do published papers always have to present an innovation?
Not really, it has to make a contribution to existing understanding, and this can be a very specific contribution in a very specific area. Look at this paper for example:
Gauthier, Joseph P., Eamonn P. Glennon, and Andrew G. Dempster. "Timing Performance of V2R3 Namuru Operating in Position-Hold Mode." (2013).
It provides a very useful, but very context specific, contribution.
Of course the problem which you will face in trying to replicated a paper like this is that unlike these authors, you lack equipment and data to write up such results. However, you could consider getting data from one of the research laboratories working in your area and trying to publish something with them. This could be something where you write up some part of the analysis, or, if you really know your stuff, you take the data they have and then do your own analyses. More likely than not you would have to cite them as you are building on their work/collected data and efforts.
Another option for you would be to do some sort of literature review, for instance, timing performance for GPS satellites: A review of the literature. Something like this would also qualify as a contribution and could be publishable. Plus, if you did it in an area where you would like to do research, it would serve to signal that you are serious about that research and also happen to know quite a bit about it.