Feasibility study of the event-mixing technique in the context of a SUSY search in the same-flavour opposite-sign di-lepton channel

The $t\bar{t}$ production is one of the main backgrounds for many BSM searches looking for new structures in the invariant mass distribution of di-leptons produced in LHC collisions. This report studies the feasibility of applying the data-driven estimation method, the event-mixing technique, to determine that distribution. The event assumes a complete decorrelation between the two leptons, which has seen to be false in most of the control regions. For that reason, we conclude that the method is not robust enough to be applied as a background estimation in BSM searches.