ChatterBox Secure Off-Grid Mesh Texting - Tuning/Testing via Simulation

In order to prove out the mesh-related algorithms I've built for the ChatterBox, I built a python-based simulation of the system. Here, I demonstrate a visualization of the running simulation (using "Mesa" as the simulation driver). The simulation I'm showing here uses the same algorithms I have built into the ChatterBoxes, but the algorithm needs tuned for real-world situations. My simulation randomly generates environments and ChatterBox positions/mobility/ranges, runs my algorithms in that environment, and then scores the performance. A genetic/evolutionary algorithm will be repeatedly running that simulation and tuning my algo settings to come up with some good settings.

For real-world deployments, this simulation could be run ahead of deployment, and generate an algorithm tuned precisely for that deployment, if desired. In fact, a machine running this simulation could be attached to a live cluster and tune it in real-time, although that kind of defeats the purpose of being able to just mindlessly use these devices and let them figure out all the technicalities, so I probably won't go down that route.

For more information about the ChatterBox project, visit https://chatters.io