well, i assocaite C/I with Ec/No, because, roughly speaking, you do average your useful signal over the noise, in both case.
EcNo can easily be measured in Idle mode, and is actually a replacement for how RxLev is used in 2G. EcNo is really the most common measurement in 3G, and is used to know whether a cell is good or bad (much more so than the RSCP)
Each cell sends some pilot signals which can be decoded and compared against the global noise. I don’t know the exact names of those channels, but that shouldn’t be too hard to find out.