arm_cmplx_dot_prod_q31.c

00001 /* ----------------------------------------------------------------------  
00002 * Copyright (C) 2010 ARM Limited. All rights reserved.  
00003 *  
00004 * $Date:        29. November 2010  
00005 * $Revision:    V1.0.3  
00006 *  
00007 * Project:      CMSIS DSP Library  
00008 * Title:        arm_cmplx_dot_prod_q31.c  
00009 *  
00010 * Description:  Q31 complex dot product  
00011 *  
00012 * Target Processor: Cortex-M4/Cortex-M3
00013 *  
00014 * Version 1.0.3 2010/11/29 
00015 *    Re-organized the CMSIS folders and updated documentation.  
00016 *   
00017 * Version 1.0.2 2010/11/11  
00018 *    Documentation updated.   
00019 *  
00020 * Version 1.0.1 2010/10/05   
00021 *    Production release and review comments incorporated.  
00022 *  
00023 * Version 1.0.0 2010/09/20   
00024 *    Production release and review comments incorporated.  
00025 * -------------------------------------------------------------------- */ 
00026  
00027 #include "arm_math.h" 
00028  
00029 /**  
00030  * @ingroup groupCmplxMath  
00031  */ 
00032  
00033 /**  
00034  * @addtogroup cmplx_dot_prod  
00035  * @{  
00036  */ 
00037  
00038 /**  
00039  * @brief  Q31 complex dot product  
00040  * @param  *pSrcA points to the first input vector  
00041  * @param  *pSrcB points to the second input vector  
00042  * @param  numSamples number of complex samples in each vector  
00043  * @param  *realResult real part of the result returned here  
00044  * @param  *imagResult imaginary part of the result returned here  
00045  * @return none.  
00046  *  
00047  * <b>Scaling and Overflow Behavior:</b>  
00048  * \par  
00049  * The function is implemented using an internal 64-bit accumulator.  
00050  * The intermediate 1.31 by 1.31 multiplications are performed with 64-bit precision and then shifted to 16.48 format.  
00051  * The internal real and imaginary accumulators are in 16.48 format and provide 15 guard bits.  
00052  * Additions are nonsaturating and no overflow will occur as long as <code>numSamples</code> is less than 32768.  
00053  * The return results <code>realResult</code> and <code>imagResult</code> are in 16.48 format.  
00054  * Input down scaling is not required.  
00055  */ 
00056  
00057 void arm_cmplx_dot_prod_q31( 
00058   q31_t * pSrcA, 
00059   q31_t * pSrcB, 
00060   uint32_t numSamples, 
00061   q63_t * realResult, 
00062   q63_t * imagResult) 
00063 { 
00064   q63_t real_sum = 0, imag_sum = 0;              /* Temporary result storage */ 
00065   uint32_t blkCnt;                               /* loop counter */ 
00066  
00067  
00068   /*loop Unrolling */ 
00069   blkCnt = numSamples >> 2u; 
00070  
00071   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.  
00072    ** a second loop below computes the remaining 1 to 3 samples. */ 
00073   while(blkCnt > 0u) 
00074   { 
00075     /* CReal = A[0]* B[0] + A[2]* B[2] + A[4]* B[4] + .....+ A[numSamples-2]* B[numSamples-2] */ 
00076     /* Convert real data in 2.62 to 16.48 by 14 right shifts */ 
00077     real_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00078     /* CImag = A[1]* B[1] + A[3]* B[3] + A[5]* B[5] + .....+ A[numSamples-1]* B[numSamples-1] */ 
00079     /* Convert imag data in 2.62 to 16.48 by 14 right shifts */ 
00080     imag_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00081  
00082     real_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00083     imag_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00084  
00085     real_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00086     imag_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00087  
00088     real_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00089     imag_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00090  
00091  
00092     /* Decrement the loop counter */ 
00093     blkCnt--; 
00094   } 
00095  
00096   /* If the numSamples  is not a multiple of 4, compute any remaining output samples here.  
00097    ** No loop unrolling is used. */ 
00098   blkCnt = numSamples % 0x4u; 
00099  
00100   while(blkCnt > 0u) 
00101   { 
00102     /* CReal = A[0]* B[0] + A[2]* B[2] + A[4]* B[4] + .....+ A[numSamples-2]* B[numSamples-2] */ 
00103     real_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00104     /* CImag = A[1]* B[1] + A[3]* B[3] + A[5]* B[5] + .....+ A[numSamples-1]* B[numSamples-1] */ 
00105     imag_sum += (q63_t) * pSrcA++ * (*pSrcB++) >> 14; 
00106  
00107     /* Decrement the loop counter */ 
00108     blkCnt--; 
00109   } 
00110  
00111   /* Store the real and imaginary results in 16.48 format  */ 
00112   *realResult = real_sum; 
00113   *imagResult = imag_sum; 
00114 } 
00115  
00116 /**  
00117  * @} end of cmplx_dot_prod group  
00118  */