mbed official / mbed-dsp

CMSIS DSP library

Dependents:   performance_timer Surfboard_ gps2rtty Capstone ... more

Legacy Warning

This is an mbed 2 library. To learn more about mbed OS 5, visit the docs.

cmsis_dsp/FilteringFunctions/arm_fir_interpolate_f32.c@5:3762170b6d4d, 2015-11-20 (annotated)

Committer:
mbed_official
Date:
Fri Nov 20 08:45:18 2015 +0000
Revision:
5:3762170b6d4d
Parent:
3:7a284390b0ce 
Synchronized with git revision 2eb940b9a73af188d3004a2575fdfbb05febe62b



Full URL: https://github.com/mbedmicro/mbed/commit/2eb940b9a73af188d3004a2575fdfbb05febe62b/



Added option to build rpc library. closes #1426

Who changed what in which revision?

UserRevisionLine numberNew contents of line
emilmont 1:fdd22bb7aa52 1 /* ----------------------------------------------------------------------
mbed_official 5:3762170b6d4d 2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
emilmont 1:fdd22bb7aa52 3 *
mbed_official 5:3762170b6d4d 4 * $Date: 19. March 2015
mbed_official 5:3762170b6d4d 5 * $Revision: V.1.4.5
emilmont 1:fdd22bb7aa52 6 *
emilmont 2:da51fb522205 7 * Project: CMSIS DSP Library
emilmont 2:da51fb522205 8 * Title: arm_fir_interpolate_f32.c
emilmont 1:fdd22bb7aa52 9 *
emilmont 2:da51fb522205 10 * Description: FIR interpolation for floating-point sequences.
emilmont 1:fdd22bb7aa52 11 *
emilmont 1:fdd22bb7aa52 12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
emilmont 1:fdd22bb7aa52 13 *
mbed_official 3:7a284390b0ce 14 * Redistribution and use in source and binary forms, with or without
mbed_official 3:7a284390b0ce 15 * modification, are permitted provided that the following conditions
mbed_official 3:7a284390b0ce 16 * are met:
mbed_official 3:7a284390b0ce 17 * - Redistributions of source code must retain the above copyright
mbed_official 3:7a284390b0ce 18 * notice, this list of conditions and the following disclaimer.
mbed_official 3:7a284390b0ce 19 * - Redistributions in binary form must reproduce the above copyright
mbed_official 3:7a284390b0ce 20 * notice, this list of conditions and the following disclaimer in
mbed_official 3:7a284390b0ce 21 * the documentation and/or other materials provided with the
mbed_official 3:7a284390b0ce 22 * distribution.
mbed_official 3:7a284390b0ce 23 * - Neither the name of ARM LIMITED nor the names of its contributors
mbed_official 3:7a284390b0ce 24 * may be used to endorse or promote products derived from this
mbed_official 3:7a284390b0ce 25 * software without specific prior written permission.
mbed_official 3:7a284390b0ce 26 *
mbed_official 3:7a284390b0ce 27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
mbed_official 3:7a284390b0ce 28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
mbed_official 3:7a284390b0ce 29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
mbed_official 3:7a284390b0ce 30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
mbed_official 3:7a284390b0ce 31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
mbed_official 3:7a284390b0ce 32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
mbed_official 3:7a284390b0ce 33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
mbed_official 3:7a284390b0ce 34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
mbed_official 3:7a284390b0ce 35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
mbed_official 3:7a284390b0ce 36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
mbed_official 3:7a284390b0ce 37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
mbed_official 3:7a284390b0ce 38 * POSSIBILITY OF SUCH DAMAGE.
emilmont 1:fdd22bb7aa52 39 * -------------------------------------------------------------------- */
emilmont 1:fdd22bb7aa52 40
emilmont 1:fdd22bb7aa52 41 #include "arm_math.h"
emilmont 1:fdd22bb7aa52 42
emilmont 1:fdd22bb7aa52 43 /**
emilmont 1:fdd22bb7aa52 44 * @defgroup FIR_Interpolate Finite Impulse Response (FIR) Interpolator
emilmont 1:fdd22bb7aa52 45 *
emilmont 1:fdd22bb7aa52 46 * These functions combine an upsampler (zero stuffer) and an FIR filter.
emilmont 1:fdd22bb7aa52 47 * They are used in multirate systems for increasing the sample rate of a signal without introducing high frequency images.
emilmont 1:fdd22bb7aa52 48 * Conceptually, the functions are equivalent to the block diagram below:
emilmont 1:fdd22bb7aa52 49 * \image html FIRInterpolator.gif "Components included in the FIR Interpolator functions"
emilmont 1:fdd22bb7aa52 50 * After upsampling by a factor of <code>L</code>, the signal should be filtered by a lowpass filter with a normalized
emilmont 1:fdd22bb7aa52 51 * cutoff frequency of <code>1/L</code> in order to eliminate high frequency copies of the spectrum.
emilmont 1:fdd22bb7aa52 52 * The user of the function is responsible for providing the filter coefficients.
