V4.0.1 of the ARM CMSIS DSP libraries. Note that arm_bitreversal2.s, arm_cfft_f32.c and arm_rfft_fast_f32.c had to be removed. arm_bitreversal2.s will not assemble with the online tools. So, the fast f32 FFT functions are not yet available. All the other FFT functions are available.
Dependents: MPU9150_Example fir_f32 fir_f32 MPU9150_nucleo_noni2cdev ... more
FilteringFunctions/arm_lms_f32.c@0:3d9c67d97d6f, 2014-07-28 (annotated)
- Committer:
- emh203
- Date:
- Mon Jul 28 15:03:15 2014 +0000
- Revision:
- 0:3d9c67d97d6f
1st working commit. Had to remove arm_bitreversal2.s arm_cfft_f32.c and arm_rfft_fast_f32.c. The .s will not assemble. For now I removed these functions so we could at least have a library for the other functions.
Who changed what in which revision?
User | Revision | Line number | New contents of line |
---|---|---|---|
emh203 | 0:3d9c67d97d6f | 1 | /* ---------------------------------------------------------------------- |
emh203 | 0:3d9c67d97d6f | 2 | * Copyright (C) 2010-2014 ARM Limited. All rights reserved. |
emh203 | 0:3d9c67d97d6f | 3 | * |
emh203 | 0:3d9c67d97d6f | 4 | * $Date: 12. March 2014 |
emh203 | 0:3d9c67d97d6f | 5 | * $Revision: V1.4.3 |
emh203 | 0:3d9c67d97d6f | 6 | * |
emh203 | 0:3d9c67d97d6f | 7 | * Project: CMSIS DSP Library |
emh203 | 0:3d9c67d97d6f | 8 | * Title: arm_lms_f32.c |
emh203 | 0:3d9c67d97d6f | 9 | * |
emh203 | 0:3d9c67d97d6f | 10 | * Description: Processing function for the floating-point LMS filter. |
emh203 | 0:3d9c67d97d6f | 11 | * |
emh203 | 0:3d9c67d97d6f | 12 | * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0 |
emh203 | 0:3d9c67d97d6f | 13 | * |
emh203 | 0:3d9c67d97d6f | 14 | * Redistribution and use in source and binary forms, with or without |
emh203 | 0:3d9c67d97d6f | 15 | * modification, are permitted provided that the following conditions |
emh203 | 0:3d9c67d97d6f | 16 | * are met: |
emh203 | 0:3d9c67d97d6f | 17 | * - Redistributions of source code must retain the above copyright |
emh203 | 0:3d9c67d97d6f | 18 | * notice, this list of conditions and the following disclaimer. |
emh203 | 0:3d9c67d97d6f | 19 | * - Redistributions in binary form must reproduce the above copyright |
emh203 | 0:3d9c67d97d6f | 20 | * notice, this list of conditions and the following disclaimer in |
emh203 | 0:3d9c67d97d6f | 21 | * the documentation and/or other materials provided with the |
emh203 | 0:3d9c67d97d6f | 22 | * distribution. |
emh203 | 0:3d9c67d97d6f | 23 | * - Neither the name of ARM LIMITED nor the names of its contributors |
emh203 | 0:3d9c67d97d6f | 24 | * may be used to endorse or promote products derived from this |
emh203 | 0:3d9c67d97d6f | 25 | * software without specific prior written permission. |
emh203 | 0:3d9c67d97d6f | 26 | * |
emh203 | 0:3d9c67d97d6f | 27 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
emh203 | 0:3d9c67d97d6f | 28 | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
emh203 | 0:3d9c67d97d6f | 29 | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS |
emh203 | 0:3d9c67d97d6f | 30 | * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE |
emh203 | 0:3d9c67d97d6f | 31 | * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, |
emh203 | 0:3d9c67d97d6f | 32 | * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, |
emh203 | 0:3d9c67d97d6f | 33 | * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; |
emh203 | 0:3d9c67d97d6f | 34 | * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER |
emh203 | 0:3d9c67d97d6f | 35 | * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT |
emh203 | 0:3d9c67d97d6f | 36 | * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN |
emh203 | 0:3d9c67d97d6f | 37 | * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE |
