arm_cmplx_mult_real_f32.c

00001 /* ----------------------------------------------------------------------    
00002 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.    
00003 *    
00004 * $Date:        19. March 2015
00005 * $Revision:    V.1.4.5
00006 *    
00007 * Project:      CMSIS DSP Library    
00008 * Title:        arm_cmplx_mult_real_f32.c    
00009 *    
00010 * Description:  Floating-point complex by real multiplication    
00011 *    
00012 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
00013 *  
00014 * Redistribution and use in source and binary forms, with or without 
00015 * modification, are permitted provided that the following conditions
00016 * are met:
00017 *   - Redistributions of source code must retain the above copyright
00018 *     notice, this list of conditions and the following disclaimer.
00019 *   - Redistributions in binary form must reproduce the above copyright
00020 *     notice, this list of conditions and the following disclaimer in
00021 *     the documentation and/or other materials provided with the 
00022 *     distribution.
00023 *   - Neither the name of ARM LIMITED nor the names of its contributors
00024 *     may be used to endorse or promote products derived from this
00025 *     software without specific prior written permission.
00026 *
00027 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
00028 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
00029 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
00030 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
00031 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
00032 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
00033 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
00034 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
00035 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
00036 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
00037 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
00038 * POSSIBILITY OF SUCH DAMAGE.  
00039 * -------------------------------------------------------------------- */
00040 
00041 #include "arm_math.h"
00042 
00043 /**        
00044  * @ingroup groupCmplxMath        
00045  */
00046 
00047 /**        
00048  * @defgroup CmplxByRealMult Complex-by-Real Multiplication        
00049  *        
00050  * Multiplies a complex vector by a real vector and generates a complex result.        
00051  * The data in the complex arrays is stored in an interleaved fashion        
00052  * (real, imag, real, imag, ...).        
00053  * The parameter <code>numSamples</code> represents the number of complex        
00054  * samples processed.  The complex arrays have a total of <code>2*numSamples</code>        
00055  * real values while the real array has a total of <code>numSamples</code>        
00056  * real values.        
00057  *        
00058  * The underlying algorithm is used:        
00059  *        
00060  * <pre>        
00061  * for(n=0; n<numSamples; n++) {        
00062  *     pCmplxDst[(2*n)+0] = pSrcCmplx[(2*n)+0] * pSrcReal[n];        
00063  *     pCmplxDst[(2*n)+1] = pSrcCmplx[(2*n)+1] * pSrcReal[n];        
00064  * }        
00065  * </pre>        
00066  *        
00067  * There are separate functions for floating-point, Q15, and Q31 data types.        
00068  */
00069 
00070 /**        
00071  * @addtogroup CmplxByRealMult        
00072  * @{        
00073  */
00074 
00075 
00076 /**        
00077  * @brief  Floating-point complex-by-real multiplication        
00078  * @param[in]  *pSrcCmplx points to the complex input vector        
00079  * @param[in]  *pSrcReal points to the real input vector        
00080  * @param[out]  *pCmplxDst points to the complex output vector        
00081  * @param[in]  numSamples number of samples in each vector        
00082  * @return none.        
00083  */
00084 
00085 void arm_cmplx_mult_real_f32(
00086   float32_t * pSrcCmplx,
00087   float32_t * pSrcReal,
00088   float32_t * pCmplxDst,
00089   uint32_t numSamples)
00090 {
00091   float32_t in;                                  /* Temporary variable to store input value */
00092   uint32_t blkCnt;                               /* loop counters */
00093 
00094 #ifndef ARM_MATH_CM0_FAMILY
00095 
00096   /* Run the below code for Cortex-M4 and Cortex-M3 */
00097   float32_t inA1, inA2, inA3, inA4;              /* Temporary variables to hold input data */
00098   float32_t inA5, inA6, inA7, inA8;              /* Temporary variables to hold input data */
00099   float32_t inB1, inB2, inB3, inB4;              /* Temporary variables to hold input data */
00100   float32_t out1, out2, out3, out4;              /* Temporary variables to hold output data */
00101   float32_t out5, out6, out7, out8;              /* Temporary variables to hold output data */
00102 
00103   /* loop Unrolling */
00104   blkCnt = numSamples >> 2u;
00105 
00106   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.        