emilmont 1:fdd22bb7aa52 53 *
emilmont 1:fdd22bb7aa52 54 * The FIR interpolator functions provided in the CMSIS DSP Library combine the upsampler and FIR filter in an efficient manner.
emilmont 1:fdd22bb7aa52 55 * The upsampler inserts <code>L-1</code> zeros between each sample.
emilmont 1:fdd22bb7aa52 56 * Instead of multiplying by these zero values, the FIR filter is designed to skip them.
emilmont 1:fdd22bb7aa52 57 * This leads to an efficient implementation without any wasted effort.
emilmont 1:fdd22bb7aa52 58 * The functions operate on blocks of input and output data.
emilmont 1:fdd22bb7aa52 59 * <code>pSrc</code> points to an array of <code>blockSize</code> input values and
emilmont 1:fdd22bb7aa52 60 * <code>pDst</code> points to an array of <code>blockSize*L</code> output values.
emilmont 1:fdd22bb7aa52 61 *
emilmont 1:fdd22bb7aa52 62 * The library provides separate functions for Q15, Q31, and floating-point data types.
emilmont 1:fdd22bb7aa52 63 *
emilmont 1:fdd22bb7aa52 64 * \par Algorithm:
emilmont 1:fdd22bb7aa52 65 * The functions use a polyphase filter structure:
emilmont 1:fdd22bb7aa52 66 * <pre>
emilmont 1:fdd22bb7aa52 67 * y[n] = b[0] * x[n] + b[L] * x[n-1] + ... + b[L*(phaseLength-1)] * x[n-phaseLength+1]
emilmont 1:fdd22bb7aa52 68 * y[n+1] = b[1] * x[n] + b[L+1] * x[n-1] + ... + b[L*(phaseLength-1)+1] * x[n-phaseLength+1]
emilmont 1:fdd22bb7aa52 69 * ...
emilmont 1:fdd22bb7aa52 70 * y[n+(L-1)] = b[L-1] * x[n] + b[2*L-1] * x[n-1] + ....+ b[L*(phaseLength-1)+(L-1)] * x[n-phaseLength+1]
emilmont 1:fdd22bb7aa52 71 * </pre>
emilmont 1:fdd22bb7aa52 72 * This approach is more efficient than straightforward upsample-then-filter algorithms.
emilmont 1:fdd22bb7aa52 73 * With this method the computation is reduced by a factor of <code>1/L</code> when compared to using a standard FIR filter.
emilmont 1:fdd22bb7aa52 74 * \par
emilmont 1:fdd22bb7aa52 75 * <code>pCoeffs</code> points to a coefficient array of size <code>numTaps</code>.
emilmont 1:fdd22bb7aa52 76 * <code>numTaps</code> must be a multiple of the interpolation factor <code>L</code> and this is checked by the
emilmont 1:fdd22bb7aa52 77 * initialization functions.
emilmont 1:fdd22bb7aa52 78 * Internally, the function divides the FIR filter's impulse response into shorter filters of length
emilmont 1:fdd22bb7aa52 79 * <code>phaseLength=numTaps/L</code>.
emilmont 1:fdd22bb7aa52 80 * Coefficients are stored in time reversed order.
emilmont 1:fdd22bb7aa52 81 * \par
emilmont 1:fdd22bb7aa52 82 * <pre>
emilmont 1:fdd22bb7aa52 83 * {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}
emilmont 1:fdd22bb7aa52 84 * </pre>
emilmont 1:fdd22bb7aa52 85 * \par
emilmont 1:fdd22bb7aa52 86 * <code>pState</code> points to a state array of size <code>blockSize + phaseLength - 1</code>.
emilmont 1:fdd22bb7aa52 87 * Samples in the state buffer are stored in the order:
emilmont 1:fdd22bb7aa52 88 * \par
emilmont 1:fdd22bb7aa52 89 * <pre>
emilmont 1:fdd22bb7aa52 90 * {x[n-phaseLength+1], x[n-phaseLength], x[n-phaseLength-1], x[n-phaseLength-2]....x[0], x[1], ..., x[blockSize-1]}
emilmont 1:fdd22bb7aa52 91 * </pre>
emilmont 1:fdd22bb7aa52 92 * The state variables are updated after each block of data is processed, the coefficients are untouched.
emilmont 1:fdd22bb7aa52 93 *
emilmont 1:fdd22bb7aa52 94 * \par Instance Structure
emilmont 1:fdd22bb7aa52 95 * The coefficients and state variables for a filter are stored together in an instance data structure.
emilmont 1:fdd22bb7aa52 96 * A separate instance structure must be defined for each filter.
emilmont 1:fdd22bb7aa52 97 * Coefficient arrays may be shared among several instances while state variable array should be allocated separately.
emilmont 1:fdd22bb7aa52 98 * There are separate instance structure declarations for each of the 3 supported data types.