emh203 | 0:3d9c67d97d6f | 38 | * POSSIBILITY OF SUCH DAMAGE. |
emh203 | 0:3d9c67d97d6f | 39 | * -------------------------------------------------------------------- */ |
emh203 | 0:3d9c67d97d6f | 40 | |
emh203 | 0:3d9c67d97d6f | 41 | #include "arm_math.h" |
emh203 | 0:3d9c67d97d6f | 42 | |
emh203 | 0:3d9c67d97d6f | 43 | /** |
emh203 | 0:3d9c67d97d6f | 44 | * @ingroup groupFilters |
emh203 | 0:3d9c67d97d6f | 45 | */ |
emh203 | 0:3d9c67d97d6f | 46 | |
emh203 | 0:3d9c67d97d6f | 47 | /** |
emh203 | 0:3d9c67d97d6f | 48 | * @defgroup LMS Least Mean Square (LMS) Filters |
emh203 | 0:3d9c67d97d6f | 49 | * |
emh203 | 0:3d9c67d97d6f | 50 | * LMS filters are a class of adaptive filters that are able to "learn" an unknown transfer functions. |
emh203 | 0:3d9c67d97d6f | 51 | * LMS filters use a gradient descent method in which the filter coefficients are updated based on the instantaneous error signal. |
emh203 | 0:3d9c67d97d6f | 52 | * Adaptive filters are often used in communication systems, equalizers, and noise removal. |
emh203 | 0:3d9c67d97d6f | 53 | * The CMSIS DSP Library contains LMS filter functions that operate on Q15, Q31, and floating-point data types. |
emh203 | 0:3d9c67d97d6f | 54 | * The library also contains normalized LMS filters in which the filter coefficient adaptation is indepedent of the level of the input signal. |
emh203 | 0:3d9c67d97d6f | 55 | * |
emh203 | 0:3d9c67d97d6f | 56 | * An LMS filter consists of two components as shown below. |
emh203 | 0:3d9c67d97d6f | 57 | * The first component is a standard transversal or FIR filter. |
emh203 | 0:3d9c67d97d6f | 58 | * The second component is a coefficient update mechanism. |
emh203 | 0:3d9c67d97d6f | 59 | * The LMS filter has two input signals. |
emh203 | 0:3d9c67d97d6f | 60 | * The "input" feeds the FIR filter while the "reference input" corresponds to the desired output of the FIR filter. |
emh203 | 0:3d9c67d97d6f | 61 | * That is, the FIR filter coefficients are updated so that the output of the FIR filter matches the reference input. |
emh203 | 0:3d9c67d97d6f | 62 | * The filter coefficient update mechanism is based on the difference between the FIR filter output and the reference input. |
emh203 | 0:3d9c67d97d6f | 63 | * This "error signal" tends towards zero as the filter adapts. |
emh203 | 0:3d9c67d97d6f | 64 | * The LMS processing functions accept the input and reference input signals and generate the filter output and error signal. |
emh203 | 0:3d9c67d97d6f | 65 | * \image html LMS.gif "Internal structure of the Least Mean Square filter" |
emh203 | 0:3d9c67d97d6f | 66 | * |
emh203 | 0:3d9c67d97d6f | 67 | * The functions operate on blocks of data and each call to the function processes |
emh203 | 0:3d9c67d97d6f | 68 | * <code>blockSize</code> samples through the filter. |
emh203 | 0:3d9c67d97d6f | 69 | * <code>pSrc</code> points to input signal, <code>pRef</code> points to reference signal, |
emh203 | 0:3d9c67d97d6f | 70 | * <code>pOut</code> points to output signal and <code>pErr</code> points to error signal. |
emh203 | 0:3d9c67d97d6f | 71 | * All arrays contain <code>blockSize</code> values. |
emh203 | 0:3d9c67d97d6f | 72 | * |
emh203 | 0:3d9c67d97d6f | 73 | * The functions operate on a block-by-block basis. |
emh203 | 0:3d9c67d97d6f | 74 | * Internally, the filter coefficients <code>b[n]</code> are updated on a sample-by-sample basis. |
emh203 | 0:3d9c67d97d6f | 75 | * The convergence of the LMS filter is slower compared to the normalized LMS algorithm. |
emh203 | 0:3d9c67d97d6f | 76 | * |
emh203 | 0:3d9c67d97d6f | 77 | * \par Algorithm: |
emh203 | 0:3d9c67d97d6f | 78 | * The output signal <code>y[n]</code> is computed by a standard FIR filter: |
emh203 | 0:3d9c67d97d6f | 79 | * <pre> |
