V4.0.1 of the ARM CMSIS DSP libraries. Note that arm_bitreversal2.s, arm_cfft_f32.c and arm_rfft_fast_f32.c had to be removed. arm_bitreversal2.s will not assemble with the online tools. So, the fast f32 FFT functions are not yet available. All the other FFT functions are available.

Dependents:   MPU9150_Example fir_f32 fir_f32 MPU9150_nucleo_noni2cdev ... more

Embed: (wiki syntax)

« Back to documentation index

Show/hide line numbers arm_biquad_cascade_df1_q31.c Source File

arm_biquad_cascade_df1_q31.c

00001 /* ----------------------------------------------------------------------    
00002 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.    
00003 *    
00004 * $Date:        12. March 2014
00005 * $Revision:    V1.4.3
00006 *    
00007 * Project:      CMSIS DSP Library    
00008 * Title:        arm_biquad_cascade_df1_q31.c    
00009 *    
00010 * Description:  Processing function for the    
00011 *               Q31 Biquad cascade filter    
00012 *    
00013 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
00014 *  
00015 * Redistribution and use in source and binary forms, with or without 
00016 * modification, are permitted provided that the following conditions
00017 * are met:
00018 *   - Redistributions of source code must retain the above copyright
00019 *     notice, this list of conditions and the following disclaimer.
00020 *   - Redistributions in binary form must reproduce the above copyright
00021 *     notice, this list of conditions and the following disclaimer in
00022 *     the documentation and/or other materials provided with the 
00023 *     distribution.
00024 *   - Neither the name of ARM LIMITED nor the names of its contributors
00025 *     may be used to endorse or promote products derived from this
00026 *     software without specific prior written permission.
00027 *
00028 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
00029 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
00030 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
00031 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
00032 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
00033 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
00034 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
00035 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
00036 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
00037 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
00038 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
00039 * POSSIBILITY OF SUCH DAMAGE.     
00040 * -------------------------------------------------------------------- */
00041 
00042 #include "arm_math.h"
00043 
00044 /**    
00045  * @ingroup groupFilters    
00046  */
00047 
00048 /**    
00049  * @addtogroup BiquadCascadeDF1    
00050  * @{    
00051  */
00052 
00053 /**    
00054  * @brief Processing function for the Q31 Biquad cascade filter.    
00055  * @param[in]  *S         points to an instance of the Q31 Biquad cascade structure.    
00056  * @param[in]  *pSrc      points to the block of input data.    
00057  * @param[out] *pDst      points to the block of output data.    
00058  * @param[in]  blockSize  number of samples to process per call.    
00059  * @return none.    
00060  *    
00061  * <b>Scaling and Overflow Behavior:</b>    
00062  * \par    
00063  * The function is implemented using an internal 64-bit accumulator.    
00064  * The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit.    
00065  * Thus, if the accumulator result overflows it wraps around rather than clip.    
00066  * In order to avoid overflows completely the input signal must be scaled down by 2 bits and lie in the range [-0.25 +0.25).    
00067  * After all 5 multiply-accumulates are performed, the 2.62 accumulator is shifted by <code>postShift</code> bits and the result truncated to    
00068  * 1.31 format by discarding the low 32 bits.    
00069  *    
00070  * \par    
00071  * Refer to the function <code>arm_biquad_cascade_df1_fast_q31()</code> for a faster but less precise implementation of this filter for Cortex-M3 and Cortex-M4.    
