DSP_iir
[Filtering and convolution]

Collaboration diagram for DSP_iir:

---

Detailed Description

void	DSP_iir (short *r1, const short *x, short *r2, const short *h2, const short *h1, int nr)

---

Function Documentation

void DSP_iir	(	short *	r1,
		const short *	x,
		short *	r2,
		const short *	h2,
		const short *	h1,
		int	nr
	)

This real IIR computes nr real output samples using 4 Autoregressive filter coefficients and 5 Moving-average filter coefficients. It operates on 16-bit data with a 32-bit accumulate. The implementation is done in Direct Form I, for the below difference equation.

r1(n) = h2(0) * x(n)
+ h2(1) * x(n-1) - h1(1) * r1(n-1)
+ h2(2) * x(n-2) - h1(2) * r1(n-2)
+ h2(3) * x(n-3) - h1(3) * r1(n-3)
+ h2(4) * x(n-4) - h1(4) * r1(n-4)

Parameters:

	r1	Pointer to real output data used.
	x	Pointer to real input data.
	r2	Pointer to real output data stored.
	h2	Pointer to 5 Moving-average real filter coefficients.
	h1	Pointer to 4 Autoregressive real filter coefficients.
	nr	Holds the value of number of real output samples.

Algorithm:

DSP_iir_cn.c is the natural C equivalent of the optimized intrinsic C code without restrictions. Note that the intrinsic C code is optimized and restrictions may apply.

Assumptions:

Arrays x, h1, h2, r1 and r2 do not overlap
x must point to x[nr-4]
h1 points to h1[0], h1[0] is not used, all the coefficients are real
h2 points to h2[0], all the coefficietns are real
r1 must point to r1[nr-4], current response depends on previous 4 responses
r2 must point to r2[0], such that r1[4+i] = r2[nr], for i = 0 to nr;
nr >= 1;

Implementation Notes:

Interruptibility : The code is interruptible. Endian support : supports both Little and Big endian modes.

---

Copyright 2014, Texas Instruments Incorporated