Author's reply

The point highlighted by Shalaby is an important point that we didn't take into consideration. In the original paper we considered that when an optical signal from a broadband source is split into two, the two signals can be considered as independent. Since the paper was accepted in this journal, and none of the three reviewers gave any comments regarding the correlation between the two split signals, we considered that our work was correct. In the following we show why we assumed that the two split signals were independent. It is obvious now that the assumption given was highly optimistic and some correlation will always exist between the two split signals. However, the results given in our equation [2] can be considered as an upper limit in the performance improvement if independence between the two split signals can be achieved.