Record a signal with your own voice (use the word CAT and try to have a long A part and both the K and T sounds clear). Use 8khz sampling rate.
In all the assignment, divide the speech into 40ms (320 sample), non-overlapped frames. Or, if you find the above difficult, segment the speech manually to the basic 3 sounds (K A T)
1- Use the channel vocoder and use 20 channels over the 0-4khz frequency band. and use any pitch estimator you like. 2- Use the LPC vocoder and use the LPC order=10. Also, use the same pitch estimator as in the channel vocoder.