tmk_core/tool/mbed/mbed-sdk/libraries/dsp/cmsis_dsp/StatisticsFunctions/arm_std_q15.c

   1 /* ----------------------------------------------------------------------
   2 * Copyright (C) 2010-2013 ARM Limited. All rights reserved.
   3 *
   4 * $Date:        17. January 2013
   5 * $Revision:    V1.4.1
   6 *
   7 * Project:          CMSIS DSP Library
   8 * Title:                arm_std_q15.c
   9 *
  10 * Description:  Standard deviation of an array of Q15 type.
  11 *
  12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
  13 *
  14 * Redistribution and use in source and binary forms, with or without
  15 * modification, are permitted provided that the following conditions
  16 * are met:
  17 *   - Redistributions of source code must retain the above copyright
  18 *     notice, this list of conditions and the following disclaimer.
  19 *   - Redistributions in binary form must reproduce the above copyright
  20 *     notice, this list of conditions and the following disclaimer in
  21 *     the documentation and/or other materials provided with the
  22 *     distribution.
  23 *   - Neither the name of ARM LIMITED nor the names of its contributors
  24 *     may be used to endorse or promote products derived from this
  25 *     software without specific prior written permission.
  26 *
  27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
  30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
  31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
  32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
  33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
  37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
  38 * POSSIBILITY OF SUCH DAMAGE.
  39 * -------------------------------------------------------------------- */
  40
  41 #include "arm_math.h"
  42
  43 /**
  44  * @ingroup groupStats
  45  */
  46
  47 /**
  48  * @addtogroup STD
  49  * @{
  50  */
  51
  52 /**
  53  * @brief Standard deviation of the elements of a Q15 vector.
  54  * @param[in]       *pSrc points to the input vector
  55  * @param[in]       blockSize length of the input vector
  56  * @param[out]      *pResult standard deviation value returned here
  57  * @return none.
  58  *
  59  * @details
  60  * <b>Scaling and Overflow Behavior:</b>
  61  *
  62  * \par
  63  * The function is implemented using a 64-bit internal accumulator.
  64  * The input is represented in 1.15 format.
  65  * Intermediate multiplication yields a 2.30 format, and this
  66  * result is added without saturation to a 64-bit accumulator in 34.30 format.
  67  * With 33 guard bits in the accumulator, there is no risk of overflow, and the
  68  * full precision of the intermediate multiplication is preserved.
  69  * Finally, the 34.30 result is truncated to 34.15 format by discarding the lower
  70  * 15 bits, and then saturated to yield a result in 1.15 format.
  71  */
  72
  73 void arm_std_q15(
  74   q15_t * pSrc,
  75   uint32_t blockSize,
  76   q15_t * pResult)
  77 {
  78   q31_t sum = 0;                                 /* Accumulator */
  79   q31_t meanOfSquares, squareOfMean;             /* square of mean and mean of square */
  80   q15_t mean;                                    /* mean */
  81   uint32_t blkCnt;                               /* loop counter */
  82   q15_t t;                                       /* Temporary variable */
  83   q63_t sumOfSquares = 0;                        /* Accumulator */
  84
  85 #ifndef ARM_MATH_CM0_FAMILY
  86
  87   /* Run the below code for Cortex-M4 and Cortex-M3 */
  88
  89   q31_t in;                                      /* input value */
  90   q15_t in1;                                     /* input value */
  91
  92   /*loop Unrolling */
  93   blkCnt = blockSize >> 2u;
  94
  95   /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
  96    ** a second loop below computes the remaining 1 to 3 samples. */
