If you're doing both ends of the connection, then I'd just stick the bytes received from the microphone streamer, directly into UDP packets to send to the receiver, and on the receiving end, take the bytes out of the received UDP packet and pass them over to the audio output streamer.
Albeit perhaps you might want to add a header to each UDP packet, specifying the sample rate (8 kHz? 44.1 kHz?), sample size (8 bit? 16 bit?) and maybe a sequence number (although, over a wifi network, it's unlikely that packets will arrive out-of-sequence - that's more something that happens over long distances over the internet).