GNU Radio Manual and C++ API Reference 3.10.5.1
The Free & Open Software Radio Ecosystem
iir_filter.h
Go to the documentation of this file.
1/* -*- c++ -*- */
2/*
3 * Copyright 2002,2012 Free Software Foundation, Inc.
4 *
5 * This file is part of GNU Radio
6 *
7 * SPDX-License-Identifier: GPL-3.0-or-later
8 *
9 */
10
11#ifndef INCLUDED_IIR_FILTER_H
12#define INCLUDED_IIR_FILTER_H
13
14#include <gnuradio/filter/api.h>
15#include <gnuradio/gr_complex.h>
16#include <stdexcept>
17#include <vector>
18
19namespace gr {
20namespace filter {
21namespace kernel {
22
23/*!
24 * \brief Base class template for Infinite Impulse Response filter (IIR)
25 *
26 * \details
27 *
28 * This class provides a templated kernel for IIR filters. These
29 * iir_filters can be instantiated with a set of feed-forward
30 * and feed-back taps in the constructor. We then call the
31 * iir_filter::filter function to add a new sample to the
32 * filter, or iir_filter::filter_n to add a vector of samples to
33 * be filtered.
34 *
35 * Instantiating a filter means defining the templates for the
36 * data types being processed by the filter. There are four templates:
37 *
38 * \li i_type the data type of the input data (i.e., float).
39 * \li o_type the data type of the output data (i.e., float).
40 * \li tap_type the data type of the filter taps (i.e., double).
41 * \li acc_type the data type of the internal accumulator (i.e., double).
42 *
43 * The acc_type is specified to control how data is handled
44 * internally in the filter. This should always be the highest
45 * precision data type of any of the first three. Often, IIR
46 * filters require double-precision values in the taps for
47 * stability, and so the internal accumulator should also be
48 * double precision.
49 *
50 * Example:
51 *
52 * \code
53 * gr::filter::kernel::iir_filter<float,float,double,double> iir_filt(fftaps, fbtaps);
54 * ...
55 * float y = iir_filt.filter(x);
56 *
57 * <or>
58 *
59 * iir_filt.filter(y, x, N); // y and x are float arrays
60 * \endcode
61 *
62 * Another example for handling complex samples with
63 * double-precision taps (see filter::iir_filter_ccz):
64 *
65 * \code
66 * gr:;filter::kernel::iir_filter<gr_complex, gr_complex, gr_complexd, gr_complexd>
67 * iir_filt(fftaps, fbtaps); \endcode
68 */
69template <class i_type, class o_type, class tap_type, class acc_type>
71{
72public:
73 /*!
74 * \brief Construct an IIR with the given taps.
75 *
76 * This filter uses the Direct Form I implementation, where
77 * \p fftaps contains the feed-forward taps, and \p fbtaps the feedback ones.
78 *
79 * \p oldstyle: The old style of the IIR filter uses feedback
80 * taps that are negative of what most definitions use (scipy
81 * and Matlab among them). This parameter keeps using the old
82 * GNU Radio style and is set to TRUE by default. When taps
83 * generated from scipy, Matlab, or gr_filter_design, use the
84 * new style by setting this to FALSE.
85 *
86 * When \p oldstyle is set FALSE, the input and output satisfy a
87 * difference equation of the form
88
89 \f[
90 y[n] + \sum_{k=1}^{M} a_k y[n-k] = \sum_{k=0}^{N} b_k x[n-k]
91 \f]
92
93 * with the corresponding rational system function
94
95 \f[
96 H(z) = \frac{\sum_{k=0}^{N} b_k z^{-k}}{1 + \sum_{k=1}^{M} a_k z^{-k}}
97 \f]
98 * where:
99 * \f$x\f$ - input signal,
100 * \f$y\f$ - output signal,
101 * \f$a_k\f$ - k-th feedback tap,
102 * \f$b_k\f$ - k-th feed-forward tap,
103 * \f$M\f$ - \p len(fbtaps)-1,
104 * \f$N\f$ - \p len(fftaps)-1.
105
106 * \f$a_0\f$, that is \p fbtaps[0], is ignored.
107 */
108 iir_filter(const std::vector<tap_type>& fftaps,
109 const std::vector<tap_type>& fbtaps,
110 bool oldstyle = true) noexcept(false)
111 {
112 d_oldstyle = oldstyle;
113 set_taps(fftaps, fbtaps);
114 }
115
116 iir_filter() : d_latest_n(0), d_latest_m(0) {}
117
118 /*!
119 * \brief compute a single output value.
120 * \returns the filtered input value.
121 */
122 o_type filter(const i_type input);
123
124 /*!
125 * \brief compute an array of N output values.
126 * \p input must have N valid entries.
127 */
128 void filter_n(o_type output[], const i_type input[], long n);
129
130 /*!