emh203 | 0:3d9c67d97d6f | 80 | * y[n] = b[0] * x[n] + b[1] * x[n-1] + b[2] * x[n-2] + ...+ b[numTaps-1] * x[n-numTaps+1] |
emh203 | 0:3d9c67d97d6f | 81 | * </pre> |
emh203 | 0:3d9c67d97d6f | 82 | * |
emh203 | 0:3d9c67d97d6f | 83 | * \par |
emh203 | 0:3d9c67d97d6f | 84 | * The error signal equals the difference between the reference signal <code>d[n]</code> and the filter output: |
emh203 | 0:3d9c67d97d6f | 85 | * <pre> |
emh203 | 0:3d9c67d97d6f | 86 | * e[n] = d[n] - y[n]. |
emh203 | 0:3d9c67d97d6f | 87 | * </pre> |
emh203 | 0:3d9c67d97d6f | 88 | * |
emh203 | 0:3d9c67d97d6f | 89 | * \par |
emh203 | 0:3d9c67d97d6f | 90 | * After each sample of the error signal is computed, the filter coefficients <code>b[k]</code> are updated on a sample-by-sample basis: |
emh203 | 0:3d9c67d97d6f | 91 | * <pre> |
emh203 | 0:3d9c67d97d6f | 92 | * b[k] = b[k] + e[n] * mu * x[n-k], for k=0, 1, ..., numTaps-1 |
emh203 | 0:3d9c67d97d6f | 93 | * </pre> |
emh203 | 0:3d9c67d97d6f | 94 | * where <code>mu</code> is the step size and controls the rate of coefficient convergence. |
emh203 | 0:3d9c67d97d6f | 95 | *\par |
emh203 | 0:3d9c67d97d6f | 96 | * In the APIs, <code>pCoeffs</code> points to a coefficient array of size <code>numTaps</code>. |
emh203 | 0:3d9c67d97d6f | 97 | * Coefficients are stored in time reversed order. |
emh203 | 0:3d9c67d97d6f | 98 | * \par |
emh203 | 0:3d9c67d97d6f | 99 | * <pre> |
emh203 | 0:3d9c67d97d6f | 100 | * {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]} |
emh203 | 0:3d9c67d97d6f | 101 | * </pre> |
emh203 | 0:3d9c67d97d6f | 102 | * \par |
emh203 | 0:3d9c67d97d6f | 103 | * <code>pState</code> points to a state array of size <code>numTaps + blockSize - 1</code>. |
emh203 | 0:3d9c67d97d6f | 104 | * Samples in the state buffer are stored in the order: |
emh203 | 0:3d9c67d97d6f | 105 | * \par |
emh203 | 0:3d9c67d97d6f | 106 | * <pre> |
emh203 | 0:3d9c67d97d6f | 107 | * {x[n-numTaps+1], x[n-numTaps], x[n-numTaps-1], x[n-numTaps-2]....x[0], x[1], ..., x[blockSize-1]} |
emh203 | 0:3d9c67d97d6f | 108 | * </pre> |
emh203 | 0:3d9c67d97d6f | 109 | * \par |
emh203 | 0:3d9c67d97d6f | 110 | * Note that the length of the state buffer exceeds the length of the coefficient array by <code>blockSize-1</code> samples. |
emh203 | 0:3d9c67d97d6f | 111 | * The increased state buffer length allows circular addressing, which is traditionally used in FIR filters, |
emh203 | 0:3d9c67d97d6f | 112 | * to be avoided and yields a significant speed improvement. |
emh203 | 0:3d9c67d97d6f | 113 | * The state variables are updated after each block of data is processed. |
emh203 | 0:3d9c67d97d6f | 114 | * \par Instance Structure |
emh203 | 0:3d9c67d97d6f | 115 | * The coefficients and state variables for a filter are stored together in an instance data structure. |
emh203 | 0:3d9c67d97d6f | 116 | * A separate instance structure must be defined for each filter and |
emh203 | 0:3d9c67d97d6f | 117 | * coefficient and state arrays cannot be shared among instances. |
emh203 | 0:3d9c67d97d6f | 118 | * There are separate instance structure declarations for each of the 3 supported data types. |
emh203 | 0:3d9c67d97d6f | 119 | * |
emh203 | 0:3d9c67d97d6f | 120 | * \par Initialization Functions |
emh203 | 0:3d9c67d97d6f | 121 | * There is also an associated initialization function for each data type. |
emh203 | 0:3d9c67d97d6f | 122 | * The initialization function performs the following operations: |
emh203 | 0:3d9c67d97d6f | 123 | * - Sets the values of the internal structure fields. |
emh203 | 0:3d9c67d97d6f | 124 | * - Zeros out the values in the state buffer. |
emh203 | 0:3d9c67d97d6f | 125 | * To do this manually without calling the init function, assign the follow subfields of the instance structure: |
