Digital filter

A digital filter for filtering successive binary coded input samples X(n) in the form:

• ##EQU13## in which n is the number of the sample, B is the number of bits, per sample, and furnishing a series of coded values Y(n) related to the said input samples by the relationship:
• ##EQU14## in which m is the order of the filter, .alpha.i and .beta.i are weighting factors, comprising:
• first memory means furnishing a first value .phi..sub.j (Y) from m+1 address bits;
• second memory means furnishing a second value .phi..sub.j (W) from the same m+1 address bits;
• first means to combine the said successive first values .phi..sub.j (Y) so as to obtain a final value:
• ##EQU15## second means to combine the said successive second value .phi..sub.j (W) so as to obtain an intermediate value:
• ##EQU16## third means to apply serially to the said first and second memory means the B bits of the said input sample X(n);
• fourth means to apply serially to the said first and second memory means the B bits of each of the said intermediate values W(n-1) to W(n-m), the m j.sup.th bits of the said intermediate values constituting with the j.sup.th bit of the said input sample the m+1 address bits of the said memories;
• and in that the said first value .phi..sub.j (Y) is related to the said m+1 address bits by the relationship:
• .phi..sub.j (Y) = X.sub.i (n) + (.alpha..sub.1 -.beta..sub.1).W.sub.j (n-1) + (.alpha..sub.2 -.beta..sub.2).W.sub.j (n-2) . . . + (.alpha.m-.beta.m).W.sub.j (n-m)
• and the second value .phi..sub.j (W) is related to the said m+1 address bits by the relationship:
• .phi..sub.j (W) = X.sub.j (n) - .beta..sub.1.W.sub.j (n-1) - .beta..sub.2.W.sub.j (n-2) . . . -.beta.m.W.sub.j (n-m).