131 * \return number of taps in filter.
132 */
133 unsigned ntaps_ff() const { return d_fftaps.size(); }
134 unsigned ntaps_fb() const { return d_fbtaps.size(); }
135
136 /*!
137 * \brief install new taps.
138 */
139 void set_taps(const std::vector<tap_type>& fftaps,
140 const std::vector<tap_type>& fbtaps)
141 {
142 d_latest_n = 0;
143 d_latest_m = 0;
144 d_fftaps = fftaps;
145
146 if (d_oldstyle) {
147 d_fbtaps = fbtaps;
148 } else {
149 // New style negates taps a[1:N-1] to fit with most IIR
150 // tap generating programs.
151 d_fbtaps.resize(fbtaps.size());
152 d_fbtaps[0] = fbtaps[0];
153 for (size_t i = 1; i < fbtaps.size(); i++) {
154 d_fbtaps[i] = -fbtaps[i];
155 }
156 }
157
158 int n = fftaps.size();
159 int m = fbtaps.size();
160 d_prev_input.clear();
161 d_prev_output.clear();
162 d_prev_input.resize(2 * n, 0);
163 d_prev_output.resize(2 * m, 0);
164 }
165
166protected:
168 std::vector<tap_type> d_fftaps;
169 std::vector<tap_type> d_fbtaps;
172 std::vector<acc_type> d_prev_output;
173 std::vector<i_type> d_prev_input;
174};
175
176//
177// general case. We may want to specialize this
178//
179template <class i_type, class o_type, class tap_type, class acc_type>
181{
182 acc_type acc;
183 unsigned i = 0;
184 unsigned n = ntaps_ff();
185 unsigned m = ntaps_fb();
186
187 if (n == 0)
188 return (o_type)0;
189
190 int latest_n = d_latest_n;
191 int latest_m = d_latest_m;
192
193 acc = d_fftaps[0] * input;
194 for (i = 1; i < n; i++)
195 acc += (d_fftaps[i] * d_prev_input[latest_n + i]);
196 for (i = 1; i < m; i++)
197 acc += (d_fbtaps[i] * d_prev_output[latest_m + i]);
198
199 // store the values twice to avoid having to handle wrap-around in the loop
200 d_prev_output[latest_m] = acc;
201 d_prev_output[latest_m + m] = acc;
202 d_prev_input[latest_n] = input;
203 d_prev_input[latest_n + n] = input;
204
205 latest_n--;
206 latest_m--;
207 if (latest_n < 0)
208 latest_n += n;
209 if (latest_m < 0)
210 latest_m += m;
211
212 d_latest_m = latest_m;
213 d_latest_n = latest_n;
214 return (o_type)acc;
215}
216
217template <class i_type, class o_type, class tap_type, class acc_type>
219 const i_type input[],
220 long n)
221{
222 for (int i = 0; i < n; i++)
223 output[i] = filter(input[i]);
224}
225
226template <>
229
230template <>
233
234template <>
236 const gr_complex input);
237
238} /* namespace kernel */
239} /* namespace filter */
240} /* namespace gr */
241
242#endif /* INCLUDED_IIR_FILTER_H */
Base class template for Infinite Impulse Response filter (IIR)
Definition: iir_filter.h:71
unsigned ntaps_ff() const
Definition: iir_filter.h:133
o_type filter(const i_type input)
compute a single output value.
Definition: iir_filter.h:180
std::vector< i_type > d_prev_input
Definition: iir_filter.h:173
std::vector< tap_type > d_fbtaps
Definition: iir_filter.h:169
unsigned ntaps_fb() const
Definition: iir_filter.h:134
std::vector< acc_type > d_prev_output
Definition: iir_filter.h:172
int d_latest_n
Definition: iir_filter.h:170
iir_filter()
Definition: iir_filter.h:116
void set_taps(const std::vector< tap_type > &fftaps, const std::vector< tap_type > &fbtaps)
install new taps.
Definition: iir_filter.h:139
std::vector< tap_type > d_fftaps
Definition: iir_filter.h:168
iir_filter(const std::vector< tap_type > &fftaps, const std::vector< tap_type > &fbtaps, bool oldstyle=true) noexcept(false)
Construct an IIR with the given taps.
Definition: iir_filter.h:108
void filter_n(o_type output[], const i_type input[], long n)
compute an array of N output values. input must have N valid entries.
Definition: iir_filter.h:218
int d_latest_m
Definition: iir_filter.h:171
bool d_oldstyle
Definition: iir_filter.h:167
#define FILTER_API
Definition: gr-filter/include/gnuradio/filter/api.h:18
std::complex< float > gr_complex
Definition: gr_complex.h:15
GNU Radio logging wrapper.
Definition: basic_block.h:29