emh203 | 0:3d9c67d97d6f | 126 | * numTaps, pCoeffs, mu, postShift (not for f32), pState. Also set all of the values in pState to zero. |
emh203 | 0:3d9c67d97d6f | 127 | * |
emh203 | 0:3d9c67d97d6f | 128 | * \par |
emh203 | 0:3d9c67d97d6f | 129 | * Use of the initialization function is optional. |
emh203 | 0:3d9c67d97d6f | 130 | * However, if the initialization function is used, then the instance structure cannot be placed into a const data section. |
emh203 | 0:3d9c67d97d6f | 131 | * To place an instance structure into a const data section, the instance structure must be manually initialized. |
emh203 | 0:3d9c67d97d6f | 132 | * Set the values in the state buffer to zeros before static initialization. |
emh203 | 0:3d9c67d97d6f | 133 | * The code below statically initializes each of the 3 different data type filter instance structures |
emh203 | 0:3d9c67d97d6f | 134 | * <pre> |
emh203 | 0:3d9c67d97d6f | 135 | * arm_lms_instance_f32 S = {numTaps, pState, pCoeffs, mu}; |
emh203 | 0:3d9c67d97d6f | 136 | * arm_lms_instance_q31 S = {numTaps, pState, pCoeffs, mu, postShift}; |
emh203 | 0:3d9c67d97d6f | 137 | * arm_lms_instance_q15 S = {numTaps, pState, pCoeffs, mu, postShift}; |
emh203 | 0:3d9c67d97d6f | 138 | * </pre> |
emh203 | 0:3d9c67d97d6f | 139 | * where <code>numTaps</code> is the number of filter coefficients in the filter; <code>pState</code> is the address of the state buffer; |
emh203 | 0:3d9c67d97d6f | 140 | * <code>pCoeffs</code> is the address of the coefficient buffer; <code>mu</code> is the step size parameter; and <code>postShift</code> is the shift applied to coefficients. |
emh203 | 0:3d9c67d97d6f | 141 | * |
emh203 | 0:3d9c67d97d6f | 142 | * \par Fixed-Point Behavior: |
emh203 | 0:3d9c67d97d6f | 143 | * Care must be taken when using the Q15 and Q31 versions of the LMS filter. |
emh203 | 0:3d9c67d97d6f | 144 | * The following issues must be considered: |
emh203 | 0:3d9c67d97d6f | 145 | * - Scaling of coefficients |
emh203 | 0:3d9c67d97d6f | 146 | * - Overflow and saturation |
emh203 | 0:3d9c67d97d6f | 147 | * |
emh203 | 0:3d9c67d97d6f | 148 | * \par Scaling of Coefficients: |
emh203 | 0:3d9c67d97d6f | 149 | * Filter coefficients are represented as fractional values and |
emh203 | 0:3d9c67d97d6f | 150 | * coefficients are restricted to lie in the range <code>[-1 +1)</code>. |
emh203 | 0:3d9c67d97d6f | 151 | * The fixed-point functions have an additional scaling parameter <code>postShift</code>. |
emh203 | 0:3d9c67d97d6f | 152 | * At the output of the filter's accumulator is a shift register which shifts the result by <code>postShift</code> bits. |
emh203 | 0:3d9c67d97d6f | 153 | * This essentially scales the filter coefficients by <code>2^postShift</code> and |
emh203 | 0:3d9c67d97d6f | 154 | * allows the filter coefficients to exceed the range <code>[+1 -1)</code>. |
emh203 | 0:3d9c67d97d6f | 155 | * The value of <code>postShift</code> is set by the user based on the expected gain through the system being modeled. |
emh203 | 0:3d9c67d97d6f | 156 | * |
emh203 | 0:3d9c67d97d6f | 157 | * \par Overflow and Saturation: |
emh203 | 0:3d9c67d97d6f | 158 | * Overflow and saturation behavior of the fixed-point Q15 and Q31 versions are |
emh203 | 0:3d9c67d97d6f | 159 | * described separately as part of the function specific documentation below. |
emh203 | 0:3d9c67d97d6f | 160 | */ |
emh203 | 0:3d9c67d97d6f | 161 | |
emh203 | 0:3d9c67d97d6f | 162 | /** |
emh203 | 0:3d9c67d97d6f | 163 | * @addtogroup LMS |
emh203 | 0:3d9c67d97d6f | 164 | * @{ |
emh203 | 0:3d9c67d97d6f | 165 | */ |
emh203 | 0:3d9c67d97d6f | 166 | |
emh203 | 0:3d9c67d97d6f | 167 | /** |
emh203 | 0:3d9c67d97d6f | 168 | * @details |
