I admittedly have no idea to set up a connection to a satellite, but this story seems like there were security failures at several levels.
According to the report, the satellites were accessed via a ground control station, which is how the attackers potentially got in.
The Landsat-7 and Terra AM-1 satellites utilize the commercially operated Svalbard Satellite Station in Spitsbergen, Norway that “routinely relies on the Internet for data access and file transfers,” says the commission, quoting a NASA report.
If I were designing a system to communicate with an expensive and sensitive piece of equipment, routine access to the Internet would not be part of the spec. There should be a ground control system that only connects to the satellite. That system would be under tight physical security to ensure only authorized satellite operators had access.
If that ground control system needs any new data, or needs satellite data transferred off, it would be ideally be done through removable media (burn to DVD, external HDD). If that will not work for whatever reason, the ground control system could be networked to another system. Access would have to go through a well configured firewall and require strong, two-factor authentication.
Making a system “hacker-proof” is not a realistic goal. Anything can be hacked if someone has the resources and determination to gain access. The best we can do is set up enough checks and safe-guards to make it not worthwhile to keep trying.
In this case, it clearly wasn’t enough to dissuade the attackers.