You really need to use matched filters and to modulate the transmitted sound with a psuedo-random bit stream. Its been discussed before with regard to distinguishing signals in a noisy enviroment which is what you are trying to do.I should try using google for matched filters to give you some background of the theory.Designing and implementing matched filter recievers is not for the faint hearted.
IIve got a feeling that it may be more trouble than its worth and youll end up trying to distinguish the signals by the time delay.