emh203 | 0:3d9c67d97d6f | 169 | * This function operates on floating-point data types. |
emh203 | 0:3d9c67d97d6f | 170 | * |
emh203 | 0:3d9c67d97d6f | 171 | * @brief Processing function for floating-point LMS filter. |
emh203 | 0:3d9c67d97d6f | 172 | * @param[in] *S points to an instance of the floating-point LMS filter structure. |
emh203 | 0:3d9c67d97d6f | 173 | * @param[in] *pSrc points to the block of input data. |
emh203 | 0:3d9c67d97d6f | 174 | * @param[in] *pRef points to the block of reference data. |
emh203 | 0:3d9c67d97d6f | 175 | * @param[out] *pOut points to the block of output data. |
emh203 | 0:3d9c67d97d6f | 176 | * @param[out] *pErr points to the block of error data. |
emh203 | 0:3d9c67d97d6f | 177 | * @param[in] blockSize number of samples to process. |
emh203 | 0:3d9c67d97d6f | 178 | * @return none. |
emh203 | 0:3d9c67d97d6f | 179 | */ |
emh203 | 0:3d9c67d97d6f | 180 | |
emh203 | 0:3d9c67d97d6f | 181 | void arm_lms_f32( |
emh203 | 0:3d9c67d97d6f | 182 | const arm_lms_instance_f32 * S, |
emh203 | 0:3d9c67d97d6f | 183 | float32_t * pSrc, |
emh203 | 0:3d9c67d97d6f | 184 | float32_t * pRef, |
emh203 | 0:3d9c67d97d6f | 185 | float32_t * pOut, |
emh203 | 0:3d9c67d97d6f | 186 | float32_t * pErr, |
emh203 | 0:3d9c67d97d6f | 187 | uint32_t blockSize) |
emh203 | 0:3d9c67d97d6f | 188 | { |
emh203 | 0:3d9c67d97d6f | 189 | float32_t *pState = S->pState; /* State pointer */ |
emh203 | 0:3d9c67d97d6f | 190 | float32_t *pCoeffs = S->pCoeffs; /* Coefficient pointer */ |
emh203 | 0:3d9c67d97d6f | 191 | float32_t *pStateCurnt; /* Points to the current sample of the state */ |
emh203 | 0:3d9c67d97d6f | 192 | float32_t *px, *pb; /* Temporary pointers for state and coefficient buffers */ |
emh203 | 0:3d9c67d97d6f | 193 | float32_t mu = S->mu; /* Adaptive factor */ |
emh203 | 0:3d9c67d97d6f | 194 | uint32_t numTaps = S->numTaps; /* Number of filter coefficients in the filter */ |
emh203 | 0:3d9c67d97d6f | 195 | uint32_t tapCnt, blkCnt; /* Loop counters */ |
emh203 | 0:3d9c67d97d6f | 196 | float32_t sum, e, d; /* accumulator, error, reference data sample */ |
emh203 | 0:3d9c67d97d6f | 197 | float32_t w = 0.0f; /* weight factor */ |
emh203 | 0:3d9c67d97d6f | 198 | |
emh203 | 0:3d9c67d97d6f | 199 | e = 0.0f; |
emh203 | 0:3d9c67d97d6f | 200 | d = 0.0f; |
emh203 | 0:3d9c67d97d6f | 201 | |
emh203 | 0:3d9c67d97d6f | 202 | /* S->pState points to state array which contains previous frame (numTaps - 1) samples */ |
emh203 | 0:3d9c67d97d6f | 203 | /* pStateCurnt points to the location where the new input data should be written */ |
emh203 | 0:3d9c67d97d6f | 204 | pStateCurnt = &(S->pState[(numTaps - 1u)]); |
emh203 | 0:3d9c67d97d6f | 205 | |
emh203 | 0:3d9c67d97d6f | 206 | blkCnt = blockSize; |
emh203 | 0:3d9c67d97d6f | 207 | |
emh203 | 0:3d9c67d97d6f | 208 | |
emh203 | 0:3d9c67d97d6f | 209 | #ifndef ARM_MATH_CM0_FAMILY |
emh203 | 0:3d9c67d97d6f | 210 | |
emh203 | 0:3d9c67d97d6f | 211 | /* Run the below code for Cortex-M4 and Cortex-M3 */ |
emh203 | 0:3d9c67d97d6f | 212 | |
emh203 | 0:3d9c67d97d6f | 213 | while(blkCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 214 | { |
emh203 | 0:3d9c67d97d6f | 215 | /* Copy the new input sample into the state buffer */ |
emh203 | 0:3d9c67d97d6f | 216 | *pStateCurnt++ = *pSrc++; |
emh203 | 0:3d9c67d97d6f | 217 | |
emh203 | 0:3d9c67d97d6f | 218 | /* Initialize pState pointer */ |
emh203 | 0:3d9c67d97d6f | 219 | px = pState; |
emh203 | 0:3d9c67d97d6f | 220 | |
emh203 | 0:3d9c67d97d6f | 221 | /* Initialize coeff pointer */ |
emh203 | 0:3d9c67d97d6f | 222 | pb = (pCoeffs); |
emh203 | 0:3d9c67d97d6f | 223 | |
emh203 | 0:3d9c67d97d6f | 224 | /* Set the accumulator to zero */ |
emh203 | 0:3d9c67d97d6f | 225 | sum = 0.0f; |
emh203 | 0:3d9c67d97d6f | 226 | |
