From 96d6da4e252b06dcfdc041e7df23e86161c33007 Mon Sep 17 00:00:00 2001
From: rihab kouki <rihab.kouki@st.com>
Date: Tue, 28 Jul 2020 11:24:49 +0100
Subject: Official ARM version: v5.6.0

---
 DSP/Source/ComplexMathFunctions/CMakeLists.txt     |  16 ++
 .../ComplexMathFunctions/ComplexMathFunctions.c    |  46 ++++
 .../ComplexMathFunctions/arm_cmplx_conj_f32.c      | 182 ++++++-------
 .../ComplexMathFunctions/arm_cmplx_conj_q15.c      | 158 +++++------
 .../ComplexMathFunctions/arm_cmplx_conj_q31.c      | 166 +++++-------
 .../ComplexMathFunctions/arm_cmplx_dot_prod_f32.c  | 298 ++++++++++++---------
 .../ComplexMathFunctions/arm_cmplx_dot_prod_q15.c  | 207 +++++++-------
 .../ComplexMathFunctions/arm_cmplx_dot_prod_q31.c  | 208 +++++++-------
 .../ComplexMathFunctions/arm_cmplx_mag_f32.c       | 191 +++++++------
 .../ComplexMathFunctions/arm_cmplx_mag_q15.c       | 163 ++++++-----
 .../ComplexMathFunctions/arm_cmplx_mag_q31.c       | 165 +++++-------
 .../arm_cmplx_mag_squared_f32.c                    | 216 +++++++--------
 .../arm_cmplx_mag_squared_q15.c                    | 153 ++++++-----
 .../arm_cmplx_mag_squared_q31.c                    |  96 +++----
 .../arm_cmplx_mult_cmplx_f32.c                     | 248 +++++++++--------
 .../arm_cmplx_mult_cmplx_q15.c                     | 147 ++++------
 .../arm_cmplx_mult_cmplx_q31.c                     | 273 ++++---------------
 .../ComplexMathFunctions/arm_cmplx_mult_real_f32.c | 230 +++++++---------
 .../ComplexMathFunctions/arm_cmplx_mult_real_q15.c | 175 ++++++------
 .../ComplexMathFunctions/arm_cmplx_mult_real_q31.c | 239 ++++++-----------
 20 files changed, 1630 insertions(+), 1947 deletions(-)
 create mode 100644 DSP/Source/ComplexMathFunctions/CMakeLists.txt
 create mode 100644 DSP/Source/ComplexMathFunctions/ComplexMathFunctions.c

(limited to 'DSP/Source/ComplexMathFunctions')

diff --git a/DSP/Source/ComplexMathFunctions/CMakeLists.txt b/DSP/Source/ComplexMathFunctions/CMakeLists.txt
new file mode 100644
index 0000000..16e06c6
--- /dev/null
+++ b/DSP/Source/ComplexMathFunctions/CMakeLists.txt
@@ -0,0 +1,16 @@
+cmake_minimum_required (VERSION 3.6)
+
+project(CMSISDSPComplexMath)
+
+
+file(GLOB SRC "./*_*.c")
+
+add_library(CMSISDSPComplexMath STATIC ${SRC})
+
+configdsp(CMSISDSPComplexMath ..)
+
+### Includes
+target_include_directories(CMSISDSPComplexMath PUBLIC "${DSP}/../../Include")
+
+
+
diff --git a/DSP/Source/ComplexMathFunctions/ComplexMathFunctions.c b/DSP/Source/ComplexMathFunctions/ComplexMathFunctions.c
new file mode 100644
index 0000000..2210533
--- /dev/null
+++ b/DSP/Source/ComplexMathFunctions/ComplexMathFunctions.c
@@ -0,0 +1,46 @@
+/* ----------------------------------------------------------------------
+ * Project:      CMSIS DSP Library
+ * Title:        CompexMathFunctions.c
+ * Description:  Combination of all comlex math function source files.
+ *
+ * $Date:        18. March 2019
+ * $Revision:    V1.0.0
+ *
+ * Target Processor: Cortex-M cores
+ * -------------------------------------------------------------------- */
+/*
+ * Copyright (C) 2019 ARM Limited or its affiliates. All rights reserved.
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ *
+ * Licensed under the Apache License, Version 2.0 (the License); you may
+ * not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an AS IS BASIS, WITHOUT
+ * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include "arm_cmplx_conj_f32.c"
+#include "arm_cmplx_conj_q15.c"
+#include "arm_cmplx_conj_q31.c"
+#include "arm_cmplx_dot_prod_f32.c"
+#include "arm_cmplx_dot_prod_q15.c"
+#include "arm_cmplx_dot_prod_q31.c"
+#include "arm_cmplx_mag_f32.c"
+#include "arm_cmplx_mag_q15.c"
+#include "arm_cmplx_mag_q31.c"
+#include "arm_cmplx_mag_squared_f32.c"
+#include "arm_cmplx_mag_squared_q15.c"
+#include "arm_cmplx_mag_squared_q31.c"
+#include "arm_cmplx_mult_cmplx_f32.c"
+#include "arm_cmplx_mult_cmplx_q15.c"
+#include "arm_cmplx_mult_cmplx_q31.c"
+#include "arm_cmplx_mult_real_f32.c"
+#include "arm_cmplx_mult_real_q15.c"
+#include "arm_cmplx_mult_real_q31.c"
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_f32.c
index cfb6f1f..df5db00 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_conj_f32.c
  * Description:  Floating-point complex conjugate
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,143 +29,133 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup cmplx_conj Complex Conjugate
- *
- * Conjugates the elements of a complex data vector.
- *
- * The <code>pSrc</code> points to the source data and
- * <code>pDst</code> points to the where the result should be written.
- * <code>numSamples</code> specifies the number of complex samples
- * and the data in each array is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * Each array has a total of <code>2*numSamples</code> values.
- * The underlying algorithm is used:
- *
- * <pre>
- * for(n=0; n<numSamples; n++) {
- *     pDst[(2*n)+0)] = pSrc[(2*n)+0];     // real part
- *     pDst[(2*n)+1)] = -pSrc[(2*n)+1];    // imag part
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup cmplx_conj Complex Conjugate
+
+  Conjugates the elements of a complex data vector.
+
+  The <code>pSrc</code> points to the source data and
+  <code>pDst</code> points to the destination data where the result should be written.
+  <code>numSamples</code> specifies the number of complex samples
+  and the data in each array is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  Each array has a total of <code>2*numSamples</code> values.
+
+  The underlying algorithm is used:
+  <pre>
+  for (n = 0; n < numSamples; n++) {
+      pDst[(2*n)  ] =  pSrc[(2*n)  ];    // real part
+      pDst[(2*n)+1] = -pSrc[(2*n)+1];    // imag part
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup cmplx_conj
- * @{
+  @addtogroup cmplx_conj
+  @{
  */
 
 /**
- * @brief  Floating-point complex conjugate.
- * @param  *pSrc points to the input vector
- * @param  *pDst points to the output vector
- * @param  numSamples number of complex samples in each vector
- * @return none.
+  @brief         Floating-point complex conjugate.
+  @param[in]     pSrc        points to the input vector
+  @param[out]    pDst        points to the output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
  */
 
+
 void arm_cmplx_conj_f32(
-  float32_t * pSrc,
-  float32_t * pDst,
-  uint32_t numSamples)
+  const float32_t * pSrc,
+        float32_t * pDst,
+        uint32_t numSamples)
 {
-  uint32_t blkCnt;                               /* loop counter */
-
-#if defined (ARM_MATH_DSP)
-
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  float32_t inR1, inR2, inR3, inR4;
-  float32_t inI1, inI2, inI3, inI4;
+        uint32_t blkCnt;                               /* Loop counter */
 
-  /*loop Unrolling */
-  blkCnt = numSamples >> 2U;
+#if defined(ARM_MATH_NEON)
+   float32x4_t zero;
+   float32x4x2_t vec;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
-  while (blkCnt > 0U)
-  {
-    /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    /* read real input samples */
-    inR1 = pSrc[0];
-    /* store real samples to destination */
-    pDst[0] = inR1;
-    inR2 = pSrc[2];
-    pDst[2] = inR2;
-    inR3 = pSrc[4];
-    pDst[4] = inR3;
-    inR4 = pSrc[6];
-    pDst[6] = inR4;
+   zero = vdupq_n_f32(0.0);
 
