Simulating Decentralized Rescue with Agent-Based Modeling

In my recent post on Comparing Mesh Networking Systems, I explored how different protocols—LoRa, Wi-Fi HaLow, and hybrid approaches—shape resilience in disaster scenarios. To complement that analysis, I want to share a simulation I built during an agent-based modeling course: a NetLogo model of decentralized search-and-rescue using mesh communication.

Why Model Rescue with Mesh?

Disaster environments are unpredictable. Survivors may have limited battery life, damaged devices, or no direct line-of-sight to rescuers. A mesh network allows signals to propagate indirectly—so even if one survivor can’t broadcast, they may still be discovered if they’re connected to others in the mesh. This mirrors real-world challenges where partial connectivity can still save lives.

How the Simulation Works

• Survivors: Stationary agents who may signal for help. Their ability to broadcast is limited by battery life.
• Rescuers: Mobile agents who traverse the terrain, detect signals, and move toward survivors.
• Mesh-links: Survivors within range form undirected links, creating a communication backbone. Even when a survivor’s battery dies, their presence in the mesh can still be revealed through multi-hop propagation.

The model highlights:

• Battery-aware signaling and its impact on discoverability
• Multi-hop mesh propagation, where rescuers query entire connected clusters
• Visual tagging of discovered survivors, showing how indirect communication extends reach

What It Reveals

The simulation makes visible a key principle: resilience emerges from connectivity, not just individual signaling power. Survivors several hops away from an active signaler can still be discovered, underscoring the importance of topology-aware rescue strategies. It also shows how resource constraints—like battery drain—shift the balance between direct detection and mesh reliance.

Why It Matters

By experimenting with parameters like signal range, survivor density, and battery drain, we can explore how mesh systems behave under stress. This kind of modeling doesn’t replace field validation, but it provides a sandbox for testing ideas before hardware hits the ground. It also bridges technical research with education, making abstract concepts tangible for students, responders, and policymakers.

This simulation is a small but powerful example of how agent-based modeling can inform disaster tech design. Just as comparative protocol analysis helps us choose the right tools, simulations like this help us understand the dynamics of decentralized rescue—where every link in the mesh can mean the difference between isolation and discovery.