emh203 | 0:3d9c67d97d6f | 227 | /* Loop unrolling. Process 4 taps at a time. */ |
emh203 | 0:3d9c67d97d6f | 228 | tapCnt = numTaps >> 2; |
emh203 | 0:3d9c67d97d6f | 229 | |
emh203 | 0:3d9c67d97d6f | 230 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 231 | { |
emh203 | 0:3d9c67d97d6f | 232 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 233 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 234 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 235 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 236 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 237 | |
emh203 | 0:3d9c67d97d6f | 238 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 239 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 240 | } |
emh203 | 0:3d9c67d97d6f | 241 | |
emh203 | 0:3d9c67d97d6f | 242 | /* If the filter length is not a multiple of 4, compute the remaining filter taps */ |
emh203 | 0:3d9c67d97d6f | 243 | tapCnt = numTaps % 0x4u; |
emh203 | 0:3d9c67d97d6f | 244 | |
emh203 | 0:3d9c67d97d6f | 245 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 246 | { |
emh203 | 0:3d9c67d97d6f | 247 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 248 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 249 | |
emh203 | 0:3d9c67d97d6f | 250 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 251 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 252 | } |
emh203 | 0:3d9c67d97d6f | 253 | |
emh203 | 0:3d9c67d97d6f | 254 | /* The result in the accumulator, store in the destination buffer. */ |
emh203 | 0:3d9c67d97d6f | 255 | *pOut++ = sum; |
emh203 | 0:3d9c67d97d6f | 256 | |
emh203 | 0:3d9c67d97d6f | 257 | /* Compute and store error */ |
emh203 | 0:3d9c67d97d6f | 258 | d = (float32_t) (*pRef++); |
emh203 | 0:3d9c67d97d6f | 259 | e = d - sum; |
emh203 | 0:3d9c67d97d6f | 260 | *pErr++ = e; |
emh203 | 0:3d9c67d97d6f | 261 | |
emh203 | 0:3d9c67d97d6f | 262 | /* Calculation of Weighting factor for the updating filter coefficients */ |
emh203 | 0:3d9c67d97d6f | 263 | w = e * mu; |
emh203 | 0:3d9c67d97d6f | 264 | |
emh203 | 0:3d9c67d97d6f | 265 | /* Initialize pState pointer */ |
emh203 | 0:3d9c67d97d6f | 266 | px = pState; |
emh203 | 0:3d9c67d97d6f | 267 | |
emh203 | 0:3d9c67d97d6f | 268 | /* Initialize coeff pointer */ |
emh203 | 0:3d9c67d97d6f | 269 | pb = (pCoeffs); |
emh203 | 0:3d9c67d97d6f | 270 | |
emh203 | 0:3d9c67d97d6f | 271 | /* Loop unrolling. Process 4 taps at a time. */ |
emh203 | 0:3d9c67d97d6f | 272 | tapCnt = numTaps >> 2; |
emh203 | 0:3d9c67d97d6f | 273 | |
emh203 | 0:3d9c67d97d6f | 274 | /* Update filter coefficients */ |
emh203 | 0:3d9c67d97d6f | 275 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 276 | { |
emh203 | 0:3d9c67d97d6f | 277 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 278 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 279 | pb++; |
emh203 | 0:3d9c67d97d6f | 280 | |
emh203 | 0:3d9c67d97d6f | 281 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 282 | pb++; |
emh203 | 0:3d9c67d97d6f | 283 | |
emh203 | 0:3d9c67d97d6f | 284 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 285 | pb++; |
emh203 | 0:3d9c67d97d6f | 286 | |
emh203 | 0:3d9c67d97d6f | 287 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 288 | pb++; |
emh203 | 0:3d9c67d97d6f | 289 | |
emh203 | 0:3d9c67d97d6f | 290 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 291 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 292 | } |
emh203 | 0:3d9c67d97d6f | 293 | |
emh203 | 0:3d9c67d97d6f | 294 | /* If the filter length is not a multiple of 4, compute the remaining filter taps */ |
emh203 | 0:3d9c67d97d6f | 295 | tapCnt = numTaps % 0x4u; |
emh203 | 0:3d9c67d97d6f | 296 | |
emh203 | 0:3d9c67d97d6f | 297 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 298 | { |