-    /* read imaginary input samples */
-    inI1 = pSrc[1];
-    inI2 = pSrc[3];
+   /* Compute 4 outputs at a time */
+   blkCnt = numSamples >> 2U;
 
-    /* conjugate input */
-    inI1 = -inI1;
+   while (blkCnt > 0U)
+   {
+     /* C[0]+jC[1] = A[0]+(-1)*jA[1] */
+     /* Calculate Complex Conjugate and then store the results in the destination buffer. */
+     vec = vld2q_f32(pSrc);
+     vec.val[1] = vsubq_f32(zero,vec.val[1]);
+     vst2q_f32(pDst,vec);
 
-    /* read imaginary input samples */
-    inI3 = pSrc[5];
+     /* Increment pointers */
+     pSrc += 8;
+     pDst += 8;
+        
+     /* Decrement the loop counter */
+     blkCnt--;
+   }
 
-    /* conjugate input */
-    inI2 = -inI2;
+   /* Tail */
+   blkCnt = numSamples & 0x3;
 
-    /* read imaginary input samples */
-    inI4 = pSrc[7];
-
-    /* conjugate input */
-    inI3 = -inI3;
+#else
+#if defined (ARM_MATH_LOOPUNROLL)
 
-    /* store imaginary samples to destination */
-    pDst[1] = inI1;
-    pDst[3] = inI2;
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
-    /* conjugate input */
-    inI4 = -inI4;
+  while (blkCnt > 0U)
+  {
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
 
-    /* store imaginary samples to destination */
-    pDst[5] = inI3;
+    /* Calculate Complex Conjugate and store result in destination buffer. */
+    *pDst++ =  *pSrc++;
+    *pDst++ = -*pSrc++;
 
-    /* increment source pointer by 8 to process next sampels */
-    pSrc += 8U;
+    *pDst++ =  *pSrc++;
+    *pDst++ = -*pSrc++;
 
-    /* store imaginary sample to destination */
-    pDst[7] = inI4;
+    *pDst++ =  *pSrc++;
+    *pDst++ = -*pSrc++;
 
-    /* increment destination pointer by 8 to store next samples */
-    pDst += 8U;
+    *pDst++ =  *pSrc++;
+    *pDst++ = -*pSrc++;
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined (ARM_MATH_NEON) */
 
   while (blkCnt > 0U)
   {
-    /* realOut + j (imagOut) = realIn + j (-1) imagIn */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    *pDst++ = *pSrc++;
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
+
+    /* Calculate Complex Conjugate and store result in destination buffer. */
+    *pDst++ =  *pSrc++;
     *pDst++ = -*pSrc++;
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of cmplx_conj group
+  @} end of cmplx_conj group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q15.c
index 7950229..073a337 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_conj_q15.c
  * Description:  Q15 complex conjugate
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,68 +29,66 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_conj
- * @{
+  @addtogroup cmplx_conj
+  @{
  */
 
 /**
- * @brief  Q15 complex conjugate.
- * @param  *pSrc points to the input vector
- * @param  *pDst points to the output vector
- * @param  numSamples number of complex samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function uses saturating arithmetic.
- * The Q15 value -1 (0x8000) will be saturated to the maximum allowable positive value 0x7FFF.
+  @brief         Q15 complex conjugate.
+  @param[in]     pSrc        points to the input vector
+  @param[out]    pDst        points to the output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function uses saturating arithmetic.
+                   The Q15 value -1 (0x8000) is saturated to the maximum allowable positive value 0x7FFF.
  */
 
 void arm_cmplx_conj_q15(
-  q15_t * pSrc,
-  q15_t * pDst,
-  uint32_t numSamples)
+  const q15_t * pSrc,
+        q15_t * pDst,
+        uint32_t numSamples)
 {
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t in1;                                     /* Temporary input variable */
 
-#if defined (ARM_MATH_DSP)
+#if defined (ARM_MATH_LOOPUNROLL) && defined (ARM_MATH_DSP)
+        q31_t in2, in3, in4;                           /* Temporary input variables */
+#endif
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
-  q31_t in1, in2, in3, in4;
-  q31_t zero = 0;
 
-  /*loop Unrolling */
+#if defined (ARM_MATH_LOOPUNROLL)
+
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    in1 = *__SIMD32(pSrc)++;
-    in2 = *__SIMD32(pSrc)++;
-    in3 = *__SIMD32(pSrc)++;
-    in4 = *__SIMD32(pSrc)++;
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
 
-#ifndef ARM_MATH_BIG_ENDIAN
+    /* Calculate Complex Conjugate and store result in destination buffer. */
 
-    in1 = __QASX(zero, in1);
-    in2 = __QASX(zero, in2);
-    in3 = __QASX(zero, in3);
-    in4 = __QASX(zero, in4);
+    #if defined (ARM_MATH_DSP)
+    in1 = read_q15x2_ia ((q15_t **) &pSrc);
+    in2 = read_q15x2_ia ((q15_t **) &pSrc);
+    in3 = read_q15x2_ia ((q15_t **) &pSrc);
+    in4 = read_q15x2_ia ((q15_t **) &pSrc);
 
+#ifndef ARM_MATH_BIG_ENDIAN
+    in1 = __QASX(0, in1);
+    in2 = __QASX(0, in2);
+    in3 = __QASX(0, in3);
+    in4 = __QASX(0, in4);
 #else
-
-    in1 = __QSAX(zero, in1);
-    in2 = __QSAX(zero, in2);
-    in3 = __QSAX(zero, in3);
-    in4 = __QSAX(zero, in4);
-
+    in1 = __QSAX(0, in1);
+    in2 = __QSAX(0, in2);
+    in3 = __QSAX(0, in3);
+    in4 = __QSAX(0, in4);
 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
 
     in1 = ((uint32_t) in1 >> 16) | ((uint32_t) in1 << 16);
@@ -98,52 +96,62 @@ void arm_cmplx_conj_q15(
     in3 = ((uint32_t) in3 >> 16) | ((uint32_t) in3 << 16);
     in4 = ((uint32_t) in4 >> 16) | ((uint32_t) in4 << 16);
 
-    *__SIMD32(pDst)++ = in1;
-    *__SIMD32(pDst)++ = in2;
-    *__SIMD32(pDst)++ = in3;
-    *__SIMD32(pDst)++ = in4;
+    write_q15x2_ia (&pDst, in1);
+    write_q15x2_ia (&pDst, in2);
+    write_q15x2_ia (&pDst, in3);
+    write_q15x2_ia (&pDst, in4);
+#else
+    *pDst++ =  *pSrc++;
+    in1 = *pSrc++;
+    *pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
 
-    /* Decrement the loop counter */
-    blkCnt--;
-  }
+    *pDst++ =  *pSrc++;
+    in1 = *pSrc++;
+    *pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x4U;
+    *pDst++ =  *pSrc++;
+    in1 = *pSrc++;
+    *pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
 
-  while (blkCnt > 0U)
-  {
-    /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    *pDst++ = *pSrc++;
-    *pDst++ = __SSAT(-*pSrc++, 16);
+    *pDst++ =  *pSrc++;
+    in1 = *pSrc++;
+    *pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
 
-    /* Decrement the loop counter */
+#endif /* #if defined (ARM_MATH_DSP) */
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  q15_t in;
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  /* Run the below code for Cortex-M0 */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-  while (numSamples > 0U)
+  while (blkCnt > 0U)
   {
-    /* realOut + j (imagOut) = realIn+ j (-1) imagIn */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    *pDst++ = *pSrc++;
-    in = *pSrc++;
-    *pDst++ = (in == (q15_t) 0x8000) ? 0x7fff : -in;
-
-    /* Decrement the loop counter */
-    numSamples--;
-  }
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
 
-#endif /* #if defined (ARM_MATH_DSP) */
+    /* Calculate Complex Conjugate and store result in destination buffer. */
+    *pDst++ =  *pSrc++;
+    in1 = *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __SSAT(-in1, 16);
+#else
+    *pDst++ = (in1 == (q15_t) 0x8000) ? (q15_t) 0x7fff : -in1;
+#endif
+
+    /* Decrement loop counter */
+    blkCnt--;
+  }
 