emilmont 1:fdd22bb7aa52 99 *
emilmont 1:fdd22bb7aa52 100 * \par Initialization Functions
emilmont 1:fdd22bb7aa52 101 * There is also an associated initialization function for each data type.
emilmont 1:fdd22bb7aa52 102 * The initialization function performs the following operations:
emilmont 1:fdd22bb7aa52 103 * - Sets the values of the internal structure fields.
emilmont 1:fdd22bb7aa52 104 * - Zeros out the values in the state buffer.
emilmont 1:fdd22bb7aa52 105 * - Checks to make sure that the length of the filter is a multiple of the interpolation factor.
mbed_official 3:7a284390b0ce 106 * To do this manually without calling the init function, assign the follow subfields of the instance structure:
mbed_official 3:7a284390b0ce 107 * L (interpolation factor), pCoeffs, phaseLength (numTaps / L), pState. Also set all of the values in pState to zero.
emilmont 1:fdd22bb7aa52 108 *
emilmont 1:fdd22bb7aa52 109 * \par
emilmont 1:fdd22bb7aa52 110 * Use of the initialization function is optional.
emilmont 1:fdd22bb7aa52 111 * However, if the initialization function is used, then the instance structure cannot be placed into a const data section.
emilmont 1:fdd22bb7aa52 112 * To place an instance structure into a const data section, the instance structure must be manually initialized.
emilmont 1:fdd22bb7aa52 113 * The code below statically initializes each of the 3 different data type filter instance structures
emilmont 1:fdd22bb7aa52 114 * <pre>
emilmont 1:fdd22bb7aa52 115 * arm_fir_interpolate_instance_f32 S = {L, phaseLength, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 116 * arm_fir_interpolate_instance_q31 S = {L, phaseLength, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 117 * arm_fir_interpolate_instance_q15 S = {L, phaseLength, pCoeffs, pState};
emilmont 1:fdd22bb7aa52 118 * </pre>
emilmont 1:fdd22bb7aa52 119 * where <code>L</code> is the interpolation factor; <code>phaseLength=numTaps/L</code> is the
emilmont 1:fdd22bb7aa52 120 * length of each of the shorter FIR filters used internally,
emilmont 1:fdd22bb7aa52 121 * <code>pCoeffs</code> is the address of the coefficient buffer;
emilmont 1:fdd22bb7aa52 122 * <code>pState</code> is the address of the state buffer.
emilmont 1:fdd22bb7aa52 123 * Be sure to set the values in the state buffer to zeros when doing static initialization.
emilmont 1:fdd22bb7aa52 124 *
emilmont 1:fdd22bb7aa52 125 * \par Fixed-Point Behavior
emilmont 1:fdd22bb7aa52 126 * Care must be taken when using the fixed-point versions of the FIR interpolate filter functions.
emilmont 1:fdd22bb7aa52 127 * In particular, the overflow and saturation behavior of the accumulator used in each function must be considered.
emilmont 1:fdd22bb7aa52 128 * Refer to the function specific documentation below for usage guidelines.
emilmont 1:fdd22bb7aa52 129 */
emilmont 1:fdd22bb7aa52 130
emilmont 1:fdd22bb7aa52 131 /**
emilmont 1:fdd22bb7aa52 132 * @addtogroup FIR_Interpolate
emilmont 1:fdd22bb7aa52 133 * @{
emilmont 1:fdd22bb7aa52 134 */
emilmont 1:fdd22bb7aa52 135
emilmont 1:fdd22bb7aa52 136 /**
emilmont 1:fdd22bb7aa52 137 * @brief Processing function for the floating-point FIR interpolator.
emilmont 1:fdd22bb7aa52 138 * @param[in] *S points to an instance of the floating-point FIR interpolator structure.
emilmont 1:fdd22bb7aa52 139 * @param[in] *pSrc points to the block of input data.
emilmont 1:fdd22bb7aa52 140 * @param[out] *pDst points to the block of output data.
emilmont 1:fdd22bb7aa52 141 * @param[in] blockSize number of input samples to process per call.
emilmont 1:fdd22bb7aa52 142 * @return none.
