Résumé : The Large Hadron Collider (LHC) marks a new era for particle physics. Thanks to the very high energy of the proton beams, and to the large amount of data collected in years 2010-12, the physicists at CERN can test the standard model (SM) consistency, can explore the scalar sector, and search for any hint of new physics. A measurement of the Z boson pair production cross section using proton-proton collisions at 7 and 8 TeV center-of-mass energy, recorded by the CMS (Compact Muon Solenoid) experiment is presented. Diboson production at the LHC is of particular interest: it validates rare SM processes never tested before with high accuracy, it probes the electroweak boson self-interactions, and it constitutes a background for many physics searches such as the searches for the Higgs boson or supersymmetry. The ZZ production cross section is measured via the decay channel ZZ to 2l2v. The data used for the analysis have been recorded in years 2011 and 2012 by CMS and correspond to an integrated luminosity of about 5.1 fb-1 at 7 TeV and 19.6 fb-1 at 8 TeV. The data are selected requiring the presence of two isolated leptons (electrons or muons) of the same flavor with high transverse momentum (pT). In addition, events containing jets or additional leptons are vetoed, and it is applied a selection based on the dilepton pT, on the dilepton invariant mass, and on the transverse momentum imbalance (Emiss). The main SM backgrounds for this analysis are the Z/gamma+jet (Drell-Yan) process, the fully leptonic tt and single-top decay, and the WW and WZ diboson processes. The Z/gamma + jet process has no neutrinos in the final state, thus it has no large Emiss, but its cross section at the Z-peak is four orders of magnitude larger than the ZZ production. For this reason, even if the fraction of Z/gamma + jet events reconstructed with significant instrumental Emiss is not large, a high Emiss cut must be applied in order to improve the signal purity. High Emiss in Z/gamma + jet events is due to misreconstruction of physical objects, and to the additional energy deposits arising from the other proton-proton interactions occurring in the same bunch crossing (pileup). These effects are not well described in simulation, for this reason a high statistic control sample has to be used in order to model these tails from data. The fully leptonic tt, the single-top decay, and the WW process have been estimated in a control sample obtained requiring exactly one electron and one muon in the final state. The WZ process instead, has been estimated directly from the simulation. All the background shapes and normalizations are then constrained to data from a fit to the shape of a discriminating variable (e.g. Emiss), allowing only the ZZ signal normalization to freely vary. The selected data were also analyzed to search for anomalous triple gauge couplings (aTGC) involving the ZZ final state, and subsequently combined with the ZZ to 2l2l' final state data, to increase the sensitivity. In the absence of signs of new physics we set limits on the relevant aTGC parameters. Between 2023 and 2025, to extend its discovery potential and/or characterize any new signal possibly discovered, the LHC will increase its instantaneous luminosity by a factor of 10 beyond its design value. The increasing in luminosity will produce an average of 140 pileup interactions, this will represent an issue for the trigger and the reconstruction, degrading the jet and photon energy resolution and all the physic objects isolation quantities. In order to maintain full sensitivity, from low to high energy scales under severe pileup and radiation conditions, the L1 trigger, the tracker, the ECAL endcap and the HCAL, and the forward muon system will be upgraded. Moreover it will be fundamental to tag and remove the extra activity from PU interactions in order to correct the relevant variables measurements. The last part of the thesis will test the utility of timing in pileup mitigation and object reconstruction. Timing could be exploited for the association of photons, electrons and jets to their collision vertices, for particle identification, or to reject energy deposits coming from secondary vertices. A time measurement, extract from simulation, will be used in several reconstruction algorithms, showing improvements in particle reconstruction.