 }
 
 /**
- * @} end of cmplx_conj group
+  @} end of cmplx_conj group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q31.c
index 709ce0e..6ef1ddb 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_conj_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_conj_q31.c
  * Description:  Q31 complex conjugate
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,141 +29,109 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_conj
- * @{
+  @addtogroup cmplx_conj
+  @{
  */
 
 /**
- * @brief  Q31 complex conjugate.
- * @param  *pSrc points to the input vector
- * @param  *pDst points to the output vector
- * @param  numSamples number of complex samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function uses saturating arithmetic.
- * The Q31 value -1 (0x80000000) will be saturated to the maximum allowable positive value 0x7FFFFFFF.
+  @brief         Q31 complex conjugate.
+  @param[in]     pSrc        points to the input vector
+  @param[out]    pDst        points to the output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function uses saturating arithmetic.
+                   The Q31 value -1 (0x80000000) is saturated to the maximum allowable positive value 0x7FFFFFFF.
  */
 
 void arm_cmplx_conj_q31(
-  q31_t * pSrc,
-  q31_t * pDst,
-  uint32_t numSamples)
+  const q31_t * pSrc,
+        q31_t * pDst,
+        uint32_t numSamples)
 {
-  uint32_t blkCnt;                               /* loop counter */
-  q31_t in;                                      /* Input value */
-
-#if defined (ARM_MATH_DSP)
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t in;                                      /* Temporary input variable */
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  q31_t inR1, inR2, inR3, inR4;                  /* Temporary real variables */
-  q31_t inI1, inI2, inI3, inI4;                  /* Temporary imaginary variables */
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    /* Saturated to 0x7fffffff if the input is -1(0x80000000) */
-    /* read real input sample */
-    inR1 = pSrc[0];
-    /* store real input sample */
-    pDst[0] = inR1;
-
-    /* read imaginary input sample */
-    inI1 = pSrc[1];
-
-    /* read real input sample */
-    inR2 = pSrc[2];
-    /* store real input sample */
-    pDst[2] = inR2;
-
-    /* read imaginary input sample */
-    inI2 = pSrc[3];
-
-    /* negate imaginary input sample */
-    inI1 = __QSUB(0, inI1);
-
-    /* read real input sample */
-    inR3 = pSrc[4];
-    /* store real input sample */
-    pDst[4] = inR3;
-
-    /* read imaginary input sample */
-    inI3 = pSrc[5];
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
 
-    /* negate imaginary input sample */
-    inI2 = __QSUB(0, inI2);
-
-    /* read real input sample */
-    inR4 = pSrc[6];
-    /* store real input sample */
-    pDst[6] = inR4;
-
-    /* negate imaginary input sample */
-    inI3 = __QSUB(0, inI3);
-
-    /* store imaginary input sample */
-    inI4 = pSrc[7];
-
-    /* store imaginary input samples */
-    pDst[1] = inI1;
-
-    /* negate imaginary input sample */
-    inI4 = __QSUB(0, inI4);
-
-    /* store imaginary input samples */
-    pDst[3] = inI2;
+    /* Calculate Complex Conjugate and store result in destination buffer. */
+    *pDst++ =  *pSrc++;
+    in = *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __QSUB(0, in);
+#else
+    *pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
+#endif
 
-    /* increment source pointer by 8 to proecess next samples */
-    pSrc += 8U;
+    *pDst++ =  *pSrc++;
+    in =  *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __QSUB(0, in);
+#else
+    *pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
+#endif
 
-    /* store imaginary input samples */
-    pDst[5] = inI3;
-    pDst[7] = inI4;
+    *pDst++ =  *pSrc++;
+    in = *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __QSUB(0, in);
+#else
+    *pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
+#endif
 
-    /* increment destination pointer by 8 to process next samples */
-    pDst += 8U;
+    *pDst++ =  *pSrc++;
+    in = *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __QSUB(0, in);
+#else
+    *pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
+#endif
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
   while (blkCnt > 0U)
   {
-    /* C[0]+jC[1] = A[0]+ j (-1) A[1] */
-    /* Calculate Complex Conjugate and then store the results in the destination buffer. */
-    /* Saturated to 0x7fffffff if the input is -1(0x80000000) */
-    *pDst++ = *pSrc++;
+    /* C[0] + jC[1] = A[0]+ j(-1)A[1] */
+
+    /* Calculate Complex Conjugate and store result in destination buffer. */
+    *pDst++ =  *pSrc++;
     in = *pSrc++;
+#if defined (ARM_MATH_DSP)
+    *pDst++ = __QSUB(0, in);
+#else
     *pDst++ = (in == INT32_MIN) ? INT32_MAX : -in;
+#endif
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of cmplx_conj group
+  @} end of cmplx_conj group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_f32.c
index bfc352b..06f1bfa 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_dot_prod_f32.c
  * Description:  Floating-point complex dot product
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,163 +29,205 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup cmplx_dot_prod Complex Dot Product
- *
- * Computes the dot product of two complex vectors.
- * The vectors are multiplied element-by-element and then summed.
- *
- * The <code>pSrcA</code> points to the first complex input vector and
- * <code>pSrcB</code> points to the second complex input vector.
- * <code>numSamples</code> specifies the number of complex samples
- * and the data in each array is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * Each array has a total of <code>2*numSamples</code> values.
- *
- * The underlying algorithm is used:
- * <pre>
- * realResult=0;
- * imagResult=0;
- * for(n=0; n<numSamples; n++) {
- *     realResult += pSrcA[(2*n)+0]*pSrcB[(2*n)+0] - pSrcA[(2*n)+1]*pSrcB[(2*n)+1];
- *     imagResult += pSrcA[(2*n)+0]*pSrcB[(2*n)+1] + pSrcA[(2*n)+1]*pSrcB[(2*n)+0];
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup cmplx_dot_prod Complex Dot Product
+
+  Computes the dot product of two complex vectors.
+  The vectors are multiplied element-by-element and then summed.
+
+  The <code>pSrcA</code> points to the first complex input vector and
+  <code>pSrcB</code> points to the second complex input vector.
+  <code>numSamples</code> specifies the number of complex samples
+  and the data in each array is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  Each array has a total of <code>2*numSamples</code> values.
+
+  The underlying algorithm is used:
+
+  <pre>
+  realResult = 0;
+  imagResult = 0;
+  for (n = 0; n < numSamples; n++) {
+      realResult += pSrcA[(2*n)+0] * pSrcB[(2*n)+0] - pSrcA[(2*n)+1] * pSrcB[(2*n)+1];
+      imagResult += pSrcA[(2*n)+0] * pSrcB[(2*n)+1] + pSrcA[(2*n)+1] * pSrcB[(2*n)+0];
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup cmplx_dot_prod
- * @{
+  @addtogroup cmplx_dot_prod
+  @{
  */
 
 /**
- * @brief  Floating-point complex dot product
- * @param  *pSrcA points to the first input vector
- * @param  *pSrcB points to the second input vector
- * @param  numSamples number of complex samples in each vector
- * @param  *realResult real part of the result returned here
- * @param  *imagResult imaginary part of the result returned here
- * @return none.
+  @brief         Floating-point complex dot product.
+  @param[in]     pSrcA       points to the first input vector
+  @param[in]     pSrcB       points to the second input vector
+  @param[in]     numSamples  number of samples in each vector
+  @param[out]    realResult  real part of the result returned here
+  @param[out]    imagResult  imaginary part of the result returned here
+  @return        none
  */
 
 void arm_cmplx_dot_prod_f32(
-  float32_t * pSrcA,
-  float32_t * pSrcB,
-  uint32_t numSamples,
-  float32_t * realResult,
-  float32_t * imagResult)
+  const float32_t * pSrcA,
+  const float32_t * pSrcB,
+        uint32_t numSamples,
+        float32_t * realResult,
+        float32_t * imagResult)
 {
-  float32_t real_sum = 0.0f, imag_sum = 0.0f;    /* Temporary result storage */
-  float32_t a0,b0,c0,d0;
+        uint32_t blkCnt;                               /* Loop counter */
+        float32_t real_sum = 0.0f, imag_sum = 0.0f;    /* Temporary result variables */
+        float32_t a0,b0,c0,d0;
 