00107    ** a second loop below computes the remaining 1 to 3 samples. */
00108   while(blkCnt > 0u)
00109   {
00110     /* C[2 * i] = A[2 * i] * B[i].            */
00111     /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
00112     /* read input from complex input buffer */
00113     inA1 = pSrcCmplx[0];
00114     inA2 = pSrcCmplx[1];
00115     /* read input from real input buffer */
00116     inB1 = pSrcReal[0];
00117 
00118     /* read input from complex input buffer */
00119     inA3 = pSrcCmplx[2];
00120 
00121     /* multiply complex buffer real input with real buffer input */
00122     out1 = inA1 * inB1;
00123 
00124     /* read input from complex input buffer */
00125     inA4 = pSrcCmplx[3];
00126 
00127     /* multiply complex buffer imaginary input with real buffer input */
00128     out2 = inA2 * inB1;
00129 
00130     /* read input from real input buffer */
00131     inB2 = pSrcReal[1];
00132     /* read input from complex input buffer */
00133     inA5 = pSrcCmplx[4];
00134 
00135     /* multiply complex buffer real input with real buffer input */
00136     out3 = inA3 * inB2;
00137 
00138     /* read input from complex input buffer */
00139     inA6 = pSrcCmplx[5];
00140     /* read input from real input buffer */
00141     inB3 = pSrcReal[2];
00142 
00143     /* multiply complex buffer imaginary input with real buffer input */
00144     out4 = inA4 * inB2;
00145 
00146     /* read input from complex input buffer */
00147     inA7 = pSrcCmplx[6];
00148 
00149     /* multiply complex buffer real input with real buffer input */
00150     out5 = inA5 * inB3;
00151 
00152     /* read input from complex input buffer */
00153     inA8 = pSrcCmplx[7];
00154 
00155     /* multiply complex buffer imaginary input with real buffer input */
00156     out6 = inA6 * inB3;
00157 
00158     /* read input from real input buffer */
00159     inB4 = pSrcReal[3];
00160 
00161     /* store result to destination bufer */
00162     pCmplxDst[0] = out1;
00163 
00164     /* multiply complex buffer real input with real buffer input */
00165     out7 = inA7 * inB4;
00166 
00167     /* store result to destination bufer */
00168     pCmplxDst[1] = out2;
00169 
00170     /* multiply complex buffer imaginary input with real buffer input */
00171     out8 = inA8 * inB4;
00172 
00173     /* store result to destination bufer */
00174     pCmplxDst[2] = out3;
00175     pCmplxDst[3] = out4;
00176     pCmplxDst[4] = out5;
00177 
00178     /* incremnet complex input buffer by 8 to process next samples */
00179     pSrcCmplx += 8u;
00180 
00181     /* store result to destination bufer */
00182     pCmplxDst[5] = out6;
00183 
00184     /* increment real input buffer by 4 to process next samples */
00185     pSrcReal += 4u;
00186 
00187     /* store result to destination bufer */
00188     pCmplxDst[6] = out7;
00189     pCmplxDst[7] = out8;
00190 
00191     /* increment destination buffer by 8 to process next sampels */
00192     pCmplxDst += 8u;
00193 
00194     /* Decrement the numSamples loop counter */
00195     blkCnt--;
00196   }
00197 
00198   /* If the numSamples is not a multiple of 4, compute any remaining output samples here.        
00199    ** No loop unrolling is used. */
00200   blkCnt = numSamples % 0x4u;
00201 
00202 #else
00203 
00204   /* Run the below code for Cortex-M0 */
00205   blkCnt = numSamples;
00206 
00207 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
00208 
00209   while(blkCnt > 0u)
00210   {
00211     /* C[2 * i] = A[2 * i] * B[i].            */
00212     /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
00213     in = *pSrcReal++;
00214     /* store the result in the destination buffer. */
00215     *pCmplxDst++ = (*pSrcCmplx++) * (in);
00216     *pCmplxDst++ = (*pSrcCmplx++) * (in);
00217 
00218     /* Decrement the numSamples loop counter */
00219     blkCnt--;
00220   }
00221 }
00222 
00223 /**        
00224  * @} end of CmplxByRealMult group        
00225  */