  97   while(blkCnt > 0u)
  98   {
  99     /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1])  */
 100     /* Compute Sum of squares of the input samples
 101      * and then store the result in a temporary variable, sum. */
 102     in = *__SIMD32(pSrc)++;
 103     sum += ((in << 16) >> 16);
 104     sum += (in >> 16);
 105     sumOfSquares = __SMLALD(in, in, sumOfSquares);
 106     in = *__SIMD32(pSrc)++;
 107     sum += ((in << 16) >> 16);
 108     sum += (in >> 16);
 109     sumOfSquares = __SMLALD(in, in, sumOfSquares);
 110
 111     /* Decrement the loop counter */
 112     blkCnt--;
 113   }
 114
 115   /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
 116    ** No loop unrolling is used. */
 117   blkCnt = blockSize % 0x4u;
 118
 119   while(blkCnt > 0u)
 120   {
 121     /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
 122     /* Compute Sum of squares of the input samples
 123      * and then store the result in a temporary variable, sum. */
 124     in1 = *pSrc++;
 125     sumOfSquares = __SMLALD(in1, in1, sumOfSquares);
 126     sum += in1;
 127
 128     /* Decrement the loop counter */
 129     blkCnt--;
 130   }
 131
 132   /* Compute Mean of squares of the input samples
 133    * and then store the result in a temporary variable, meanOfSquares. */
 134   t = (q15_t) ((1.0 / (blockSize - 1)) * 16384LL);
 135   sumOfSquares = __SSAT((sumOfSquares >> 15u), 16u);
 136
 137   meanOfSquares = (q31_t) ((sumOfSquares * t) >> 14u);
 138
 139   /* Compute mean of all input values */
 140   t = (q15_t) ((1.0 / (blockSize * (blockSize - 1))) * 32768LL);
 141   mean = (q15_t) __SSAT(sum, 16u);
 142
 143   /* Compute square of mean */
 144   squareOfMean = ((q31_t) mean * mean) >> 15;
 145   squareOfMean = (q31_t) (((q63_t) squareOfMean * t) >> 15);
 146
 147   /* mean of the squares minus the square of the mean. */
 148   in1 = (q15_t) (meanOfSquares - squareOfMean);
 149
 150   /* Compute standard deviation and store the result to the destination */
 151   arm_sqrt_q15(in1, pResult);
 152
 153 #else
 154
 155   /* Run the below code for Cortex-M0 */
 156   q15_t in;                                      /* input value */
 157
 158   /* Loop over blockSize number of values */
 159   blkCnt = blockSize;
 160
 161   while(blkCnt > 0u)
 162   {
 163     /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
 164     /* Compute Sum of squares of the input samples
 165      * and then store the result in a temporary variable, sumOfSquares. */
 166     in = *pSrc++;
 167     sumOfSquares += (in * in);
 168
 169     /* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
 170     /* Compute sum of all input values and then store the result in a temporary variable, sum. */
 171     sum += in;
 172
 173     /* Decrement the loop counter */
 174     blkCnt--;
 175   }
 176
 177   /* Compute Mean of squares of the input samples
 178    * and then store the result in a temporary variable, meanOfSquares. */
 179   t = (q15_t) ((1.0 / (blockSize - 1)) * 16384LL);
 180   sumOfSquares = __SSAT((sumOfSquares >> 15u), 16u);
 181   meanOfSquares = (q31_t) ((sumOfSquares * t) >> 14u);
 182
 183   /* Compute mean of all input values */
 184   mean = (q15_t) __SSAT(sum, 16u);
 185
 186   /* Compute square of mean of the input samples
 187    * and then store the result in a temporary variable, squareOfMean.*/
 188   t = (q15_t) ((1.0 / (blockSize * (blockSize - 1))) * 32768LL);
 189   squareOfMean = ((q31_t) mean * mean) >> 15;
 190   squareOfMean = (q31_t) (((q63_t) squareOfMean * t) >> 15);
 191
 192   /* mean of the squares minus the square of the mean. */
 193   in = (q15_t) (meanOfSquares - squareOfMean);
 194
 195   /* Compute standard deviation and store the result to the destination */
 196   arm_sqrt_q15(in, pResult);
 197
 198 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
 199
 200
 201 }
 202
 203 /**
 204  * @} end of STD group
 205  */