The WIKI Higgs boson site has an interesting diagram illustrating likely Higgs mass intervals that experiments (LEP & Tevatron) or indirect measurements have determined with a 95% confidence level. It posits that the feasible mass ranges are in the intervals of about 115-158 GeV or 175-185 GeV. The Tevatron confirmed the existance of the top quark (172 GeV) in 1995, so it had the energy and time to detect the Higgs in the 115-158 GeV ranges, but could not. Given LHC's greater energy and luminosity, why haven't there been anouncements of the Higgs discovery? Do we understand enough about potential particle creation to definitively know they are Higgs decays?
There are two considerations:
a) if you follow a bit the current theories, the manifestation of the Higgs decays can vary, depending on the parameters of the models, and there are many models.
b) Hadron colliders, in contrast to e+e- ones are "dirty", there are enormous backgrounds that have to be understood, and the few clear decay channels proposed by theories have small crossections and have to be fished out of these backgrounds. So more data than the ones existing are necessary to be able to answer definitively the Higgs question.
LHC is now building up the statistics necessary.
So what I foresee will happen, if we are lucky, is first a new resonance in some of the channels checked will be declared with at least 4 sigma. Then everybody will fall on it, experimentalists to find other decay channels and theorists to fit specific models to decide whether it is the Higgs or not.
The decay channels were discussed here:
Approximately, the Higgs decays to the heaviest pair of particles that is still consistent with energy conservation. The precise rates are calculable by the Standard Model - if you substitute the so-far unknown Higgs boson mass as input. Of course, if the world isn't described by the ordinary Standard Model, predictions may change even though (for the lightest Higgs, if one exists) the difference is usually not qualitative.
It takes some time - more precisely, some number of collisions - for a collider to distinguish the decaying Higgses from other, Higgs-free processes which may produce the same final particles (the so-called background), to be sufficiently sure that some Higgses were produced.
The required number of collisions depends on the Higgs mass. You get graphs like this one:
The higher the curve is in the $y$-direction for a given mass on the $x$-axis, the easier it is to find the Higgs if the mass is there, and also the easier it is to show that it's not there if it is not there.
A very light Higgs, near 115 GeV for example, is very hard to be found at the LHC. The work that the Tevatron has done - several inverse femtobarns - will not be surpassed soon. For any substantially higher masses, the LHC energetic advantage is so huge that the LHC, with a much smaller amount of collisions, has already done more to decide whether the Higgs is there than the Tevatron did, or it will soon be the case.
In the whole intervals where the Higgs is conceivable, it should be found by the end of 2012 - at least at 4 sigma - which is why the run was extended so that the Higgs is "almost certainly" found before the upgrade.