-#if defined (ARM_MATH_DSP)
+#if defined(ARM_MATH_NEON)
+    float32x4x2_t vec1,vec2,vec3,vec4;
+    float32x4_t accR,accI;
+    float32x2_t accum = vdup_n_f32(0);
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
+    accR = vdupq_n_f32(0.0);
+    accI = vdupq_n_f32(0.0);
 
-  /*loop Unrolling */
-  blkCnt = numSamples >> 2U;
+    /* Loop unrolling: Compute 8 outputs at a time */
+    blkCnt = numSamples >> 3U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
-  while (blkCnt > 0U)
-  {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      /* Decrement the loop counter */
-      blkCnt--;
-  }
+    while (blkCnt > 0U)
+    {
+	/* C = (A[0]+jA[1])*(B[0]+jB[1]) + ...  */
+        /* Calculate dot product and then store the result in a temporary buffer. */
+
+	vec1 = vld2q_f32(pSrcA);
+        vec2 = vld2q_f32(pSrcB);
+
+	/* Increment pointers */
+        pSrcA += 8;
+        pSrcB += 8;
+
+	/* Re{C} = Re{A}*Re{B} - Im{A}*Im{B} */
+        accR = vmlaq_f32(accR,vec1.val[0],vec2.val[0]);
+        accR = vmlsq_f32(accR,vec1.val[1],vec2.val[1]);
+
+	/* Im{C} = Re{A}*Im{B} + Im{A}*Re{B} */
+        accI = vmlaq_f32(accI,vec1.val[1],vec2.val[0]);
+        accI = vmlaq_f32(accI,vec1.val[0],vec2.val[1]);
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples & 0x3U;
+        vec3 = vld2q_f32(pSrcA);
+        vec4 = vld2q_f32(pSrcB);
+	
+	/* Increment pointers */
+        pSrcA += 8;
+        pSrcB += 8;
+
+	/* Re{C} = Re{A}*Re{B} - Im{A}*Im{B} */
+        accR = vmlaq_f32(accR,vec3.val[0],vec4.val[0]);
+        accR = vmlsq_f32(accR,vec3.val[1],vec4.val[1]);
+
+	/* Im{C} = Re{A}*Im{B} + Im{A}*Re{B} */
+        accI = vmlaq_f32(accI,vec3.val[1],vec4.val[0]);
+        accI = vmlaq_f32(accI,vec3.val[0],vec4.val[1]);
+
+        /* Decrement the loop counter */
+        blkCnt--;
+    }
+
+    accum = vpadd_f32(vget_low_f32(accR), vget_high_f32(accR));
+    real_sum += accum[0] + accum[1];
+
+    accum = vpadd_f32(vget_low_f32(accI), vget_high_f32(accI));
+    imag_sum += accum[0] + accum[1];
+
+    /* Tail */
+    blkCnt = numSamples & 0x7;
+
+#else
+#if defined (ARM_MATH_LOOPUNROLL)
+
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
   while (blkCnt > 0U)
   {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      /* Decrement the loop counter */
-      blkCnt--;
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += a0 * c0;
+    imag_sum += a0 * d0;
+    real_sum -= b0 * d0;
+    imag_sum += b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += a0 * c0;
+    imag_sum += a0 * d0;
+    real_sum -= b0 * d0;
+    imag_sum += b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += a0 * c0;
+    imag_sum += a0 * d0;
+    real_sum -= b0 * d0;
+    imag_sum += b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += a0 * c0;
+    imag_sum += a0 * d0;
+    real_sum -= b0 * d0;
+    imag_sum += b0 * c0;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  while (numSamples > 0U)
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined(ARM_MATH_NEON) */
+
+  while (blkCnt > 0U)
   {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
-
-      /* Decrement the loop counter */
-      numSamples--;
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += a0 * c0;
+    imag_sum += a0 * d0;
+    real_sum -= b0 * d0;
+    imag_sum += b0 * c0;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
-  /* Store the real and imaginary results in the destination buffers */
+  /* Store real and imaginary result in destination buffer. */
   *realResult = real_sum;
   *imagResult = imag_sum;
 }
 
 /**
- * @} end of cmplx_dot_prod group
+  @} end of cmplx_dot_prod group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q15.c
index 9e23a01..2ecd801 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_dot_prod_q15.c
  * Description:  Processing function for the Q15 Complex Dot product
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,143 +29,120 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_dot_prod
- * @{
+  @addtogroup cmplx_dot_prod
+  @{
  */
 
 /**
- * @brief  Q15 complex dot product
- * @param  *pSrcA points to the first input vector
- * @param  *pSrcB points to the second input vector
- * @param  numSamples number of complex samples in each vector
- * @param  *realResult real part of the result returned here
- * @param  *imagResult imaginary part of the result returned here
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function is implemented using an internal 64-bit accumulator.
- * The intermediate 1.15 by 1.15 multiplications are performed with full precision and yield a 2.30 result.
- * These are accumulated in a 64-bit accumulator with 34.30 precision.
- * As a final step, the accumulators are converted to 8.24 format.
- * The return results <code>realResult</code> and <code>imagResult</code> are in 8.24 format.
+  @brief         Q15 complex dot product.
+  @param[in]     pSrcA       points to the first input vector
+  @param[in]     pSrcB       points to the second input vector
+  @param[in]     numSamples  number of samples in each vector
+  @param[out]    realResult  real part of the result returned here
+  @param[out]    imagResult  imaginary part of the result returned her
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function is implemented using an internal 64-bit accumulator.
+                   The intermediate 1.15 by 1.15 multiplications are performed with full precision and yield a 2.30 result.
+                   These are accumulated in a 64-bit accumulator with 34.30 precision.
+                   As a final step, the accumulators are converted to 8.24 format.
+                   The return results <code>realResult</code> and <code>imagResult</code> are in 8.24 format.
  */
 
 void arm_cmplx_dot_prod_q15(
-  q15_t * pSrcA,
-  q15_t * pSrcB,
-  uint32_t numSamples,
-  q31_t * realResult,
-  q31_t * imagResult)
+  const q15_t * pSrcA,
+  const q15_t * pSrcB,
+        uint32_t numSamples,
+        q31_t * realResult,
+        q31_t * imagResult)
 {
-  q63_t real_sum = 0, imag_sum = 0;              /* Temporary result storage */
-  q15_t a0,b0,c0,d0;
-
-#if defined (ARM_MATH_DSP)
+        uint32_t blkCnt;                               /* Loop counter */
+        q63_t real_sum = 0, imag_sum = 0;              /* Temporary result variables */
+        q15_t a0,b0,c0,d0;
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
+#if defined (ARM_MATH_LOOPUNROLL)
 
-
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += (q31_t)a0 * c0;
-      imag_sum += (q31_t)a0 * d0;
-      real_sum -= (q31_t)b0 * d0;
-      imag_sum += (q31_t)b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += (q31_t)a0 * c0;
-      imag_sum += (q31_t)a0 * d0;
-      real_sum -= (q31_t)b0 * d0;
-      imag_sum += (q31_t)b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += (q31_t)a0 * c0;
-      imag_sum += (q31_t)a0 * d0;
-      real_sum -= (q31_t)b0 * d0;
-      imag_sum += (q31_t)b0 * c0;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += (q31_t)a0 * c0;
-      imag_sum += (q31_t)a0 * d0;
-      real_sum -= (q31_t)b0 * d0;
-      imag_sum += (q31_t)b0 * c0;
-
-      /* Decrement the loop counter */
-      blkCnt--;
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += (q31_t)a0 * c0;
+    imag_sum += (q31_t)a0 * d0;
+    real_sum -= (q31_t)b0 * d0;
+    imag_sum += (q31_t)b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += (q31_t)a0 * c0;
+    imag_sum += (q31_t)a0 * d0;
+    real_sum -= (q31_t)b0 * d0;
+    imag_sum += (q31_t)b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += (q31_t)a0 * c0;
+    imag_sum += (q31_t)a0 * d0;
+    real_sum -= (q31_t)b0 * d0;
+    imag_sum += (q31_t)b0 * c0;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += (q31_t)a0 * c0;
+    imag_sum += (q31_t)a0 * d0;
+    real_sum -= (q31_t)b0 * d0;
+    imag_sum += (q31_t)b0 * c0;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
-  while (blkCnt > 0U)
-  {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += (q31_t)a0 * c0;
-      imag_sum += (q31_t)a0 * d0;
-      real_sum -= (q31_t)b0 * d0;
-      imag_sum += (q31_t)b0 * c0;
-
-      /* Decrement the loop counter */
-      blkCnt--;
-  }
-
 #else
 
