Inputs
Link budget
TAV products in scope
Switch a product on and set its quantity. Mbps per unit are low–high ranges; the highlighted figure is the one the scenario and estimate point select.
| Product | Sizing unit | In scope | Qty | Typical low | Typical high | Peak low | Peak high | Diversity | Selected Mbps | Share |
|---|---|---|---|---|---|---|---|---|---|---|
| Total platform load | ||||||||||
Core rows (FMS to Authentication) take their rates from the source data below, scaled by Airport size. Only PFM (1 Mbps per camera) is a TAV-published rate; the others, and every diversity factor, are planning assumptions to replace with measured data.
Link designs compared
Same demand on every design. Pick a card to test it above. Each bar reads utilisation of the nominal rate: green up to the comfort limit, amber up to usable capacity, red past it (the dark line is line rate). Active-active only counts with verified load sharing.
Unit converter
Volumes assume the link runs at 100% all the time over a 30-day month. 1 byte = 8 bits; decimal units (1 GB = 1,000 MB).
Tips and definitions
- Mbps vs MB/s
- Links are sold in bits, files are measured in bytes. 1,000 Mbps is about 125 MB/s.
- Nominal vs usable
- Usable = nominal minus the headroom reserve.
- Which scenario
- Pilot: Peak, High. Production: Peak + Growth, High with a 50% growth margin, which reproduces the original 450 Mbps design basis.
- The 80% rule
- Sustained load above ~80% of usable capacity causes queueing, latency spikes and packet loss. The verdict turns amber there.
- HA vs active-active
- HA dual links must carry the full load if one fails, so plan on one link.
- Avoid double counting
- FMS, RMS and A-CDM already include standard integrations; use the Integration Broker row only for extra interfaces. Staff counted as workstations or users in a product row should not be counted again under Concurrent active users.
- Peak diversity
- Hundreds of check-in desks, handhelds or phones do not all peak in the same second. Diversity factors below 1 reduce their peak load; set a product to 1 for the worst case.
- Where traffic flows
- Rates describe WAN and data-centre traffic. Local LAN traffic counts only if it crosses this link.
- Latency and QoS
- FMS, RMS, A-CDM and CUPPS are time-critical. Bandwidth does not fix latency; prioritise operational traffic with QoS.
Source data
The embedded figures every calculation reads. Edits here flow through the whole page.
Core platform rates, link designs and user profiles
editable
| Core product | Typical low | Typical high | Peak low | Peak high |
|---|
| Link design | Nominal Mbps |
|---|
| User profile | Mbps / active user |
|---|
| Airport size (core multiplier) | × core rates |
|---|
Standard link tiers used for the recommendation: 100, 200, 300, 500 Mbps, 1, 2, 3, 4, 5, 10, 20, 40 and 100 Gbps. 3 and 4 Gbps are usually n × 1 Gbps aggregates or a 10 Gbps port with a lower committed rate; check availability with the carrier.
Original FMS / RMS / A-CDM sizing (Sheet1)
recomputed
| Scenario | Sum of core rows | Workbook said | Check |
|---|
Workbook recommendations: 500 Mbps VC for the pilot, dual 1 Gbps connections for production.
Additional concurrent users per design (User Capacity sheet)
recomputed
Fixed platform load of 300–450 Mbps (the original growth-margin envelope) on each design, with the current headroom and user profiles. Users = max(0, usable − fixed load) ÷ Mbps per user, rounded down. Ranges run from the high-load to the low-load case.
| Design | Nominal | Usable | Spare @ 450 | Spare @ 300 | Lean | Standard | Heavy |
|---|
Corrections to the v3 workbook
What changed from Bandwidth Calculations v3.xlsx and why. The same fixes are in v8 of the workbook.
Typical total is 48–95 Mbps, not 50–95
Sheet1 rows add up to 10 + 15 + 10 + 10 + 3 = 48 Mbps at the low end of Typical. The peak totals (180–300 Mbps) were correct.
Growth margin now 50% by default, and adjustable
The calculator used 10%, but its own tip and the User Capacity sheet are built on the 300 → 450 Mbps basis (+50%). At 10% the two sheets disagreed: 940 versus 700 extra users on 1 Gbps. At 50% they agree. Use the slider or presets to test other margins.
The recommended link now passes its own verdict
The recommendation only removed the 20% headroom, while a green verdict also needs demand at or below 80% of usable capacity. At 750 Mbps demand it recommended 1 Gbps and then rated that link "tight, plan an upgrade". The recommendation is now the smallest standard tier that scores green: demand ÷ (0.8 × (1 − headroom)). The red verdict quotes the same figure.
Max users is shown for both limits
The workbook's max users filled the link to 100% of usable capacity, which the verdict rates amber. The budget now shows the green-fit limit next to the usable limit.
Peak diversity
Peak loads used to be added as if every device peaked in the same second, which overstated large fleets (CUPPS, BRS, phones). Each product now has a diversity factor, applied in Peak and Peak + Growth, with an on/off switch. Tests on five synthetic airports cut demand by 13–43%.
Core rates scale with airport size
The FMS, RMS and A-CDM rates are fixed at medium-airport values, so a regional airport carried 555 Mbps per million passengers against 161 at a hub. An Airport size multiplier (editable in Source data) now scales them; Medium is ×1, so the original basis is unchanged.
More link tiers
The recommendation jumped from 2 to 5 Gbps and stopped at 10 Gbps. Tiers 3 and 4 Gbps and 20, 40 and 100 Gbps were added.
Default quantities lowered
600 CUPPS workstations and 410 BRS readers were about five times a medium airport's likely count. They are now 100 and 60, marked as planning quantities to confirm.
Off-link and double-counting warnings
The passenger app and GHS mobile rows carry an "often off-link" tag because that traffic usually uses the internet or 4G/5G. The users hint and tips say not to count staff twice.
Safety stack and zero demand
A line under the products shows how growth, the 80% comfort limit and headroom multiply into the required link size. With nothing switched on the recommendation now shows a dash instead of 100 Mbps.
Checked and correct as built: per-product selection (Mid = average of low and high), Peak + Growth reading the peak columns, link utilisation, spare capacity, the 4-design comparison and the unit converter (1 Gbps = 125 MB/s = 450 GB/h = 324 TB per 30-day month).