Basically the iOS SDK should do the following:
1. Record the audio from my smartphone for short interval and repeat this process throughout.
2. Process the recorded audio data in each recording cycle by taking the FFT and then taking out the frequency from that data
3. Check if the frequency range belong to a particular frequency band, if yes then store the frequency value in a array.
4. Send this array to my server.
NOTE : Find a way to bypass the red bar that comes up when i record audio in the background. The service should run in the background, regardless of the state of the application, with which it came as integrated. It should not be a hack and should pass app store check.