-  /* Run the below code for Cortex-M0 */
-
-  while (numSamples > 0U)
-  {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += a0 * c0;
-      imag_sum += a0 * d0;
-      real_sum -= b0 * d0;
-      imag_sum += b0 * c0;
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-      /* Decrement the loop counter */
-      numSamples--;
+  while (blkCnt > 0U)
+  {
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += (q31_t)a0 * c0;
+    imag_sum += (q31_t)a0 * d0;
+    real_sum -= (q31_t)b0 * d0;
+    imag_sum += (q31_t)b0 * c0;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
-  /* Store the real and imaginary results in 8.24 format  */
+  /* Store real and imaginary result in 8.24 format  */
   /* Convert real data in 34.30 to 8.24 by 6 right shifts */
   *realResult = (q31_t) (real_sum >> 6);
   /* Convert imaginary data in 34.30 to 8.24 by 6 right shifts */
@@ -173,5 +150,5 @@ void arm_cmplx_dot_prod_q15(
 }
 
 /**
- * @} end of cmplx_dot_prod group
+  @} end of cmplx_dot_prod group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q31.c
index 6eb5b6e..d715d98 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_dot_prod_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_dot_prod_q31.c
  * Description:  Q31 complex dot product
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,147 +29,125 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_dot_prod
- * @{
+  @addtogroup cmplx_dot_prod
+  @{
  */
 
 /**
- * @brief  Q31 complex dot product
- * @param  *pSrcA points to the first input vector
- * @param  *pSrcB points to the second input vector
- * @param  numSamples number of complex samples in each vector
- * @param  *realResult real part of the result returned here
- * @param  *imagResult imaginary part of the result returned here
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function is implemented using an internal 64-bit accumulator.
- * The intermediate 1.31 by 1.31 multiplications are performed with 64-bit precision and then shifted to 16.48 format.
- * The internal real and imaginary accumulators are in 16.48 format and provide 15 guard bits.
- * Additions are nonsaturating and no overflow will occur as long as <code>numSamples</code> is less than 32768.
- * The return results <code>realResult</code> and <code>imagResult</code> are in 16.48 format.
- * Input down scaling is not required.
+  @brief         Q31 complex dot product.
+  @param[in]     pSrcA       points to the first input vector
+  @param[in]     pSrcB       points to the second input vector
+  @param[in]     numSamples  number of samples in each vector
+  @param[out]    realResult  real part of the result returned here
+  @param[out]    imagResult  imaginary part of the result returned here
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function is implemented using an internal 64-bit accumulator.
+                   The intermediate 1.31 by 1.31 multiplications are performed with 64-bit precision and then shifted to 16.48 format.
+                   The internal real and imaginary accumulators are in 16.48 format and provide 15 guard bits.
+                   Additions are nonsaturating and no overflow will occur as long as <code>numSamples</code> is less than 32768.
+                   The return results <code>realResult</code> and <code>imagResult</code> are in 16.48 format.
+                   Input down scaling is not required.
  */
 
 void arm_cmplx_dot_prod_q31(
-  q31_t * pSrcA,
-  q31_t * pSrcB,
-  uint32_t numSamples,
-  q63_t * realResult,
-  q63_t * imagResult)
+  const q31_t * pSrcA,
+  const q31_t * pSrcB,
+        uint32_t numSamples,
+        q63_t * realResult,
+        q63_t * imagResult)
 {
-  q63_t real_sum = 0, imag_sum = 0;              /* Temporary result storage */
-  q31_t a0,b0,c0,d0;
-
-#if defined (ARM_MATH_DSP)
-
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
+        uint32_t blkCnt;                               /* Loop counter */
+        q63_t real_sum = 0, imag_sum = 0;              /* Temporary result variables */
+        q31_t a0,b0,c0,d0;
 
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      /* Decrement the loop counter */
-      blkCnt--;
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += ((q63_t)a0 * c0) >> 14;
+    imag_sum += ((q63_t)a0 * d0) >> 14;
+    real_sum -= ((q63_t)b0 * d0) >> 14;
+    imag_sum += ((q63_t)b0 * c0) >> 14;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += ((q63_t)a0 * c0) >> 14;
+    imag_sum += ((q63_t)a0 * d0) >> 14;
+    real_sum -= ((q63_t)b0 * d0) >> 14;
+    imag_sum += ((q63_t)b0 * c0) >> 14;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += ((q63_t)a0 * c0) >> 14;
+    imag_sum += ((q63_t)a0 * d0) >> 14;
+    real_sum -= ((q63_t)b0 * d0) >> 14;
+    imag_sum += ((q63_t)b0 * c0) >> 14;
+
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += ((q63_t)a0 * c0) >> 14;
+    imag_sum += ((q63_t)a0 * d0) >> 14;
+    real_sum -= ((q63_t)b0 * d0) >> 14;
+    imag_sum += ((q63_t)b0 * c0) >> 14;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-  /* If the numSamples  is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
-  while (blkCnt > 0U)
-  {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      /* Decrement the loop counter */
-      blkCnt--;
-  }
-
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
+
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-  while (numSamples > 0U)
+  while (blkCnt > 0U)
   {
-      a0 = *pSrcA++;
-      b0 = *pSrcA++;
-      c0 = *pSrcB++;
-      d0 = *pSrcB++;
-
-      real_sum += ((q63_t)a0 * c0) >> 14;
-      imag_sum += ((q63_t)a0 * d0) >> 14;
-      real_sum -= ((q63_t)b0 * d0) >> 14;
-      imag_sum += ((q63_t)b0 * c0) >> 14;
-
-      /* Decrement the loop counter */
-      numSamples--;
+    a0 = *pSrcA++;
+    b0 = *pSrcA++;
+    c0 = *pSrcB++;
+    d0 = *pSrcB++;
+
+    real_sum += ((q63_t)a0 * c0) >> 14;
+    imag_sum += ((q63_t)a0 * d0) >> 14;
+    real_sum -= ((q63_t)b0 * d0) >> 14;
+    imag_sum += ((q63_t)b0 * c0) >> 14;
+
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
-  /* Store the real and imaginary results in 16.48 format  */
+  /* Store real and imaginary result in 16.48 format  */
   *realResult = real_sum;
   *imagResult = imag_sum;
 }
 
 /**
- * @} end of cmplx_dot_prod group
+  @} end of cmplx_dot_prod group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_f32.c
index 95aaf1e..84812dc 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_f32.c
  * Description:  Floating-point complex magnitude
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,125 +29,160 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup cmplx_mag Complex Magnitude
- *
- * Computes the magnitude of the elements of a complex data vector.
- *
- * The <code>pSrc</code> points to the source data and
- * <code>pDst</code> points to the where the result should be written.
- * <code>numSamples</code> specifies the number of complex samples
- * in the input array and the data is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * The input array has a total of <code>2*numSamples</code> values;
- * the output array has a total of <code>numSamples</code> values.
- * The underlying algorithm is used:
- *
- * <pre>
- * for(n=0; n<numSamples; n++) {
- *     pDst[n] = sqrt(pSrc[(2*n)+0]^2 + pSrc[(2*n)+1]^2);
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup cmplx_mag Complex Magnitude
+
+  Computes the magnitude of the elements of a complex data vector.
+
+  The <code>pSrc</code> points to the source data and
+  <code>pDst</code> points to the where the result should be written.
+  <code>numSamples</code> specifies the number of complex samples
+  in the input array and the data is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  The input array has a total of <code>2*numSamples</code> values;
+  the output array has a total of <code>numSamples</code> values.
+
+  The underlying algorithm is used:
+
+  <pre>
+  for (n = 0; n < numSamples; n++) {
+      pDst[n] = sqrt(pSrc[(2*n)+0]^2 + pSrc[(2*n)+1]^2);
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup cmplx_mag
- * @{
+  @addtogroup cmplx_mag
+  @{
  */
+
 /**
- * @brief Floating-point complex magnitude.
- * @param[in]       *pSrc points to complex input buffer
- * @param[out]      *pDst points to real output buffer
- * @param[in]       numSamples number of complex samples in the input vector
- * @return none.
- *
+  @brief         Floating-point complex magnitude.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
  */
 