00072  */
00073 
00074 void arm_biquad_cascade_df1_q31(
00075   const arm_biquad_casd_df1_inst_q31 * S,
00076   q31_t * pSrc,
00077   q31_t * pDst,
00078   uint32_t blockSize)
00079 {
00080   q63_t acc;                                     /*  accumulator                   */
00081   uint32_t uShift = ((uint32_t) S->postShift + 1u);
00082   uint32_t lShift = 32u - uShift;                /*  Shift to be applied to the output */
00083   q31_t *pIn = pSrc;                             /*  input pointer initialization  */
00084   q31_t *pOut = pDst;                            /*  output pointer initialization */
00085   q31_t *pState = S->pState;                     /*  pState pointer initialization */
00086   q31_t *pCoeffs = S->pCoeffs;                   /*  coeff pointer initialization  */
00087   q31_t Xn1, Xn2, Yn1, Yn2;                      /*  Filter state variables        */
00088   q31_t b0, b1, b2, a1, a2;                      /*  Filter coefficients           */
00089   q31_t Xn;                                      /*  temporary input               */
00090   uint32_t sample, stage = S->numStages;         /*  loop counters                     */
00091 
00092 
00093 #ifndef ARM_MATH_CM0_FAMILY_FAMILY
00094 
00095   q31_t acc_l, acc_h;                            /*  temporary output variables    */
00096 
00097   /* Run the below code for Cortex-M4 and Cortex-M3 */
00098 
00099   do
00100   {
00101     /* Reading the coefficients */
00102     b0 = *pCoeffs++;
00103     b1 = *pCoeffs++;
00104     b2 = *pCoeffs++;
00105     a1 = *pCoeffs++;
00106     a2 = *pCoeffs++;
00107 
00108     /* Reading the state values */
00109     Xn1 = pState[0];
00110     Xn2 = pState[1];
00111     Yn1 = pState[2];
00112     Yn2 = pState[3];
00113 
00114     /* Apply loop unrolling and compute 4 output values simultaneously. */
00115     /*      The variable acc hold output values that are being computed:    
00116      *    
00117      *    acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2]    
00118      */
00119 
00120     sample = blockSize >> 2u;
00121 
00122     /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.    
00123      ** a second loop below computes the remaining 1 to 3 samples. */
00124     while(sample > 0u)
00125     {
00126       /* Read the input */
00127       Xn = *pIn++;
00128 
00129       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00130 
00131       /* acc =  b0 * x[n] */
00132       acc = (q63_t) b0 *Xn;
00133       /* acc +=  b1 * x[n-1] */
00134       acc += (q63_t) b1 *Xn1;
00135       /* acc +=  b[2] * x[n-2] */
00136       acc += (q63_t) b2 *Xn2;
00137       /* acc +=  a1 * y[n-1] */
00138       acc += (q63_t) a1 *Yn1;
00139       /* acc +=  a2 * y[n-2] */
00140       acc += (q63_t) a2 *Yn2;
00141 
00142       /* The result is converted to 1.31 , Yn2 variable is reused */
00143 
00144       /* Calc lower part of acc */
00145       acc_l = acc & 0xffffffff;
00146 
00147       /* Calc upper part of acc */
00148       acc_h = (acc >> 32) & 0xffffffff;
00149 
00150       /* Apply shift for lower part of acc and upper part of acc */
00151       Yn2 = (uint32_t) acc_l >> lShift | acc_h << uShift;
00152 
00153       /* Store the output in the destination buffer. */
00154       *pOut++ = Yn2;
00155 
00156       /* Read the second input */
00157       Xn2 = *pIn++;
00158 
00159       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00160 