emh203 | 0:3d9c67d97d6f | 299 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 300 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 301 | pb++; |
emh203 | 0:3d9c67d97d6f | 302 | |
emh203 | 0:3d9c67d97d6f | 303 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 304 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 305 | } |
emh203 | 0:3d9c67d97d6f | 306 | |
emh203 | 0:3d9c67d97d6f | 307 | /* Advance state pointer by 1 for the next sample */ |
emh203 | 0:3d9c67d97d6f | 308 | pState = pState + 1; |
emh203 | 0:3d9c67d97d6f | 309 | |
emh203 | 0:3d9c67d97d6f | 310 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 311 | blkCnt--; |
emh203 | 0:3d9c67d97d6f | 312 | } |
emh203 | 0:3d9c67d97d6f | 313 | |
emh203 | 0:3d9c67d97d6f | 314 | |
emh203 | 0:3d9c67d97d6f | 315 | /* Processing is complete. Now copy the last numTaps - 1 samples to the |
emh203 | 0:3d9c67d97d6f | 316 | satrt of the state buffer. This prepares the state buffer for the |
emh203 | 0:3d9c67d97d6f | 317 | next function call. */ |
emh203 | 0:3d9c67d97d6f | 318 | |
emh203 | 0:3d9c67d97d6f | 319 | /* Points to the start of the pState buffer */ |
emh203 | 0:3d9c67d97d6f | 320 | pStateCurnt = S->pState; |
emh203 | 0:3d9c67d97d6f | 321 | |
emh203 | 0:3d9c67d97d6f | 322 | /* Loop unrolling for (numTaps - 1u) samples copy */ |
emh203 | 0:3d9c67d97d6f | 323 | tapCnt = (numTaps - 1u) >> 2u; |
emh203 | 0:3d9c67d97d6f | 324 | |
emh203 | 0:3d9c67d97d6f | 325 | /* copy data */ |
emh203 | 0:3d9c67d97d6f | 326 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 327 | { |
emh203 | 0:3d9c67d97d6f | 328 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 329 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 330 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 331 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 332 | |
emh203 | 0:3d9c67d97d6f | 333 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 334 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 335 | } |
emh203 | 0:3d9c67d97d6f | 336 | |
emh203 | 0:3d9c67d97d6f | 337 | /* Calculate remaining number of copies */ |
emh203 | 0:3d9c67d97d6f | 338 | tapCnt = (numTaps - 1u) % 0x4u; |
emh203 | 0:3d9c67d97d6f | 339 | |
emh203 | 0:3d9c67d97d6f | 340 | /* Copy the remaining q31_t data */ |
emh203 | 0:3d9c67d97d6f | 341 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 342 | { |
emh203 | 0:3d9c67d97d6f | 343 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 344 | |
emh203 | 0:3d9c67d97d6f | 345 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 346 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 347 | } |
emh203 | 0:3d9c67d97d6f | 348 | |
emh203 | 0:3d9c67d97d6f | 349 | #else |
emh203 | 0:3d9c67d97d6f | 350 | |
emh203 | 0:3d9c67d97d6f | 351 | /* Run the below code for Cortex-M0 */ |
emh203 | 0:3d9c67d97d6f | 352 | |
emh203 | 0:3d9c67d97d6f | 353 | while(blkCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 354 | { |
emh203 | 0:3d9c67d97d6f | 355 | /* Copy the new input sample into the state buffer */ |
emh203 | 0:3d9c67d97d6f | 356 | *pStateCurnt++ = *pSrc++; |
emh203 | 0:3d9c67d97d6f | 357 | |
emh203 | 0:3d9c67d97d6f | 358 | /* Initialize pState pointer */ |
emh203 | 0:3d9c67d97d6f | 359 | px = pState; |
emh203 | 0:3d9c67d97d6f | 360 | |
emh203 | 0:3d9c67d97d6f | 361 | /* Initialize pCoeffs pointer */ |
emh203 | 0:3d9c67d97d6f | 362 | pb = pCoeffs; |
emh203 | 0:3d9c67d97d6f | 363 | |
emh203 | 0:3d9c67d97d6f | 364 | /* Set the accumulator to zero */ |
emh203 | 0:3d9c67d97d6f | 365 | sum = 0.0f; |
emh203 | 0:3d9c67d97d6f | 366 | |
emh203 | 0:3d9c67d97d6f | 367 | /* Loop over numTaps number of values */ |
emh203 | 0:3d9c67d97d6f | 368 | tapCnt = numTaps; |
emh203 | 0:3d9c67d97d6f | 369 | |
