The foreseen increase of Secondary Surveillance Radar (SSR) replies, as well the replies emanating from the ADS-B systems will cause many collisions ending into garbling the received replies. While with a traditional receiver the leading reply might be decoded, the trailing ones will suffer a higher failure rate, and the reply will need to be re-issued and lead to an increase of replies density. To combat this and the loss of data, we propose a new mathematical criterion, the Disjoint Component Analysis (DCA), which allows to separate the replies regardless of the SSR protocol, the power ratio or the delay between replies, as demonstrated by several simulations and real data analysis. As we are still missing a computationally efficient algorithmic solution to minimize the cost function of the DCA criterion, this paper mostly demonstrates the potential of said criterion.