emilmont 1:fdd22bb7aa52 143 */
mbed_official 3:7a284390b0ce 144 #ifndef ARM_MATH_CM0_FAMILY
emilmont 1:fdd22bb7aa52 145
emilmont 1:fdd22bb7aa52 146 /* Run the below code for Cortex-M4 and Cortex-M3 */
emilmont 1:fdd22bb7aa52 147
emilmont 1:fdd22bb7aa52 148 void arm_fir_interpolate_f32(
emilmont 1:fdd22bb7aa52 149 const arm_fir_interpolate_instance_f32 * S,
emilmont 1:fdd22bb7aa52 150 float32_t * pSrc,
emilmont 1:fdd22bb7aa52 151 float32_t * pDst,
emilmont 1:fdd22bb7aa52 152 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 153 {
emilmont 1:fdd22bb7aa52 154 float32_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 155 float32_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 156 float32_t *pStateCurnt; /* Points to the current sample of the state */
emilmont 1:fdd22bb7aa52 157 float32_t *ptr1, *ptr2; /* Temporary pointers for state and coefficient buffers */
emilmont 1:fdd22bb7aa52 158 float32_t sum0; /* Accumulators */
emilmont 1:fdd22bb7aa52 159 float32_t x0, c0; /* Temporary variables to hold state and coefficient values */
emilmont 1:fdd22bb7aa52 160 uint32_t i, blkCnt, j; /* Loop counters */
emilmont 1:fdd22bb7aa52 161 uint16_t phaseLen = S->phaseLength, tapCnt; /* Length of each polyphase filter component */
emilmont 1:fdd22bb7aa52 162 float32_t acc0, acc1, acc2, acc3;
emilmont 1:fdd22bb7aa52 163 float32_t x1, x2, x3;
emilmont 1:fdd22bb7aa52 164 uint32_t blkCntN4;
emilmont 1:fdd22bb7aa52 165 float32_t c1, c2, c3;
emilmont 1:fdd22bb7aa52 166
emilmont 1:fdd22bb7aa52 167 /* S->pState buffer contains previous frame (phaseLen - 1) samples */
emilmont 1:fdd22bb7aa52 168 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 169 pStateCurnt = S->pState + (phaseLen - 1u);
emilmont 1:fdd22bb7aa52 170
emilmont 1:fdd22bb7aa52 171 /* Initialise blkCnt */
emilmont 1:fdd22bb7aa52 172 blkCnt = blockSize / 4;
emilmont 1:fdd22bb7aa52 173 blkCntN4 = blockSize - (4 * blkCnt);
emilmont 1:fdd22bb7aa52 174
emilmont 1:fdd22bb7aa52 175 /* Samples loop unrolled by 4 */
emilmont 1:fdd22bb7aa52 176 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 177 {
emilmont 1:fdd22bb7aa52 178 /* Copy new input sample into the state buffer */
emilmont 1:fdd22bb7aa52 179 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 180 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 181 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 182 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 183
emilmont 1:fdd22bb7aa52 184 /* Address modifier index of coefficient buffer */
emilmont 1:fdd22bb7aa52 185 j = 1u;
emilmont 1:fdd22bb7aa52 186
emilmont 1:fdd22bb7aa52 187 /* Loop over the Interpolation factor. */
emilmont 1:fdd22bb7aa52 188 i = (S->L);
emilmont 1:fdd22bb7aa52 189
emilmont 1:fdd22bb7aa52 190 while(i > 0u)
emilmont 1:fdd22bb7aa52 191 {
emilmont 1:fdd22bb7aa52 192 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 193 acc0 = 0.0f;
emilmont 1:fdd22bb7aa52 194 acc1 = 0.0f;
emilmont 1:fdd22bb7aa52 195 acc2 = 0.0f;
emilmont 1:fdd22bb7aa52 196 acc3 = 0.0f;
emilmont 1:fdd22bb7aa52 197
emilmont 1:fdd22bb7aa52 198 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 199 ptr1 = pState;
emilmont 1:fdd22bb7aa52 200
emilmont 1:fdd22bb7aa52 201 /* Initialize coefficient pointer */
emilmont 1:fdd22bb7aa52 202 ptr2 = pCoeffs + (S->L - j);
emilmont 1:fdd22bb7aa52 203
emilmont 1:fdd22bb7aa52 204 /* Loop over the polyPhase length. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 205 ** Repeat until we've computed numTaps-(4*S->L) coefficients. */
emilmont 1:fdd22bb7aa52 206 tapCnt = phaseLen >> 2u;
emilmont 1:fdd22bb7aa52 207
emilmont 1:fdd22bb7aa52 208 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 209 x1 = *(ptr1++);
emilmont 1:fdd22bb7aa52 210 x2 = *(ptr1++);
emilmont 1:fdd22bb7aa52 211
emilmont 1:fdd22bb7aa52 212 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 213 {
emilmont 1:fdd22bb7aa52 214
emilmont 1:fdd22bb7aa52 215 /* Read the input sample */
emilmont 1:fdd22bb7aa52 216 x3 = *(ptr1++);
emilmont 1:fdd22bb7aa52 217
emilmont 1:fdd22bb7aa52 218 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 219 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 220
emilmont 1:fdd22bb7aa52 221 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 222 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 223 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 224 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 225 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 226
emilmont 1:fdd22bb7aa52 227 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 228 c1 = *(ptr2 + S->L);
emilmont 1:fdd22bb7aa52 229
emilmont 1:fdd22bb7aa52 230 /* Read the input sample */
emilmont 1:fdd22bb7aa52 231 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 232
emilmont 1:fdd22bb7aa52 233 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 234 acc0 += x1 * c1;
emilmont 1:fdd22bb7aa52 235 acc1 += x2 * c1;
