First the citation issue. Citing from the ieeetran manual
Citations are made with the \cite command as usual.
IEEEtran will produce citation numbers that are individually
bracketed in IEEE style. (“[1], [5]” as opposed to the more
common “[1, 5]” form.) The base IEEEtran does not sort or
produce compressed “ranges” when there are three or more
adjacent citation numbers. However, IEEEtran pre-defines
some format control macros to facilitate easy use with Donald
Arseneau’s cite package [13]. So, all an author has to do
is to call cite package:
\usepackage{cite}
and the adjacent citation numbers will automatically be sorted
and compressed (ranged) IEEE style. (Of course, multiple
adjacent citations should always all be declared within a
single \cite, comma separated, for this to work.) Invoke
cite.sty’s noadjust option to prevent an unwanted leading space
from occurring should a citation ever need to be enclosed in
parenthesis.
.......
Note that, if needed (e.g., next to a non-punctuation/nonspace
character), cite.sty’s \cite command will automatically
add a leading space. i.e., “(\cite{mshell01})” will become
like “( [1])”. If this behavior is not desired, use the cite
package’s noadjust option (cite.sty V3.8 and later) which will
turn off the added spaces:
\usepackage[noadjust]{cite}
However, one should keep in mind that IEEE has an unexplainable
what-you-submit-is-NOT-what-you-get type of workflow. I can understand that they might have reasons for that but still, sometimes the changes are too dramatic. Citation style is also affected from that in particular they are probably modified manually by a copy editor. Most often, the missing Oxford commas are added. Moreover, the compressed and uncompressed citations are separated from each other as groups. Some examples from published articles
Long story short, it seems that IEEE does not fully automate their citation style which is context dependent, therefore, for the users simply following the manual is sufficient. A copy editor will fix your article hence there is no need on the user side to tweak the underlying citation mechanism. You don't even need to balance the last page as they will manually insert line breaks to the bibliography items and adjust the spacing accordingly. I think this is also unfair to the author of ieeetran class as I'm sure he can come up with fully automated mechanism had IEEE decided on something concrete as a standard.
\begin{rant}
As it happened to me with siunitx usage, they actually don't like extra bits, for example don't even think of leaving your TikZ code in, they will be replaced with bitmap images. They also tend to remove TeX modifications (in my case changed all the per-mode=frac typeset units into hard coded N/m and I'm not even going into the things that they do to your vector images).
Moreover, they don't allow you to receive the modified TeX file but you annotate the corrected proof PDF result. Hence, you only send your rough idea about the article and they create it for you. That's cutting edge technology right there. It might happen that some people just want to see their hard work in a mildly pleasing format.
The counter argument for this is the idiot academician who can't manage to write a proper minimal LaTeX code but somehow manage to prove things using sophisticated math arguments. Scientific publication is part of the job description of the academician. They are responsible to write proper articles as much as they are responsible for the contribution of their results. The whole IEEE community can not resort back to its weakest link just because some person refuses to read the manual (probably with an all-time classic I don't have time for that nonsense argument invoked; Firm steps towards the fourth Nobel prize ladies and gentlemen step aside!).
So IEEE can send the article back to the author and ask for a better code. Somehow the article payments and extra page charges are fully automated to the smallest detail and the procedure just stops until we pay... a short dramatic silence.... So same care can go into the content too. I'm not asking much; just the 175$ extra page charge for my online article which is probably enriched with BeCu compounds such that the paper size is invariant under all adverse climate changes and remains in its optimal readability conditions for at least 400 hundred more years.
\end{rant}
\usepackage[noadjust]{cite}to remove the trailing space from the citation numbers. In fact in the accepted proof version, they will further make it like[1],[2] and [3]if less than three citations are compressed. They will also do terrible things to your article but they are not relevant here. – percusse Oct 18 '12 at 01:47Have a nice day!
– Ales Oct 18 '12 at 12:50