Adaptive interference cancelling

I have selected my problem for the project as follows. In a car we have the audio system. There will be some noise coming from the engine when we try to play a song. We can use an adaptive filter in order to reduce the noise. The system configuration is shown in the figure 1.

In the figure S(k) is the speech signal, n(k) is the noise signal which is coming from the engine vibrations, n'(k) is the reference signal that is being used to reduce the noise, Sys(k), is the primary input to the filter that is the corrupt signal, H(z) is the transfer function of the adaptive filter, Y(k), is the output of the adaptive filter and c(k) is the output signal i.e., is the difference of the corrupt signal and the output of the adaptive filter. The refernce input i.e., n'(k) is in someway correlated with the noise and is uncorrelated with the original signal. If these conditions are violated, the adaptive filter may cancel the signal, s(k) in place of (or in additon to) the noise or could fail to cancel the noise. It can be proved that n(k) and n'(k) are correlated but n'(k) and s(k) are uncorrelated from the derivation given in appendix [1].