One way to do this is to build a vector of (n+1) edge values (the first being the minimum signal value and the last being the maximum signal value), and to use discretize() to get out the number of which "bin" each sample would fall into. Then use that number to index back into the list of levels to get the quantized version of the signal.
bins = discretize(YourSignal, edges);
Quantized_signal = edges(bins);
