What I need is to take the address from the first .csv and match it to the GPS co-ordinates in the second .csv file.
...
Being GPS co-ordinates there is some margin of difference -/+10

So you want to merge (JOIN) two data sets, keyed on integer data within a range. That sounds very much like a job for a database. I would expect any decent RDBMS to handle this task with ease (and most will have a load-from-CSV option saving you effort there too). Is there anything you aren't telling us which would preclude solving this in the database and then using Perl to process the results?