This is a good question, and one many users have. In principle linking the same household between CPS samples is possible a unique household ID. Linking the March Basic samples (i.e. the October education supplement) to their corresponding ASEC (i.e. the poverty and income supplement) samples, however, is complicated by the fact that the ASEC includes an oversample of households from adjacent months. The IPUMS-CPS Team is currently working on a variable called “MARBASECID” to facilitate linking the two data sets. The variable is still undergoing testing, but here is a link to a working paper that goes into the details of how this variable was created as well as walks through some of the complicating factors. Additionally, this question and answer walks through an imperfect but useful alternative process that may be of use to you in the meantime.