00161       /* acc =  b0 * x[n] */
00162       acc = (q63_t) b0 *Xn2;
00163       /* acc +=  b1 * x[n-1] */
00164       acc += (q63_t) b1 *Xn;
00165       /* acc +=  b[2] * x[n-2] */
00166       acc += (q63_t) b2 *Xn1;
00167       /* acc +=  a1 * y[n-1] */
00168       acc += (q63_t) a1 *Yn2;
00169       /* acc +=  a2 * y[n-2] */
00170       acc += (q63_t) a2 *Yn1;
00171 
00172 
00173       /* The result is converted to 1.31, Yn1 variable is reused  */
00174 
00175       /* Calc lower part of acc */
00176       acc_l = acc & 0xffffffff;
00177 
00178       /* Calc upper part of acc */
00179       acc_h = (acc >> 32) & 0xffffffff;
00180 
00181 
00182       /* Apply shift for lower part of acc and upper part of acc */
00183       Yn1 = (uint32_t) acc_l >> lShift | acc_h << uShift;
00184 
00185       /* Store the output in the destination buffer. */
00186       *pOut++ = Yn1;
00187 
00188       /* Read the third input  */
00189       Xn1 = *pIn++;
00190 
00191       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00192 
00193       /* acc =  b0 * x[n] */
00194       acc = (q63_t) b0 *Xn1;
00195       /* acc +=  b1 * x[n-1] */
00196       acc += (q63_t) b1 *Xn2;
00197       /* acc +=  b[2] * x[n-2] */
00198       acc += (q63_t) b2 *Xn;
00199       /* acc +=  a1 * y[n-1] */
00200       acc += (q63_t) a1 *Yn1;
00201       /* acc +=  a2 * y[n-2] */
00202       acc += (q63_t) a2 *Yn2;
00203 
00204       /* The result is converted to 1.31, Yn2 variable is reused  */
00205       /* Calc lower part of acc */
00206       acc_l = acc & 0xffffffff;
00207 
00208       /* Calc upper part of acc */
00209       acc_h = (acc >> 32) & 0xffffffff;
00210 
00211 
00212       /* Apply shift for lower part of acc and upper part of acc */
00213       Yn2 = (uint32_t) acc_l >> lShift | acc_h << uShift;
00214 
00215       /* Store the output in the destination buffer. */
00216       *pOut++ = Yn2;
00217 
00218       /* Read the forth input */
00219       Xn = *pIn++;
00220 
00221       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00222 
00223       /* acc =  b0 * x[n] */
00224       acc = (q63_t) b0 *Xn;
00225       /* acc +=  b1 * x[n-1] */
00226       acc += (q63_t) b1 *Xn1;
00227       /* acc +=  b[2] * x[n-2] */
00228       acc += (q63_t) b2 *Xn2;
00229       /* acc +=  a1 * y[n-1] */
00230       acc += (q63_t) a1 *Yn2;
00231       /* acc +=  a2 * y[n-2] */
00232       acc += (q63_t) a2 *Yn1;
00233 
00234       /* The result is converted to 1.31, Yn1 variable is reused  */
00235       /* Calc lower part of acc */
00236       acc_l = acc & 0xffffffff;
00237 
00238       /* Calc upper part of acc */
00239       acc_h = (acc >> 32) & 0xffffffff;
00240 
00241       /* Apply shift for lower part of acc and upper part of acc */
00242       Yn1 = (uint32_t) acc_l >> lShift | acc_h << uShift;
00243 
00244       /* Every time after the output is computed state should be updated. */
00245       /* The states should be updated as:  */
00246       /* Xn2 = Xn1    */
00247       /* Xn1 = Xn     */
00248       /* Yn2 = Yn1    */
00249       /* Yn1 = acc    */
00250       Xn2 = Xn1;
00251       Xn1 = Xn;
00252 
00253       /* Store the output in the destination buffer. */
00254       *pOut++ = Yn1;
00255 
00256       /* decrement the loop counter */
00257       sample--;
00258     }
00259 
00260     /* If the blockSize is not a multiple of 4, compute any remaining output samples here.    
