BGPy
Network-security simulation work with BGPy, modeling routing attacks, defenses, and policy behavior across configurable topologies.
What this project is
BGPy is a Python simulator for studying BGP security. It can model Internet-scale topologies, routing attacks, defensive policies, and routing-origin-validation scenarios without treating a simulation as a production network.
My work with it was part of a research assistant role through the UConn School of Computing and WSRAP. I used the simulator to model routing attacks, defenses, and policy behavior, then ran repeatable scenarios across configurable network topologies.
The work
The useful question was not whether one simulated network looked safe. It was how a policy behaved when the topology, attack, or defensive rule changed.
I worked on scenarios that evaluated attack and defense behavior, kept the experiments repeatable, and improved the simulation workflow so more cases could be run without losing track of what changed. The result was analysis of policy behavior, not a claim about how the live Internet would respond.
Attribution boundary
BGPy is an upstream open-source project created by its original maintainers. Henry's public repository is a fork of jfuruness/bgpy_pkg; the fork is evidence of access and experimentation, not original authorship of the simulator.
The related bgpsimulator fork is separate and does not provide evidence of a completed Henry-authored implementation, so it is not presented as one here.
What I learned
Network-security experiments are only useful when the topology, policy, attack, and result can be repeated. A simulator makes those variables easier to change, but it does not remove the gap between a modeled route and a real network.
The work left me with a better habit: describe the experiment that ran, preserve the boundary around the result, and avoid turning a tool's upstream capabilities into a personal project claim.