-
 void arm_cmplx_mag_f32(
-  float32_t * pSrc,
-  float32_t * pDst,
-  uint32_t numSamples)
+  const float32_t * pSrc,
+        float32_t * pDst,
+        uint32_t numSamples)
 {
-  float32_t realIn, imagIn;                      /* Temporary variables to hold input values */
+  uint32_t blkCnt;                               /* loop counter */
+  float32_t real, imag;                      /* Temporary variables to hold input values */
 
-#if defined (ARM_MATH_DSP)
+#if defined(ARM_MATH_NEON)
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
+  float32x4x2_t vecA;
+  float32x4_t vRealA;
+  float32x4_t vImagA;
+  float32x4_t vMagSqA;
 
-  /*loop Unrolling */
-  blkCnt = numSamples >> 2U;
+  float32x4x2_t vecB;
+  float32x4_t vRealB;
+  float32x4_t vImagB;
+  float32x4_t vMagSqB;
+
+  /* Loop unrolling: Compute 8 outputs at a time */
+  blkCnt = numSamples >> 3;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
+    /* out = sqrt((real * real) + (imag * imag)) */
 
-    /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    /* store the result in the destination buffer. */
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
+    vecA = vld2q_f32(pSrc);
+    pSrc += 8;
+
+    vecB = vld2q_f32(pSrc);
+    pSrc += 8;
 
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
+    vRealA = vmulq_f32(vecA.val[0], vecA.val[0]);
+    vImagA = vmulq_f32(vecA.val[1], vecA.val[1]);
+    vMagSqA = vaddq_f32(vRealA, vImagA);
 
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
+    vRealB = vmulq_f32(vecB.val[0], vecB.val[0]);
+    vImagB = vmulq_f32(vecB.val[1], vecB.val[1]);
+    vMagSqB = vaddq_f32(vRealB, vImagB);
 
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
+    /* Store the result in the destination buffer. */
+    vst1q_f32(pDst, __arm_vec_sqrt_f32_neon(vMagSqA));
+    pDst += 4;
 
+    vst1q_f32(pDst, __arm_vec_sqrt_f32_neon(vMagSqB));
+    pDst += 4;
 
     /* Decrement the loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x4U;
+  blkCnt = numSamples & 7;
+
+#else
+
+#if defined (ARM_MATH_LOOPUNROLL)
+
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
   while (blkCnt > 0U)
   {
     /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    /* store the result in the destination buffer. */
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
 
-    /* Decrement the loop counter */
+    real = *pSrc++;
+    imag = *pSrc++;
+
+    /* store result in destination buffer. */
+    arm_sqrt_f32((real * real) + (imag * imag), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    arm_sqrt_f32((real * real) + (imag * imag), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    arm_sqrt_f32((real * real) + (imag * imag), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    arm_sqrt_f32((real * real) + (imag * imag), pDst++);
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
+
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined(ARM_MATH_NEON) */
 
-  while (numSamples > 0U)
+  while (blkCnt > 0U)
   {
-    /* out = sqrt((real * real) + (imag * imag)) */
-    realIn = *pSrc++;
-    imagIn = *pSrc++;
-    /* store the result in the destination buffer. */
-    arm_sqrt_f32((realIn * realIn) + (imagIn * imagIn), pDst++);
+    /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
 
-    /* Decrement the loop counter */
-    numSamples--;
-  }
+    real = *pSrc++;
+    imag = *pSrc++;
+
+    /* store result in destination buffer. */
+    arm_sqrt_f32((real * real) + (imag * imag), pDst++);
 
-#endif /* #if defined (ARM_MATH_DSP) */
+    /* Decrement loop counter */
+    blkCnt--;
+  }
 
 }
 
 /**
- * @} end of cmplx_mag group
+  @} end of cmplx_mag group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q15.c
index 03d9b2a..a493274 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_q15.c
  * Description:  Q15 complex magnitude
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,113 +29,134 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_mag
- * @{
+  @addtogroup cmplx_mag
+  @{
  */
 
-
 /**
- * @brief  Q15 complex magnitude
- * @param  *pSrc points to the complex input vector
- * @param  *pDst points to the real output vector
- * @param  numSamples number of complex samples in the input vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.15 by 1.15 multiplications and finally output is converted into 2.14 format.
+  @brief         Q15 complex magnitude.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.15 by 1.15 multiplications and finally output is converted into 2.14 format.
  */
 
 void arm_cmplx_mag_q15(
-  q15_t * pSrc,
-  q15_t * pDst,
-  uint32_t numSamples)
+  const q15_t * pSrc,
+        q15_t * pDst,
+        uint32_t numSamples)
 {
-  q31_t acc0, acc1;                              /* Accumulators */
+        uint32_t blkCnt;                               /* Loop counter */
 
 #if defined (ARM_MATH_DSP)
+        q31_t in;
+        q31_t acc0;                                    /* Accumulators */
+#else
+       q15_t real, imag;                              /* Temporary input variables */
+       q31_t acc0, acc1;                              /* Accumulators */
+#endif
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
-  q31_t in1, in2, in3, in4;
-  q31_t acc2, acc3;
-
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-
     /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
-    in1 = *__SIMD32(pSrc)++;
-    in2 = *__SIMD32(pSrc)++;
-    in3 = *__SIMD32(pSrc)++;
-    in4 = *__SIMD32(pSrc)++;
-
-    acc0 = __SMUAD(in1, in1);
-    acc1 = __SMUAD(in2, in2);
-    acc2 = __SMUAD(in3, in3);
-    acc3 = __SMUAD(in4, in4);
-
-    /* store the result in 2.14 format in the destination buffer. */
-    arm_sqrt_q15((q15_t) ((acc0) >> 17), pDst++);
-    arm_sqrt_q15((q15_t) ((acc1) >> 17), pDst++);
-    arm_sqrt_q15((q15_t) ((acc2) >> 17), pDst++);
-    arm_sqrt_q15((q15_t) ((acc3) >> 17), pDst++);
-
-    /* Decrement the loop counter */
-    blkCnt--;
-  }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x4U;
+#if defined (ARM_MATH_DSP)
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    /* store result in 2.14 format in destination buffer. */
+    arm_sqrt_q15((q15_t) (acc0 >> 17), pDst++);
 
-  while (blkCnt > 0U)
-  {
-    /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
-    in1 = *__SIMD32(pSrc)++;
-    acc0 = __SMUAD(in1, in1);
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    arm_sqrt_q15((q15_t) (acc0 >> 17), pDst++);
+
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    arm_sqrt_q15((q15_t) (acc0 >> 17), pDst++);
 
-    /* store the result in 2.14 format in the destination buffer. */
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
     arm_sqrt_q15((q15_t) (acc0 >> 17), pDst++);
+#else
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+
+    /* store result in 2.14 format in destination buffer. */
+    arm_sqrt_q15((q15_t) (((q63_t) acc0 + acc1) >> 17), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    arm_sqrt_q15((q15_t) (((q63_t) acc0 + acc1) >> 17), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    arm_sqrt_q15((q15_t) (((q63_t) acc0 + acc1) >> 17), pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    arm_sqrt_q15((q15_t) (((q63_t) acc0 + acc1) >> 17), pDst++);
+#endif /* #if defined (ARM_MATH_DSP) */
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  /* Run the below code for Cortex-M0 */
-  q15_t real, imag;                              /* Temporary variables to hold input values */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  while (numSamples > 0U)
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+
+  while (blkCnt > 0U)
   {
-    /* out = sqrt(real * real + imag * imag) */
+    /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
+
+#if defined (ARM_MATH_DSP)
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+
+    /* store result in 2.14 format in destination buffer. */
+    arm_sqrt_q15((q15_t) (acc0 >> 17), pDst++);
+#else
     real = *pSrc++;
     imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
 