00261      ** No loop unrolling is used. */
00262     sample = (blockSize & 0x3u);
00263 
00264     while(sample > 0u)
00265     {
00266       /* Read the input */
00267       Xn = *pIn++;
00268 
00269       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00270 
00271       /* acc =  b0 * x[n] */
00272       acc = (q63_t) b0 *Xn;
00273       /* acc +=  b1 * x[n-1] */
00274       acc += (q63_t) b1 *Xn1;
00275       /* acc +=  b[2] * x[n-2] */
00276       acc += (q63_t) b2 *Xn2;
00277       /* acc +=  a1 * y[n-1] */
00278       acc += (q63_t) a1 *Yn1;
00279       /* acc +=  a2 * y[n-2] */
00280       acc += (q63_t) a2 *Yn2;
00281 
00282       /* The result is converted to 1.31  */
00283       acc = acc >> lShift;
00284 
00285       /* Every time after the output is computed state should be updated. */
00286       /* The states should be updated as:  */
00287       /* Xn2 = Xn1    */
00288       /* Xn1 = Xn     */
00289       /* Yn2 = Yn1    */
00290       /* Yn1 = acc    */
00291       Xn2 = Xn1;
00292       Xn1 = Xn;
00293       Yn2 = Yn1;
00294       Yn1 = (q31_t) acc;
00295 
00296       /* Store the output in the destination buffer. */
00297       *pOut++ = (q31_t) acc;
00298 
00299       /* decrement the loop counter */
00300       sample--;
00301     }
00302 
00303     /*  The first stage goes from the input buffer to the output buffer. */
00304     /*  Subsequent stages occur in-place in the output buffer */
00305     pIn = pDst;
00306 
00307     /* Reset to destination pointer */
00308     pOut = pDst;
00309 
00310     /*  Store the updated state variables back into the pState array */
00311     *pState++ = Xn1;
00312     *pState++ = Xn2;
00313     *pState++ = Yn1;
00314     *pState++ = Yn2;
00315 
00316   } while(--stage);
00317 
00318 #else
00319 
00320   /* Run the below code for Cortex-M0 */
00321 
00322   do
00323   {
00324     /* Reading the coefficients */
00325     b0 = *pCoeffs++;
00326     b1 = *pCoeffs++;
00327     b2 = *pCoeffs++;
00328     a1 = *pCoeffs++;
00329     a2 = *pCoeffs++;
00330 
00331     /* Reading the state values */
00332     Xn1 = pState[0];
00333     Xn2 = pState[1];
00334     Yn1 = pState[2];
00335     Yn2 = pState[3];
00336 
00337     /*      The variables acc holds the output value that is computed:         
00338      *    acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2]         
00339      */
00340 
00341     sample = blockSize;
00342 
00343     while(sample > 0u)
00344     {
00345       /* Read the input */
00346       Xn = *pIn++;
00347 
00348       /* acc =  b0 * x[n] + b1 * x[n-1] + b2 * x[n-2] + a1 * y[n-1] + a2 * y[n-2] */
00349       /* acc =  b0 * x[n] */
00350       acc = (q63_t) b0 *Xn;
00351 
00352       /* acc +=  b1 * x[n-1] */
00353       acc += (q63_t) b1 *Xn1;
00354       /* acc +=  b[2] * x[n-2] */
00355       acc += (q63_t) b2 *Xn2;
00356       /* acc +=  a1 * y[n-1] */
00357       acc += (q63_t) a1 *Yn1;
00358       /* acc +=  a2 * y[n-2] */
00359       acc += (q63_t) a2 *Yn2;
00360 
00361       /* The result is converted to 1.31  */
00362       acc = acc >> lShift;
00363 
00364       /* Every time after the output is computed state should be updated. */
00365       /* The states should be updated as:  */
00366       /* Xn2 = Xn1    */
00367       /* Xn1 = Xn     */
00368       /* Yn2 = Yn1    */
00369       /* Yn1 = acc    */
00370       Xn2 = Xn1;
00371       Xn1 = Xn;
00372       Yn2 = Yn1;
00373       Yn1 = (q31_t) acc;
00374 
00375       /* Store the output in the destination buffer. */
00376       *pOut++ = (q31_t) acc;
00377 
00378       /* decrement the loop counter */
00379       sample--;
00380     }
00381 
00382     /*  The first stage goes from the input buffer to the output buffer. */
00383     /*  Subsequent stages occur in-place in the output buffer */
00384     pIn = pDst;
00385 
00386     /* Reset to destination pointer */
00387     pOut = pDst;
00388 
00389     /*  Store the updated state variables back into the pState array */
00390     *pState++ = Xn1;
00391     *pState++ = Xn2;
00392     *pState++ = Yn1;
00393     *pState++ = Yn2;
00394 
00395   } while(--stage);
00396 
00397 #endif /*  #ifndef ARM_MATH_CM0_FAMILY_FAMILY */
00398 }
00399 
00400 
00401 
00402 
00403 /**    
00404   * @} end of BiquadCascadeDF1 group    
00405   */