Tcp Congestion Control Visualizer

A visual walkthrough of TCP's slow start and AIMD congestion control.

Built an interactive TCP congestion control simulator visualizing how AIMD (Additive Increase, Multiplicative Decrease) converges to fairness across multiple flows sharing a bottleneck link. Models slow start, congestion avoidance, and synchronized congestion signals — rendering the classic sawtooth pattern live on a scrolling chart. Senders can be added manually or auto-spawned/despawned in a churning population, letting you watch the network re-converge toward equal share in real time. A live Jain's Fairness Index and fair-share reference line quantify convergence at every tick. The simulation engine is fully decoupled from the UI and unit-tested with Vitest, with deterministic, injectable randomness for reproducible tests.