emilmont 1:fdd22bb7aa52 236 acc2 += x3 * c1;
emilmont 1:fdd22bb7aa52 237 acc3 += x0 * c1;
emilmont 1:fdd22bb7aa52 238
emilmont 1:fdd22bb7aa52 239 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 240 c2 = *(ptr2 + S->L * 2);
emilmont 1:fdd22bb7aa52 241
emilmont 1:fdd22bb7aa52 242 /* Read the input sample */
emilmont 1:fdd22bb7aa52 243 x1 = *(ptr1++);
emilmont 1:fdd22bb7aa52 244
emilmont 1:fdd22bb7aa52 245 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 246 acc0 += x2 * c2;
emilmont 1:fdd22bb7aa52 247 acc1 += x3 * c2;
emilmont 1:fdd22bb7aa52 248 acc2 += x0 * c2;
emilmont 1:fdd22bb7aa52 249 acc3 += x1 * c2;
emilmont 1:fdd22bb7aa52 250
emilmont 1:fdd22bb7aa52 251 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 252 c3 = *(ptr2 + S->L * 3);
emilmont 1:fdd22bb7aa52 253
emilmont 1:fdd22bb7aa52 254 /* Read the input sample */
emilmont 1:fdd22bb7aa52 255 x2 = *(ptr1++);
emilmont 1:fdd22bb7aa52 256
emilmont 1:fdd22bb7aa52 257 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 258 acc0 += x3 * c3;
emilmont 1:fdd22bb7aa52 259 acc1 += x0 * c3;
emilmont 1:fdd22bb7aa52 260 acc2 += x1 * c3;
emilmont 1:fdd22bb7aa52 261 acc3 += x2 * c3;
emilmont 1:fdd22bb7aa52 262
emilmont 1:fdd22bb7aa52 263
emilmont 1:fdd22bb7aa52 264 /* Upsampling is done by stuffing L-1 zeros between each sample.
emilmont 1:fdd22bb7aa52 265 * So instead of multiplying zeros with coefficients,
emilmont 1:fdd22bb7aa52 266 * Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 267 ptr2 += 4 * S->L;
emilmont 1:fdd22bb7aa52 268
emilmont 1:fdd22bb7aa52 269 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 270 tapCnt--;
emilmont 1:fdd22bb7aa52 271 }
emilmont 1:fdd22bb7aa52 272
emilmont 1:fdd22bb7aa52 273 /* If the polyPhase length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 274 tapCnt = phaseLen % 0x4u;
emilmont 1:fdd22bb7aa52 275
emilmont 1:fdd22bb7aa52 276 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 277 {
emilmont 1:fdd22bb7aa52 278
emilmont 1:fdd22bb7aa52 279 /* Read the input sample */
emilmont 1:fdd22bb7aa52 280 x3 = *(ptr1++);
emilmont 1:fdd22bb7aa52 281
emilmont 1:fdd22bb7aa52 282 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 283 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 284
emilmont 1:fdd22bb7aa52 285 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 286 acc0 += x0 * c0;
emilmont 1:fdd22bb7aa52 287 acc1 += x1 * c0;
emilmont 1:fdd22bb7aa52 288 acc2 += x2 * c0;
emilmont 1:fdd22bb7aa52 289 acc3 += x3 * c0;
emilmont 1:fdd22bb7aa52 290
emilmont 1:fdd22bb7aa52 291 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 292 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 293
emilmont 1:fdd22bb7aa52 294 /* update states for next sample processing */
emilmont 1:fdd22bb7aa52 295 x0 = x1;
emilmont 1:fdd22bb7aa52 296 x1 = x2;
emilmont 1:fdd22bb7aa52 297 x2 = x3;
emilmont 1:fdd22bb7aa52 298
emilmont 1:fdd22bb7aa52 299 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 300 tapCnt--;
emilmont 1:fdd22bb7aa52 301 }
emilmont 1:fdd22bb7aa52 302
emilmont 1:fdd22bb7aa52 303 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 304 *pDst = acc0;
emilmont 1:fdd22bb7aa52 305 *(pDst + S->L) = acc1;
emilmont 1:fdd22bb7aa52 306 *(pDst + 2 * S->L) = acc2;
emilmont 1:fdd22bb7aa52 307 *(pDst + 3 * S->L) = acc3;
emilmont 1:fdd22bb7aa52 308
emilmont 1:fdd22bb7aa52 309 pDst++;
emilmont 1:fdd22bb7aa52 310
emilmont 1:fdd22bb7aa52 311 /* Increment the address modifier index of coefficient buffer */
emilmont 1:fdd22bb7aa52 312 j++;
emilmont 1:fdd22bb7aa52 313
emilmont 1:fdd22bb7aa52 314 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 315 i--;
emilmont 1:fdd22bb7aa52 316 }
emilmont 1:fdd22bb7aa52 317
emilmont 1:fdd22bb7aa52 318 /* Advance the state pointer by 1
emilmont 1:fdd22bb7aa52 319 * to process the next group of interpolation factor number samples */
emilmont 1:fdd22bb7aa52 320 pState = pState + 4;
emilmont 1:fdd22bb7aa52 321
emilmont 1:fdd22bb7aa52 322 pDst += S->L * 3;
emilmont 1:fdd22bb7aa52 323
emilmont 1:fdd22bb7aa52 324 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 325 blkCnt--;
emilmont 1:fdd22bb7aa52 326 }
emilmont 1:fdd22bb7aa52 327
emilmont 1:fdd22bb7aa52 328 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
emilmont 1:fdd22bb7aa52 329 ** No loop unrolling is used. */
emilmont 1:fdd22bb7aa52 330
emilmont 1:fdd22bb7aa52 331 while(blkCntN4 > 0u)
emilmont 1:fdd22bb7aa52 332 {
emilmont 1:fdd22bb7aa52 333 /* Copy new input sample into the state buffer */
emilmont 1:fdd22bb7aa52 334 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 335
emilmont 1:fdd22bb7aa52 336 /* Address modifier index of coefficient buffer */
emilmont 1:fdd22bb7aa52 337 j = 1u;
emilmont 1:fdd22bb7aa52 338
emilmont 1:fdd22bb7aa52 339 /* Loop over the Interpolation factor. */
emilmont 1:fdd22bb7aa52 340 i = S->L;
emilmont 1:fdd22bb7aa52 341 while(i > 0u)
emilmont 1:fdd22bb7aa52 342 {
emilmont 1:fdd22bb7aa52 343 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 344 sum0 = 0.0f;
