This is a defensive model of how a cyberattack on one critical-infrastructure sector can cascade into others through shared operational dependencies (power feeds water pumps, water cools hospital equipment, telecom carries SCADA control traffic, and so on) — the same interdependency risk that national CERTs and ICS-CERT advisories analyse. It uses the independent cascade model from network science: at each discrete step, every newly-compromised node gets exactly one chance to "infect" each of its still-healthy neighbours.
P(i → j) = p₀ · I · w_ij · (1 − S) · (1 − 0.6·R_j)
p₀ base transmission probability per edge
I attack intensity slider (0–1)
w_ij dependency weight of the edge i→j
S network segmentation (0–1) — cuts every edge's chance
R_j redundancy investment at node j — backup systems that
absorb a compromised dependency without failing over
Segmentation (Zero-Trust microsegmentation, Purdue-model zone separation) multiplies down every edge uniformly — it is the defender's main lever to stop a compromise from crossing sector boundaries. Redundancy (backup power, failover water treatment, offline hospital protocols) makes an individual node resistant even when an upstream sector it depends on goes down. A node recovers to "contained" once its incident response completes, roughly proportional to redundancy.
- Attack intensity — how aggressively the initial compromise attempts to spread (attacker skill/resourcing).
- Segmentation — defender's network isolation between sectors; raises it to sever cascade paths in real time.
- Redundancy — defender's investment in backup systems per sector; raises it to shorten containment time.
- Launch Attack — starts the cascade from the Power Grid node with the current parameters; particles show live control-traffic flow along the dependency graph, red pulses show an active exploit attempt.