What exactly is missing that requires reverse-engineering?
There must be lots of satellites that are older than 18 years that are still being actively monitored and controlled. I would not have thought the basic transmission methods and protocols would differ hugely from satellite to satellite (apart from deliberately obfuscated military stuff anyway), because ground stations will be monitoring and controlling many satellites, so you would not want to design every satellite so it needed its own unique set of ground equipment.
I would have thought that all you would need is the data format of the particular satellite, and the control codes (which would be unique for each satellite and possibly encrypted to avoid hacking), but surely the hardware needed to physically receive and transmit data to/from the satellite would be pretty generic?
I cannot believe that the technical information on the data formats and codes etc. would be lost from an expensive project less than 20 years old. If nothing else, surely the source code of the satellite's on-board computers (which have to generate & interpret the data) will be archived somewhere? That wouldn't need large amounts of storage space or have to be counted in every audit.
Biting the hand that feeds IT