-    acc0 = (real * real);
-    acc1 = (imag * imag);
-
-    /* store the result in 2.14 format in the destination buffer. */
+    /* store result in 2.14 format in destination buffer. */
     arm_sqrt_q15((q15_t) (((q63_t) acc0 + acc1) >> 17), pDst++);
+#endif
 
-    /* Decrement the loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of cmplx_mag group
+  @} end of cmplx_mag group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q31.c
index 830ecb9..873e566 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_q31.c
  * Description:  Q31 complex magnitude
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,145 +29,102 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_mag
- * @{
+  @addtogroup cmplx_mag
+  @{
  */
 
 /**
- * @brief  Q31 complex magnitude
- * @param  *pSrc points to the complex input vector
- * @param  *pDst points to the real output vector
- * @param  numSamples number of complex samples in the input vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.31 by 1.31 multiplications and finally output is converted into 2.30 format.
- * Input down scaling is not required.
+  @brief         Q31 complex magnitude.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.31 by 1.31 multiplications and finally output is converted into 2.30 format.
+                   Input down scaling is not required.
  */
 
 void arm_cmplx_mag_q31(
-  q31_t * pSrc,
-  q31_t * pDst,
-  uint32_t numSamples)
+  const q31_t * pSrc,
+        q31_t * pDst,
+        uint32_t numSamples)
 {
-  q31_t real, imag;                              /* Temporary variables to hold input values */
-  q31_t acc0, acc1;                              /* Accumulators */
-  uint32_t blkCnt;                               /* loop counter */
-
-#if defined (ARM_MATH_DSP)
-
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  q31_t real1, real2, imag1, imag2;              /* Temporary variables to hold input values */
-  q31_t out1, out2, out3, out4;                  /* Accumulators */
-  q63_t mul1, mul2, mul3, mul4;                  /* Temporary variables */
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t real, imag;                              /* Temporary input variables */
+        q31_t acc0, acc1;                              /* Accumulators */
 
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* read complex input from source buffer */
-    real1 = pSrc[0];
-    imag1 = pSrc[1];
-    real2 = pSrc[2];
-    imag2 = pSrc[3];
-
-    /* calculate power of input values */
-    mul1 = (q63_t) real1 *real1;
-    mul2 = (q63_t) imag1 *imag1;
-    mul3 = (q63_t) real2 *real2;
-    mul4 = (q63_t) imag2 *imag2;
-
-    /* get the result to 3.29 format */
-    out1 = (q31_t) (mul1 >> 33);
-    out2 = (q31_t) (mul2 >> 33);
-    out3 = (q31_t) (mul3 >> 33);
-    out4 = (q31_t) (mul4 >> 33);
-
-    /* add real and imaginary accumulators */
-    out1 = out1 + out2;
-    out3 = out3 + out4;
-
-    /* read complex input from source buffer */
-    real1 = pSrc[4];
-    imag1 = pSrc[5];
-    real2 = pSrc[6];
-    imag2 = pSrc[7];
-
-    /* calculate square root */
-    arm_sqrt_q31(out1, &pDst[0]);
-
-    /* calculate power of input values */
-    mul1 = (q63_t) real1 *real1;
-
-    /* calculate square root */
-    arm_sqrt_q31(out3, &pDst[1]);
-
-    /* calculate power of input values */
-    mul2 = (q63_t) imag1 *imag1;
-    mul3 = (q63_t) real2 *real2;
-    mul4 = (q63_t) imag2 *imag2;
-
-    /* get the result to 3.29 format */
-    out1 = (q31_t) (mul1 >> 33);
-    out2 = (q31_t) (mul2 >> 33);
-    out3 = (q31_t) (mul3 >> 33);
-    out4 = (q31_t) (mul4 >> 33);
-
-    /* add real and imaginary accumulators */
-    out1 = out1 + out2;
-    out3 = out3 + out4;
-
-    /* calculate square root */
-    arm_sqrt_q31(out1, &pDst[2]);
-
-    /* increment destination by 8 to process next samples */
-    pSrc += 8U;
-
-    /* calculate square root */
-    arm_sqrt_q31(out3, &pDst[3]);
-
-    /* increment destination by 4 to process next samples */
-    pDst += 4U;
-
-    /* Decrement the loop counter */
+    /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = (q31_t) (((q63_t) real * real) >> 33);
+    acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
+
+    /* store result in 2.30 format in destination buffer. */
+    arm_sqrt_q31(acc0 + acc1, pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = (q31_t) (((q63_t) real * real) >> 33);
+    acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
+    arm_sqrt_q31(acc0 + acc1, pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = (q31_t) (((q63_t) real * real) >> 33);
+    acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
+    arm_sqrt_q31(acc0 + acc1, pDst++);
+
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = (q31_t) (((q63_t) real * real) >> 33);
+    acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
+    arm_sqrt_q31(acc0 + acc1, pDst++);
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
   while (blkCnt > 0U)
   {
     /* C[0] = sqrt(A[0] * A[0] + A[1] * A[1]) */
+
     real = *pSrc++;
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
     acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 2.30 format in the destination buffer. */
+
+    /* store result in 2.30 format in destination buffer. */
     arm_sqrt_q31(acc0 + acc1, pDst++);
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of cmplx_mag group
+  @} end of cmplx_mag group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_f32.c
index 59127a2..99f051c 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_squared_f32.c
  * Description:  Floating-point complex magnitude squared
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,176 +29,156 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup cmplx_mag_squared Complex Magnitude Squared
- *
- * Computes the magnitude squared of the elements of a complex data vector.
- *
- * The <code>pSrc</code> points to the source data and
- * <code>pDst</code> points to the where the result should be written.
- * <code>numSamples</code> specifies the number of complex samples
- * in the input array and the data is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * The input array has a total of <code>2*numSamples</code> values;
- * the output array has a total of <code>numSamples</code> values.
- *
- * The underlying algorithm is used:
- *
- * <pre>
- * for(n=0; n<numSamples; n++) {
- *     pDst[n] = pSrc[(2*n)+0]^2 + pSrc[(2*n)+1]^2;
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup cmplx_mag_squared Complex Magnitude Squared
+
+  Computes the magnitude squared of the elements of a complex data vector.
+
+  The <code>pSrc</code> points to the source data and
+  <code>pDst</code> points to the where the result should be written.
+  <code>numSamples</code> specifies the number of complex samples
+  in the input array and the data is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  The input array has a total of <code>2*numSamples</code> values;
+  the output array has a total of <code>numSamples</code> values.
+
+  The underlying algorithm is used:
+
+  <pre>
+  for (n = 0; n < numSamples; n++) {
+      pDst[n] = pSrc[(2*n)+0]^2 + pSrc[(2*n)+1]^2;
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup cmplx_mag_squared
- * @{
+  @addtogroup cmplx_mag_squared
+  @{
  */
 
-
 /**
- * @brief  Floating-point complex magnitude squared
- * @param[in]  *pSrc points to the complex input vector
- * @param[out]  *pDst points to the real output vector
- * @param[in]  numSamples number of complex samples in the input vector
- * @return none.
+  @brief         Floating-point complex magnitude squared.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
  */
 
 void arm_cmplx_mag_squared_f32(
-  float32_t * pSrc,
-  float32_t * pDst,
-  uint32_t numSamples)
+  const float32_t * pSrc,
+        float32_t * pDst,
+        uint32_t numSamples)
 {
-  float32_t real, imag;                          /* Temporary variables to store real and imaginary values */
-  uint32_t blkCnt;                               /* loop counter */
+        uint32_t blkCnt;                               /* Loop counter */
+        float32_t real, imag;                          /* Temporary input variables */
 
-#if defined (ARM_MATH_DSP)
-  float32_t real1, real2, real3, real4;          /* Temporary variables to hold real values */
-  float32_t imag1, imag2, imag3, imag4;          /* Temporary variables to hold imaginary values */
-  float32_t mul1, mul2, mul3, mul4;              /* Temporary variables */
-  float32_t mul5, mul6, mul7, mul8;              /* Temporary variables */
-  float32_t out1, out2, out3, out4;              /* Temporary variables to hold output values */
+#if defined(ARM_MATH_NEON)
+  float32x4x2_t vecA;
+  float32x4_t vRealA;
+  float32x4_t vImagA;
+  float32x4_t vMagSqA;
 
