arm_cmplx_dot_prod_q15.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_q15.c  
00009 *  
00010 * Description:  Processing function for the Q15 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  Q15 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 1.15 by 1.15 multiplications are performed with full precision and yield a 2.30 result.  
00051  * These are accumulated in a 64-bit accumulator with 34.30 precision.  
00052  * As a final step, the accumulators are converted to 8.24 format.  
00053  * The return results <code>realResult</code> and <code>imagResult</code> are in 8.24 format.  
00054  */ 
00055  
00056 void arm_cmplx_dot_prod_q15( 
00057   q15_t * pSrcA, 
00058   q15_t * pSrcB, 
00059   uint32_t numSamples, 
00060   q31_t * realResult, 
00061   q31_t * imagResult) 
00062 { 
00063   q63_t real_sum = 0, imag_sum = 0;              /* Temporary result storage */ 
00064   uint32_t blkCnt;                               /* loop counter */ 
00065  
00066  
00067   /*loop Unrolling */ 
00068   blkCnt = numSamples >> 2u; 
00069  
00070   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.  
00071    ** a second loop below computes the remaining 1 to 3 samples. */ 
00072   while(blkCnt > 0u) 
00073   { 
00074     /* CReal = A[0]* B[0] + A[2]* B[2] + A[4]* B[4] + .....+ A[numSamples-2]* B[numSamples-2] */ 
00075     real_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00076  
00077     /* CImag = A[1]* B[1] + A[3]* B[3] + A[5]* B[5] + .....+ A[numSamples-1]* B[numSamples-1] */ 
00078     imag_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00079  
00080     real_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00081     imag_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00082  
00083     real_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00084     imag_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00085  
00086     real_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00087     imag_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00088  
00089     /* Decrement the loop counter */ 
00090     blkCnt--; 
00091   } 
00092  
00093   /* If the numSamples is not a multiple of 4, compute any remaining output samples here.  
00094    ** No loop unrolling is used. */ 
00095   blkCnt = numSamples % 0x4u; 
00096  
00097   while(blkCnt > 0u) 
00098   { 
00099     /* CReal = A[0]* B[0] + A[2]* B[2] + A[4]* B[4] + .....+ A[numSamples-2]* B[numSamples-2] */ 
00100     real_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00101     /* CImag = A[1]* B[1] + A[3]* B[3] + A[5]* B[5] + .....+ A[numSamples-1]* B[numSamples-1] */ 
00102     imag_sum += ((q31_t) * pSrcA++ * *pSrcB++); 
00103  
00104     /* Decrement the loop counter */ 
00105     blkCnt--; 
00106   } 
00107  
00108   /* Store the real and imaginary results in 8.24 format  */ 
00109   /* Convert real data in 34.30 to 8.24 by 6 right shifts */ 
00110   *realResult = (q31_t) (real_sum) >> 6; 
00111   /* Convert imaginary data in 34.30 to 8.24 by 6 right shifts */ 
00112   *imagResult = (q31_t) (imag_sum) >> 6; 
00113 } 
00114  
00115 /**  
00116  * @} end of cmplx_dot_prod group  
00117  */