emilmont 1:fdd22bb7aa52 345
emilmont 1:fdd22bb7aa52 346 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 347 ptr1 = pState;
emilmont 1:fdd22bb7aa52 348
emilmont 1:fdd22bb7aa52 349 /* Initialize coefficient pointer */
emilmont 1:fdd22bb7aa52 350 ptr2 = pCoeffs + (S->L - j);
emilmont 1:fdd22bb7aa52 351
emilmont 1:fdd22bb7aa52 352 /* Loop over the polyPhase length. Unroll by a factor of 4.
emilmont 1:fdd22bb7aa52 353 ** Repeat until we've computed numTaps-(4*S->L) coefficients. */
emilmont 1:fdd22bb7aa52 354 tapCnt = phaseLen >> 2u;
emilmont 1:fdd22bb7aa52 355 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 356 {
emilmont 1:fdd22bb7aa52 357
emilmont 1:fdd22bb7aa52 358 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 359 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 360
emilmont 1:fdd22bb7aa52 361 /* Upsampling is done by stuffing L-1 zeros between each sample.
emilmont 1:fdd22bb7aa52 362 * So instead of multiplying zeros with coefficients,
emilmont 1:fdd22bb7aa52 363 * Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 364 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 365
emilmont 1:fdd22bb7aa52 366 /* Read the input sample */
emilmont 1:fdd22bb7aa52 367 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 368
emilmont 1:fdd22bb7aa52 369 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 370 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 371
emilmont 1:fdd22bb7aa52 372 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 373 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 374
emilmont 1:fdd22bb7aa52 375 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 376 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 377
emilmont 1:fdd22bb7aa52 378 /* Read the input sample */
emilmont 1:fdd22bb7aa52 379 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 380
emilmont 1:fdd22bb7aa52 381 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 382 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 383
emilmont 1:fdd22bb7aa52 384 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 385 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 386
emilmont 1:fdd22bb7aa52 387 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 388 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 389
emilmont 1:fdd22bb7aa52 390 /* Read the input sample */
emilmont 1:fdd22bb7aa52 391 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 392
emilmont 1:fdd22bb7aa52 393 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 394 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 395
emilmont 1:fdd22bb7aa52 396 /* Read the coefficient */
emilmont 1:fdd22bb7aa52 397 c0 = *(ptr2);
emilmont 1:fdd22bb7aa52 398
emilmont 1:fdd22bb7aa52 399 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 400 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 401
emilmont 1:fdd22bb7aa52 402 /* Read the input sample */
emilmont 1:fdd22bb7aa52 403 x0 = *(ptr1++);
emilmont 1:fdd22bb7aa52 404
emilmont 1:fdd22bb7aa52 405 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 406 sum0 += x0 * c0;
emilmont 1:fdd22bb7aa52 407
emilmont 1:fdd22bb7aa52 408 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 409 tapCnt--;
emilmont 1:fdd22bb7aa52 410 }
emilmont 1:fdd22bb7aa52 411
emilmont 1:fdd22bb7aa52 412 /* If the polyPhase length is not a multiple of 4, compute the remaining filter taps */
emilmont 1:fdd22bb7aa52 413 tapCnt = phaseLen % 0x4u;
emilmont 1:fdd22bb7aa52 414
emilmont 1:fdd22bb7aa52 415 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 416 {
emilmont 1:fdd22bb7aa52 417 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 418 sum0 += *(ptr1++) * (*ptr2);
emilmont 1:fdd22bb7aa52 419
emilmont 1:fdd22bb7aa52 420 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 421 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 422
emilmont 1:fdd22bb7aa52 423 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 424 tapCnt--;
emilmont 1:fdd22bb7aa52 425 }
emilmont 1:fdd22bb7aa52 426
emilmont 1:fdd22bb7aa52 427 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 428 *pDst++ = sum0;
emilmont 1:fdd22bb7aa52 429
emilmont 1:fdd22bb7aa52 430 /* Increment the address modifier index of coefficient buffer */
emilmont 1:fdd22bb7aa52 431 j++;
emilmont 1:fdd22bb7aa52 432
emilmont 1:fdd22bb7aa52 433 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 434 i--;
emilmont 1:fdd22bb7aa52 435 }
emilmont 1:fdd22bb7aa52 436
emilmont 1:fdd22bb7aa52 437 /* Advance the state pointer by 1
emilmont 1:fdd22bb7aa52 438 * to process the next group of interpolation factor number samples */
emilmont 1:fdd22bb7aa52 439 pState = pState + 1;
emilmont 1:fdd22bb7aa52 440
emilmont 1:fdd22bb7aa52 441 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 442 blkCntN4--;
emilmont 1:fdd22bb7aa52 443 }
emilmont 1:fdd22bb7aa52 444
emilmont 1:fdd22bb7aa52 445 /* Processing is complete.
