End-to-End Path Designer
A whole transport service in one place — both paths, their latency, their availability and what they share — because the failures that survive good per-hop design are the ones that only exist across the whole path.
The service
Working path
Protection path
Both paths, side by side
Fade margin to availability
Every 10 dB of fade margin divides the outage by ten — that part is dependable. The occurrence factor in front of it is not: it depends on path length, terrain roughness, climate and frequency, and should come from ITU-R P.530 rather than from a guess. This is multipath only and says nothing about rain fade, which dominates above about 10 GHz.
Why per-hop design misses this
Each hop gets designed against its own limit by the right tool. A microwave link has its fade margin, a fibre span its loss budget, an amplified line its OSNR, a protected service its second path. All of them pass, and the design review moves on.
What no hop owns is the total. Latency accumulates across every span and every box, and the number that matters — round trip, end to end — belongs to nobody in particular. Availability multiplies down the chain, so a path of five good elements is worse than any of them. And the protection path is designed as a path, not as a comparison with the one it protects.
Latency is mostly geography
Light in standard single-mode fibre travels at about two thirds of its speed in vacuum, which works out at 4.90 microseconds per kilometre — roughly a millisecond for every 204 km, or two milliseconds round trip. That figure is not negotiable; it is a property of glass.
Microwave is faster over the same ground, at 3.34 µs/km, because radio through air travels at very nearly the vacuum speed. That is the entire reason low-latency routes are built out of towers rather than trenches, despite everything else about microwave being harder.
Equipment delay is the part you can change. On a long path it is a rounding error; on a short one with many hops it can be most of the total, and the answer there is fewer boxes rather than a shorter route.
Protection is a different path, not the same path twice
Diversity means the protection route avoids what the working route uses, and avoiding things means going further. A protection path 50% longer than the working path is entirely normal and usually a sign the diversity is real.
The consequence is that latency changes when protection operates. For most services that does not matter. For synchronous storage replication, for a trading feed, for anything with a tight acknowledgement loop, it can turn a successful protection switch into a service outage of a different kind — one where every link is up, every alarm is green, and the application is broken.
The fix is usually not technical. It is knowing the number in advance, writing it into the service description, and deciding deliberately whether the protection path is fit for the traffic it will carry.
What this does not do
It composes figures you supply; it does not design hops. Per-hop budgets belong to the tools that do them properly — the microwave link budget, the fibre loss budget, OSNR and shared-risk analysis.
It does not model protection switching time, restoration behaviour, queueing or congestion delay, jitter, or anything about the traffic itself. Propagation and equipment delay are the floor, not the whole story: a congested link adds delay that no amount of geography explains.
Everything is calculated in your browser. Route lengths, site names, risk tags and availability figures describe your network and are not uploaded.