emh203 | 0:3d9c67d97d6f | 370 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 371 | { |
emh203 | 0:3d9c67d97d6f | 372 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 373 | sum += (*px++) * (*pb++); |
emh203 | 0:3d9c67d97d6f | 374 | |
emh203 | 0:3d9c67d97d6f | 375 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 376 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 377 | } |
emh203 | 0:3d9c67d97d6f | 378 | |
emh203 | 0:3d9c67d97d6f | 379 | /* The result is stored in the destination buffer. */ |
emh203 | 0:3d9c67d97d6f | 380 | *pOut++ = sum; |
emh203 | 0:3d9c67d97d6f | 381 | |
emh203 | 0:3d9c67d97d6f | 382 | /* Compute and store error */ |
emh203 | 0:3d9c67d97d6f | 383 | d = (float32_t) (*pRef++); |
emh203 | 0:3d9c67d97d6f | 384 | e = d - sum; |
emh203 | 0:3d9c67d97d6f | 385 | *pErr++ = e; |
emh203 | 0:3d9c67d97d6f | 386 | |
emh203 | 0:3d9c67d97d6f | 387 | /* Weighting factor for the LMS version */ |
emh203 | 0:3d9c67d97d6f | 388 | w = e * mu; |
emh203 | 0:3d9c67d97d6f | 389 | |
emh203 | 0:3d9c67d97d6f | 390 | /* Initialize pState pointer */ |
emh203 | 0:3d9c67d97d6f | 391 | px = pState; |
emh203 | 0:3d9c67d97d6f | 392 | |
emh203 | 0:3d9c67d97d6f | 393 | /* Initialize pCoeffs pointer */ |
emh203 | 0:3d9c67d97d6f | 394 | pb = pCoeffs; |
emh203 | 0:3d9c67d97d6f | 395 | |
emh203 | 0:3d9c67d97d6f | 396 | /* Loop over numTaps number of values */ |
emh203 | 0:3d9c67d97d6f | 397 | tapCnt = numTaps; |
emh203 | 0:3d9c67d97d6f | 398 | |
emh203 | 0:3d9c67d97d6f | 399 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 400 | { |
emh203 | 0:3d9c67d97d6f | 401 | /* Perform the multiply-accumulate */ |
emh203 | 0:3d9c67d97d6f | 402 | *pb = *pb + (w * (*px++)); |
emh203 | 0:3d9c67d97d6f | 403 | pb++; |
emh203 | 0:3d9c67d97d6f | 404 | |
emh203 | 0:3d9c67d97d6f | 405 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 406 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 407 | } |
emh203 | 0:3d9c67d97d6f | 408 | |
emh203 | 0:3d9c67d97d6f | 409 | /* Advance state pointer by 1 for the next sample */ |
emh203 | 0:3d9c67d97d6f | 410 | pState = pState + 1; |
emh203 | 0:3d9c67d97d6f | 411 | |
emh203 | 0:3d9c67d97d6f | 412 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 413 | blkCnt--; |
emh203 | 0:3d9c67d97d6f | 414 | } |
emh203 | 0:3d9c67d97d6f | 415 | |
emh203 | 0:3d9c67d97d6f | 416 | |
emh203 | 0:3d9c67d97d6f | 417 | /* Processing is complete. Now copy the last numTaps - 1 samples to the |
emh203 | 0:3d9c67d97d6f | 418 | * start of the state buffer. This prepares the state buffer for the |
emh203 | 0:3d9c67d97d6f | 419 | * next function call. */ |
emh203 | 0:3d9c67d97d6f | 420 | |
emh203 | 0:3d9c67d97d6f | 421 | /* Points to the start of the pState buffer */ |
emh203 | 0:3d9c67d97d6f | 422 | pStateCurnt = S->pState; |
emh203 | 0:3d9c67d97d6f | 423 | |
emh203 | 0:3d9c67d97d6f | 424 | /* Copy (numTaps - 1u) samples */ |
emh203 | 0:3d9c67d97d6f | 425 | tapCnt = (numTaps - 1u); |
emh203 | 0:3d9c67d97d6f | 426 | |
emh203 | 0:3d9c67d97d6f | 427 | /* Copy the data */ |
emh203 | 0:3d9c67d97d6f | 428 | while(tapCnt > 0u) |
emh203 | 0:3d9c67d97d6f | 429 | { |
emh203 | 0:3d9c67d97d6f | 430 | *pStateCurnt++ = *pState++; |
emh203 | 0:3d9c67d97d6f | 431 | |
emh203 | 0:3d9c67d97d6f | 432 | /* Decrement the loop counter */ |
emh203 | 0:3d9c67d97d6f | 433 | tapCnt--; |
emh203 | 0:3d9c67d97d6f | 434 | } |
emh203 | 0:3d9c67d97d6f | 435 | |
emh203 | 0:3d9c67d97d6f | 436 | #endif /* #ifndef ARM_MATH_CM0_FAMILY */ |
emh203 | 0:3d9c67d97d6f | 437 | |
emh203 | 0:3d9c67d97d6f | 438 | } |
emh203 | 0:3d9c67d97d6f | 439 | |
emh203 | 0:3d9c67d97d6f | 440 | /** |
emh203 | 0:3d9c67d97d6f | 441 | * @} end of LMS group |
emh203 | 0:3d9c67d97d6f | 442 | */ |