposed to ensure the applicability of the proposed approach in real-world scenarios. A number of simulation experiments with data loaded from real environments have been conducted. The results show that our proposed approach is not only successful in establishing multihop ad-hoc routes but also meets the requirements for real-time deployment of UAVs.

Motivation & Objective

• To address the challenge of establishing reliable, real-time multihop ad-hoc communication in disaster areas where infrastructure is destroyed.
• To optimize UAV node placement for maximum communication range and network reliability in unknown, obstacle-rich environments.
• To enable safe, real-time navigation of UAVs to assigned positions using a behavior-based control strategy that avoids obstacles and maintains network connectivity.
• To ensure scalability and computational feasibility for real-world emergency deployment scenarios.

Proposed method

• Formulate UAV deployment as a constrained optimization problem using PSO to find optimal positions that maximize multihop network connectivity and minimize link length.
• Model UAVs as particles with position $P=[x,y,z]^T$, heading $\psi$, and constraints on communication range ($R_C=300$ m), sensing range ($R_S=50$ m), and safe distance ($S=10D$).
• Use a behavior-based target tracking controller with parameters $a_{m2g}, b_{m2g}, a_{ath}, b_{ath}, a_{adr}, b_{adr}$ to guide UAVs along safe, obstacle-free paths to their target positions.
• Implement a sequential deployment strategy where each UAV searches within its sensing range for the optimal position closest to the communication boundary before the next UAV is dispatched.
• Integrate PSO with a two-phase navigation process: first, flying toward the last deployed UAV (assigned state), then exploring intermediate points (explore state) to reach the final position.
• Define a fitness function based on proximity to the line-of-sight path and network connectivity, optimized over 100 iterations with $w=1$, $c_1=c_2=1.5$, and population size 100.

Experimental results

Research questions

• RQ1 Can PSO effectively identify optimal UAV positions for multihop ad-hoc network formation in unknown, obstacle-rich environments?
• RQ2 Can a behavior-based controller enable safe, real-time navigation of UAVs to their target positions while avoiding obstacles and maintaining network connectivity?
• RQ3 How accurate and efficient is the proposed deployment strategy in terms of position deviation, path safety, and computation time?
• RQ4 To what extent does the method scale to longer distances and complex terrains with realistic sensor and communication constraints?

Key findings

• The PSO algorithm converged within 30–50 iterations, demonstrating fast convergence for real-time deployment.
• The average position deviation from the ideal line-of-sight path was 70 m, and the average target angle deviation was 0.2 rad, primarily due to environmental obstacles.
• The system required an average of 5 minutes to deploy 4 UAVs at 15 m/s, with PSO computation taking only 1 second per search, well within UAV flight time between points.
• The behavior-based controller computed control commands in milliseconds, confirming suitability for real-time operation.
• In both scenarios, the UAVs successfully formed a multihop network with an average link length of 295 m, close to the 300 m communication range.
• The deployment paths maintained safe distances from obstacles in complex environments, as confirmed by simulation results and video demonstrations.