Budget a service for real users

Airbus’s prelaunch release describes a design life exceeding fifteen years, a launch mass of 5,600 kilograms and electrical power capability above 9 kilowatts. These are different kinds of specifications. A design-life target is not a current health report, and spacecraft electrical power is not the same quantity as radio power received at a terminal or useful data delivered to a school.

Consider an invented 10-kilowatt spacecraft budget. Platform functions require 3 kilowatts and the plan reserves 1, leaving 6 for the modeled payload allocation. Asking for 7 would exceed that allocation. These are fictional numbers, not a reconstruction of TIBA’s internal power system. Changing the allocation also requires checking thermal and other limits.

Suppose a fictional link carries 30 megabits per second of useful application traffic after overhead. Three independent users request 8, 10 and 15 megabits per second at the same time. Their combined demand is 33, exceeding supply by 3. Each request fitting individually does not mean they fit together.

An illustrative equal-share cap of 10 lets the users receive 8, 10 and 10: 28 total, leaving 2 unused until another scheduling rule allocates it. Giving those 2 to the third user would produce 8, 10 and 12. These examples are choices for a simplified scheduler, not TIBA’s contractual allocation or a claim that equal sharing is always the right policy.

Add the users before promising the rate

A fictional school needs a 120-megabit lesson file within 30 seconds. At a sustained application rate of 5 megabits per second, transfer alone takes 24 seconds. If setup and all remaining delays total 4 seconds, the task takes 28 and meets the stated deadline. At 4 megabits per second, transfer alone takes 30; adding the same delay produces 34 and misses it.

A useful acceptance record identifies the endpoints, file size, concurrent traffic, timing method and conditions. Measure repeat runs if the requirement concerns repeatability. This turns broad ambitions for connectivity into a task that can be evaluated, while leaving broader educational outcomes to their own assessment. A successful file delivery does not itself establish improved learning.