Abstract
Recent years have seen an increased interest in the use of Non-terrestrial networks (NTNs), especially the unmanned aerial vehicles (UAVs) to provide cost-effective global connectivity in next-generation wireless networks. We introduce a resilient, adaptive, self-healing network design (RASHND) to optimize signal quality under dynamic interference and adversarial conditions. RASHND leverages inter-node communication and an intelligent algorithm selection process, incorporating combining techniques like distributed-Maximal Ratio Combining (d-MRC), distributed-Linear Minimum Mean Squared Error Estimation(d-LMMSE), and Selection Combining (SC). These algorithms are selected to improve performance by adapting to changing network conditions. To evaluate the effectiveness of the proposed RASHND solutions, a software-defined radio (SDR)-based hardware testbed afforded initial testing and evaluations. Additionally, we present results from UAV tests conducted on the AERPAW testbed to validate our solutions in real-world scenarios. The results demonstrate that RASHND significantly enhances the reliability and interference resilience of UAV networks, making them well-suited for critical communications.
Introduction
As we move toward 6G, Non-Terrestrial Networks (NTNs)—particularly Unmanned Aerial Vehicles (UAVs)—are becoming crucial for providing cost-effective, on-demand broadband coverage to remote, rural, or disaster-stricken areas. UAVs can operate collaboratively in swarms to form distributed antenna systems, improving signal quality through spatial diversity.
However, UAV swarms operate in highly dynamic and potentially adversarial environments. They face unpredictable interference, node failures, and active jamming. Because manual troubleshooting mid-flight is impossible, these networks require "self-healing" capabilities. They must be able to autonomously detect abnormalities, isolate compromised nodes, and recover signal integrity to keep critical communications online.
System Design and Methodology
The authors propose the Resilient, Adaptive, Self-Healing Network Design (RASHND). The setup uses a Single-Input Multiple-Output (SIMO) architecture where a ground transmitter sends data to a swarm of UAV receivers. The core innovation is an intelligent state machine that monitors the Bit Error Rate (BER) of each receiver in real-time.
Based on how many receivers cross a critical error threshold ($N_s$), the swarm dynamically selects one of three combining algorithms:
- d-MRC (distributed-Maximal Ratio Combining): Used when interference is low and all receivers have good signals ($N_s = N$). It combines data from all nodes to maximize the overall Signal-to-Noise Ratio (SNR).
- d-LMMSE (distributed-Linear Minimum Mean Square Error Estimation): Triggered when specific nodes face high interference ($1 < N_s < N$). It applies weighted combinations to filter out the interference while still utilizing the clean data.
- SC (Selection Combining): Activated under severe jamming when most receivers are heavily compromised ($N_s = 1$). It relies entirely on the single best-performing receiver.
To make this autonomous, the system relies on Inter-Device Links (IDL). The UAVs (equipped with USRP B210 software-defined radios and Raspberry Pis) communicate via an internal ad-hoc Wi-Fi network using TCP sockets. To prevent a single point of failure, the "leader node" role—which aggregates the data and runs the combining algorithm—rotates sequentially among the UAVs.
Experimental Results and Discussion
The system was evaluated across three environments: an indoor testbed, an outdoor tripod testbed, and the AERPAW testbed featuring actual moving UAVs.
Key findings from the hardware-in-the-loop evaluations include:
- Dynamic Adaptation Beats Static Algorithms: No single signal-combining algorithm is optimal for all conditions. As shown in the performance data, the system maintains a high combined data rate by transitioning from d-MRC to d-LMMSE and finally to Selection Combining as jamming power (Jx) increases.
- Jamming Resilience: In outdoor tests, introducing a -30 dBm jammer crippled one receiver, triggering d-LMMSE to maintain the signal. Increasing the jammer to 0 dBm crippled two receivers, triggering SC to isolate the surviving node. In all cases, the swarm maintained a viable connection where individual nodes failed.
- Decentralized Leadership: By rotating the "leader node" role among the Raspberry Pis, the system achieves true fault tolerance. If the current leader is jammed or loses connection, the system passes the baton to the next node, ensuring swarm survival.
- Real-World Flight Validation: Tests on the AERPAW platform with UAVs flying at 5 m/s at 50-60m altitudes proved the algorithms inherently account for real-world physical challenges like Doppler shifts and dynamic channel fading.
Conclusion
RASHND demonstrates that complex, real-time signal processing and device-to-device coordination can run effectively on lightweight, low-power commercial off-the-shelf hardware (Raspberry Pis), making it highly practical for payload-constrained UAVs. While the current study validates the concept on a small scale, future work will involve scaling the swarm size and testing the architecture against more sophisticated, non-stationary adversarial attacks in denser 6G network topologies.
Summary updated