That talks about the OSes (primarily) involved, and also discusses why agencies are slow to adopt newer OSes and languages (tried-and-true beats novelty, in general, due to the cost in time and money of testing/verifying/validating).
Intra-system shouldn't be too hard with today's technology. NASA has highly accurate models that could run on a computer that tracks the ship's movement using similar parameters to the calculations they do for space flight today. So instead of GPS satellites you could rely on internal systems. If everyone is calibrated the same, then the results/coordinates should be sufficiently accurate.
Inter-system will be very difficult. If you're talking full universe, then you would need to map the universe. If possible, picking a stationary reference to use as the core of the navigation coordinates would be good, but I wouldn't hold my breath since everything is moving. Otherwise we could use the reference we know the most about - earth. This could work, but would require extensive mapping and testing since at the further distances we would essentially be calculating our trip to where a planet will be based on where it was given that they light we use to observe it took so long to get here.
So really, mapping is the big deal and there's no way of knowing if the greatly distant object we are travelling to is in the place we expect it to be, or if it even still exists today (the light is so "old"). It's a very interesting problem. One I probably will never live to see the answer to.
> NASA has highly accurate models that could run on a computer that tracks the ship's movement using similar parameters to the calculations they do for space flight today
The information useful to navigation has been available in encyclopedias for a long time, but not to the same accuracy as we know today. All we really need is the measures of the orbits and their distances. This would allow calculations for travel between planets. If keeping track of the mass, thrust, trajectory, and any external influences (impact of space debris), then determining one's position should be fairly accurate.
Google Sky has intricate models of the stars from the perspective of earth. Just something I found to be cool and thought I'd mention it.
https://arstechnica.com/features/2020/10/the-space-operating...
https://news.ycombinator.com/item?id=24661998 (HN submission of the above, but no actual discussion yet)
That talks about the OSes (primarily) involved, and also discusses why agencies are slow to adopt newer OSes and languages (tried-and-true beats novelty, in general, due to the cost in time and money of testing/verifying/validating).