floating point usage in verilog

hello, i have a doubt in giving inputs to flip-flop as floating point. say, H(z)=h(0)+h(2)z(-1)+h(4)z(-2). inputs are predefined... how to implement this??????