emilmont 1:fdd22bb7aa52 446 ** Now copy the last phaseLen - 1 samples to the satrt of the state buffer.
emilmont 1:fdd22bb7aa52 447 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 448
emilmont 1:fdd22bb7aa52 449 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 450 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 451
emilmont 1:fdd22bb7aa52 452 tapCnt = (phaseLen - 1u) >> 2u;
emilmont 1:fdd22bb7aa52 453
emilmont 1:fdd22bb7aa52 454 /* copy data */
emilmont 1:fdd22bb7aa52 455 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 456 {
emilmont 1:fdd22bb7aa52 457 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 458 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 459 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 460 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 461
emilmont 1:fdd22bb7aa52 462 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 463 tapCnt--;
emilmont 1:fdd22bb7aa52 464 }
emilmont 1:fdd22bb7aa52 465
emilmont 1:fdd22bb7aa52 466 tapCnt = (phaseLen - 1u) % 0x04u;
emilmont 1:fdd22bb7aa52 467
emilmont 1:fdd22bb7aa52 468 /* copy data */
emilmont 1:fdd22bb7aa52 469 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 470 {
emilmont 1:fdd22bb7aa52 471 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 472
emilmont 1:fdd22bb7aa52 473 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 474 tapCnt--;
emilmont 1:fdd22bb7aa52 475 }
emilmont 1:fdd22bb7aa52 476 }
emilmont 1:fdd22bb7aa52 477
emilmont 1:fdd22bb7aa52 478 #else
emilmont 1:fdd22bb7aa52 479
emilmont 1:fdd22bb7aa52 480 /* Run the below code for Cortex-M0 */
emilmont 1:fdd22bb7aa52 481
emilmont 1:fdd22bb7aa52 482 void arm_fir_interpolate_f32(
emilmont 1:fdd22bb7aa52 483 const arm_fir_interpolate_instance_f32 * S,
emilmont 1:fdd22bb7aa52 484 float32_t * pSrc,
emilmont 1:fdd22bb7aa52 485 float32_t * pDst,
emilmont 1:fdd22bb7aa52 486 uint32_t blockSize)
emilmont 1:fdd22bb7aa52 487 {
emilmont 1:fdd22bb7aa52 488 float32_t *pState = S->pState; /* State pointer */
emilmont 1:fdd22bb7aa52 489 float32_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */
emilmont 1:fdd22bb7aa52 490 float32_t *pStateCurnt; /* Points to the current sample of the state */
emilmont 1:fdd22bb7aa52 491 float32_t *ptr1, *ptr2; /* Temporary pointers for state and coefficient buffers */
emilmont 1:fdd22bb7aa52 492
emilmont 1:fdd22bb7aa52 493
emilmont 1:fdd22bb7aa52 494 float32_t sum; /* Accumulator */
emilmont 1:fdd22bb7aa52 495 uint32_t i, blkCnt; /* Loop counters */
emilmont 1:fdd22bb7aa52 496 uint16_t phaseLen = S->phaseLength, tapCnt; /* Length of each polyphase filter component */
emilmont 1:fdd22bb7aa52 497
emilmont 1:fdd22bb7aa52 498
emilmont 1:fdd22bb7aa52 499 /* S->pState buffer contains previous frame (phaseLen - 1) samples */
emilmont 1:fdd22bb7aa52 500 /* pStateCurnt points to the location where the new input data should be written */
emilmont 1:fdd22bb7aa52 501 pStateCurnt = S->pState + (phaseLen - 1u);
emilmont 1:fdd22bb7aa52 502
emilmont 1:fdd22bb7aa52 503 /* Total number of intput samples */
emilmont 1:fdd22bb7aa52 504 blkCnt = blockSize;
emilmont 1:fdd22bb7aa52 505
emilmont 1:fdd22bb7aa52 506 /* Loop over the blockSize. */
emilmont 1:fdd22bb7aa52 507 while(blkCnt > 0u)
emilmont 1:fdd22bb7aa52 508 {
emilmont 1:fdd22bb7aa52 509 /* Copy new input sample into the state buffer */
