GNU Radio 3.4.2 C++ API
|
00001 /* -------------------------------------------------------------- */ 00002 /* (C)Copyright 2008 Free Software Foundation, Inc. */ 00003 /* (C)Copyright 2001,2007, */ 00004 /* International Business Machines Corporation, */ 00005 /* Sony Computer Entertainment, Incorporated, */ 00006 /* Toshiba Corporation, */ 00007 /* */ 00008 /* All Rights Reserved. */ 00009 /* */ 00010 /* Redistribution and use in source and binary forms, with or */ 00011 /* without modification, are permitted provided that the */ 00012 /* following conditions are met: */ 00013 /* */ 00014 /* - Redistributions of source code must retain the above copyright*/ 00015 /* notice, this list of conditions and the following disclaimer. */ 00016 /* */ 00017 /* - Redistributions in binary form must reproduce the above */ 00018 /* copyright notice, this list of conditions and the following */ 00019 /* disclaimer in the documentation and/or other materials */ 00020 /* provided with the distribution. */ 00021 /* */ 00022 /* - Neither the name of IBM Corporation nor the names of its */ 00023 /* contributors may be used to endorse or promote products */ 00024 /* derived from this software without specific prior written */ 00025 /* permission. */ 00026 /* */ 00027 /* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND */ 00028 /* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, */ 00029 /* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF */ 00030 /* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE */ 00031 /* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR */ 00032 /* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, */ 00033 /* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT */ 00034 /* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; */ 00035 /* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) */ 00036 /* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN */ 00037 /* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR */ 00038 /* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, */ 00039 /* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ 00040 /* -------------------------------------------------------------- */ 00041 /* PROLOG END TAG zYx */ 00042 00043 #ifndef INCLUDED_LIBFFT_H 00044 #define INCLUDED_LIBFFT_H 00045 00046 // must be defined before inclusion of fft_1d_r2.h 00047 #define MAX_FFT_1D_SIZE 4096 00048 00049 /* fft_1d_r2 00050 * --------- 00051 * Performs a single precision, complex Fast Fourier Transform using 00052 * the DFT (Discrete Fourier Transform) with radix-2 decimation in time. 00053 * The input <in> is an array of complex numbers of length (1<<log2_size) 00054 * entries. The result is returned in the array of complex numbers specified 00055 * by <out>. Note: This routine can support an in-place transformation 00056 * by specifying <in> and <out> to be the same array. 00057 * 00058 * This implementation utilizes the Cooley-Tukey algorithm consisting 00059 * of <log2_size> stages. The basic operation is the butterfly. 00060 * 00061 * p --------------------------> P = p + q*Wi 00062 * \ / 00063 * \ / 00064 * \ / 00065 * \/ 00066 * /\ 00067 * / \ 00068 * / \ 00069 * ____ / \ 00070 * q --| Wi |-----------------> Q = p - q*Wi 00071 * ---- 00072 * 00073 * This routine also requires pre-computed twiddle values, W. W is an 00074 * array of single precision complex numbers of length 1<<(log2_size-2) 00075 * and is computed as follows: 00076 * 00077 * for (i=0; i<n/4; i++) 00078 * W[i].real = cos(i * 2*PI/n); 00079 * W[i].imag = -sin(i * 2*PI/n); 00080 * } 00081 * 00082 * This array actually only contains the first half of the twiddle 00083 * factors. Due for symmetry, the second half of the twiddle factors 00084 * are implied and equal: 00085 * 00086 * for (i=0; i<n/4; i++) 00087 * W[i+n/4].real = W[i].imag = sin(i * 2*PI/n); 00088 * W[i+n/4].imag = -W[i].real = -cos(i * 2*PI/n); 00089 * } 00090 * 00091 * Further symmetry allows one to generate the twiddle factor table 00092 * using half the number of trig computations as follows: 00093 * 00094 * W[0].real = 1.0; 00095 * W[0].imag = 0.0; 00096 * for (i=1; i<n/4; i++) 00097 * W[i].real = cos(i * 2*PI/n); 00098 * W[n/4 - i].imag = -W[i].real; 00099 * } 00100 * 00101 * The complex numbers are packed into quadwords as follows: 00102 * 00103 * quadword complex 00104 * array element array elements 00105 * ----------------------------------------------------- 00106 * i | real 2*i | imag 2*i | real 2*i+1 | imag 2*i+1 | 00107 * ----------------------------------------------------- 00108 * 00109 */ 00110 00111 void fft_1d_r2(vector float *out, vector float *in, vector float *W, int log2_size); 00112 00113 #endif /* INCLUDED_LIBFFT_H */