emilmont 1:fdd22bb7aa52 510 *pStateCurnt++ = *pSrc++;
emilmont 1:fdd22bb7aa52 511
emilmont 1:fdd22bb7aa52 512 /* Loop over the Interpolation factor. */
emilmont 1:fdd22bb7aa52 513 i = S->L;
emilmont 1:fdd22bb7aa52 514
emilmont 1:fdd22bb7aa52 515 while(i > 0u)
emilmont 1:fdd22bb7aa52 516 {
emilmont 1:fdd22bb7aa52 517 /* Set accumulator to zero */
emilmont 1:fdd22bb7aa52 518 sum = 0.0f;
emilmont 1:fdd22bb7aa52 519
emilmont 1:fdd22bb7aa52 520 /* Initialize state pointer */
emilmont 1:fdd22bb7aa52 521 ptr1 = pState;
emilmont 1:fdd22bb7aa52 522
emilmont 1:fdd22bb7aa52 523 /* Initialize coefficient pointer */
emilmont 1:fdd22bb7aa52 524 ptr2 = pCoeffs + (i - 1u);
emilmont 1:fdd22bb7aa52 525
emilmont 1:fdd22bb7aa52 526 /* Loop over the polyPhase length */
emilmont 1:fdd22bb7aa52 527 tapCnt = phaseLen;
emilmont 1:fdd22bb7aa52 528
emilmont 1:fdd22bb7aa52 529 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 530 {
emilmont 1:fdd22bb7aa52 531 /* Perform the multiply-accumulate */
emilmont 1:fdd22bb7aa52 532 sum += *ptr1++ * *ptr2;
emilmont 1:fdd22bb7aa52 533
emilmont 1:fdd22bb7aa52 534 /* Increment the coefficient pointer by interpolation factor times. */
emilmont 1:fdd22bb7aa52 535 ptr2 += S->L;
emilmont 1:fdd22bb7aa52 536
emilmont 1:fdd22bb7aa52 537 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 538 tapCnt--;
emilmont 1:fdd22bb7aa52 539 }
emilmont 1:fdd22bb7aa52 540
emilmont 1:fdd22bb7aa52 541 /* The result is in the accumulator, store in the destination buffer. */
emilmont 1:fdd22bb7aa52 542 *pDst++ = sum;
emilmont 1:fdd22bb7aa52 543
emilmont 1:fdd22bb7aa52 544 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 545 i--;
emilmont 1:fdd22bb7aa52 546 }
emilmont 1:fdd22bb7aa52 547
emilmont 1:fdd22bb7aa52 548 /* Advance the state pointer by 1
emilmont 1:fdd22bb7aa52 549 * to process the next group of interpolation factor number samples */
emilmont 1:fdd22bb7aa52 550 pState = pState + 1;
emilmont 1:fdd22bb7aa52 551
emilmont 1:fdd22bb7aa52 552 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 553 blkCnt--;
emilmont 1:fdd22bb7aa52 554 }
emilmont 1:fdd22bb7aa52 555
emilmont 1:fdd22bb7aa52 556 /* Processing is complete.
emilmont 1:fdd22bb7aa52 557 ** Now copy the last phaseLen - 1 samples to the start of the state buffer.
emilmont 1:fdd22bb7aa52 558 ** This prepares the state buffer for the next function call. */
emilmont 1:fdd22bb7aa52 559
emilmont 1:fdd22bb7aa52 560 /* Points to the start of the state buffer */
emilmont 1:fdd22bb7aa52 561 pStateCurnt = S->pState;
emilmont 1:fdd22bb7aa52 562
emilmont 1:fdd22bb7aa52 563 tapCnt = phaseLen - 1u;
emilmont 1:fdd22bb7aa52 564
emilmont 1:fdd22bb7aa52 565 while(tapCnt > 0u)
emilmont 1:fdd22bb7aa52 566 {
emilmont 1:fdd22bb7aa52 567 *pStateCurnt++ = *pState++;
emilmont 1:fdd22bb7aa52 568
emilmont 1:fdd22bb7aa52 569 /* Decrement the loop counter */
emilmont 1:fdd22bb7aa52 570 tapCnt--;
emilmont 1:fdd22bb7aa52 571 }
emilmont 1:fdd22bb7aa52 572
emilmont 1:fdd22bb7aa52 573 }
emilmont 1:fdd22bb7aa52 574
mbed_official 3:7a284390b0ce 575 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
emilmont 1:fdd22bb7aa52 576
emilmont 1:fdd22bb7aa52 577
emilmont 1:fdd22bb7aa52 578
emilmont 1:fdd22bb7aa52 579 /**
emilmont 1:fdd22bb7aa52 580 * @} end of FIR_Interpolate group
emilmont 1:fdd22bb7aa52 581 */