-  /*loop Unrolling */
-  blkCnt = numSamples >> 2U;
+  float32x4x2_t vecB;
+  float32x4_t vRealB;
+  float32x4_t vImagB;
+  float32x4_t vMagSqB;
+
+  /* Loop unrolling: Compute 8 outputs at a time */
+  blkCnt = numSamples >> 3;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
-    /* read real input sample from source buffer */
-    real1 = pSrc[0];
-    /* read imaginary input sample from source buffer */
-    imag1 = pSrc[1];
-
-    /* calculate power of real value */
-    mul1 = real1 * real1;
-
-    /* read real input sample from source buffer */
-    real2 = pSrc[2];
-
-    /* calculate power of imaginary value */
-    mul2 = imag1 * imag1;
+    /* out = sqrt((real * real) + (imag * imag)) */
 
-    /* read imaginary input sample from source buffer */
-    imag2 = pSrc[3];
+    vecA = vld2q_f32(pSrc);
+    pSrc += 8;
 
-    /* calculate power of real value */
-    mul3 = real2 * real2;
+    vRealA = vmulq_f32(vecA.val[0], vecA.val[0]);
+    vImagA = vmulq_f32(vecA.val[1], vecA.val[1]);
+    vMagSqA = vaddq_f32(vRealA, vImagA);
 
-    /* read real input sample from source buffer */
-    real3 = pSrc[4];
+    vecB = vld2q_f32(pSrc);
+    pSrc += 8;
 
-    /* calculate power of imaginary value */
-    mul4 = imag2 * imag2;
+    vRealB = vmulq_f32(vecB.val[0], vecB.val[0]);
+    vImagB = vmulq_f32(vecB.val[1], vecB.val[1]);
+    vMagSqB = vaddq_f32(vRealB, vImagB);
 
-    /* read imaginary input sample from source buffer */
-    imag3 = pSrc[5];
+    /* Store the result in the destination buffer. */
+    vst1q_f32(pDst, vMagSqA);
+    pDst += 4;
 
-    /* calculate power of real value */
-    mul5 = real3 * real3;
-    /* calculate power of imaginary value */
-    mul6 = imag3 * imag3;
+    vst1q_f32(pDst, vMagSqB);
+    pDst += 4;
 
-    /* read real input sample from source buffer */
-    real4 = pSrc[6];
-
-    /* accumulate real and imaginary powers */
-    out1 = mul1 + mul2;
-
-    /* read imaginary input sample from source buffer */
-    imag4 = pSrc[7];
-
-    /* accumulate real and imaginary powers */
-    out2 = mul3 + mul4;
-
-    /* calculate power of real value */
-    mul7 = real4 * real4;
-    /* calculate power of imaginary value */
-    mul8 = imag4 * imag4;
+    /* Decrement the loop counter */
+    blkCnt--;
+  }
 
-    /* store output to destination */
-    pDst[0] = out1;
+  blkCnt = numSamples & 7;
 
-    /* accumulate real and imaginary powers */
-    out3 = mul5 + mul6;
+#else
+#if defined (ARM_MATH_LOOPUNROLL)
 
-    /* store output to destination */
-    pDst[1] = out2;
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
-    /* accumulate real and imaginary powers */
-    out4 = mul7 + mul8;
+  while (blkCnt > 0U)
+  {
+    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
 
-    /* store output to destination */
-    pDst[2] = out3;
+    real = *pSrc++;
+    imag = *pSrc++;
+    *pDst++ = (real * real) + (imag * imag);
 
-    /* increment destination pointer by 8 to process next samples */
-    pSrc += 8U;
+    real = *pSrc++;
+    imag = *pSrc++;
+    *pDst++ = (real * real) + (imag * imag);
 
-    /* store output to destination */
-    pDst[3] = out4;
+    real = *pSrc++;
+    imag = *pSrc++;
+    *pDst++ = (real * real) + (imag * imag);
 
-    /* increment destination pointer by 4 to process next samples */
-    pDst += 4U;
+    real = *pSrc++;
+    imag = *pSrc++;
+    *pDst++ = (real * real) + (imag * imag);
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
-
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined(ARM_MATH_NEON) */
 
   while (blkCnt > 0U)
   {
     /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
+
     real = *pSrc++;
     imag = *pSrc++;
 
-    /* out = (real * real) + (imag * imag) */
-    /* store the result in the destination buffer. */
+    /* store result in destination buffer. */
     *pDst++ = (real * real) + (imag * imag);
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of cmplx_mag_squared group
+  @} end of cmplx_mag_squared group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q15.c
index 3f740c3..fa5f4e6 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_squared_q15.c
  * Description:  Q15 complex magnitude squared
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,108 +29,133 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_mag_squared
- * @{
+  @addtogroup cmplx_mag_squared
+  @{
  */
 
 /**
- * @brief  Q15 complex magnitude squared
- * @param  *pSrc points to the complex input vector
- * @param  *pDst points to the real output vector
- * @param  numSamples number of complex samples in the input vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.15 by 1.15 multiplications and finally output is converted into 3.13 format.
+  @brief         Q15 complex magnitude squared.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.15 by 1.15 multiplications and finally output is converted into 3.13 format.
  */
 
 void arm_cmplx_mag_squared_q15(
-  q15_t * pSrc,
-  q15_t * pDst,
-  uint32_t numSamples)
+  const q15_t * pSrc,
+        q15_t * pDst,
+        uint32_t numSamples)
 {
-  q31_t acc0, acc1;                              /* Accumulators */
+        uint32_t blkCnt;                               /* Loop counter */
 
 #if defined (ARM_MATH_DSP)
+        q31_t in;
+        q31_t acc0;                                    /* Accumulators */
+#else
+        q15_t real, imag;                              /* Temporary input variables */
+        q31_t acc0, acc1;                              /* Accumulators */
+#endif
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
-  q31_t in1, in2, in3, in4;
-  q31_t acc2, acc3;
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /*loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
     /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
-    in1 = *__SIMD32(pSrc)++;
-    in2 = *__SIMD32(pSrc)++;
-    in3 = *__SIMD32(pSrc)++;
-    in4 = *__SIMD32(pSrc)++;
 
-    acc0 = __SMUAD(in1, in1);
-    acc1 = __SMUAD(in2, in2);
-    acc2 = __SMUAD(in3, in3);
-    acc3 = __SMUAD(in4, in4);
+#if defined (ARM_MATH_DSP)
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    /* store result in 3.13 format in destination buffer. */
+    *pDst++ = (q15_t) (acc0 >> 17);
 
-    /* store the result in 3.13 format in the destination buffer. */
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
     *pDst++ = (q15_t) (acc0 >> 17);
-    *pDst++ = (q15_t) (acc1 >> 17);
-    *pDst++ = (q15_t) (acc2 >> 17);
-    *pDst++ = (q15_t) (acc3 >> 17);
 
-    /* Decrement the loop counter */
-    blkCnt--;
-  }
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    *pDst++ = (q15_t) (acc0 >> 17);
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x4U;
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+    *pDst++ = (q15_t) (acc0 >> 17);
+#else
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    /* store result in 3.13 format in destination buffer. */
+    *pDst++ = (q15_t) (((q63_t) acc0 + acc1) >> 17);
 
-  while (blkCnt > 0U)
-  {
-    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
-    in1 = *__SIMD32(pSrc)++;
-    acc0 = __SMUAD(in1, in1);
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    *pDst++ = (q15_t) (((q63_t) acc0 + acc1) >> 17);
 
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ = (q15_t) (acc0 >> 17);
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    *pDst++ = (q15_t) (((q63_t) acc0 + acc1) >> 17);
 
-    /* Decrement the loop counter */
+    real = *pSrc++;
+    imag = *pSrc++;
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+    *pDst++ = (q15_t) (((q63_t) acc0 + acc1) >> 17);
+#endif /* #if defined (ARM_MATH_DSP) */
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  /* Run the below code for Cortex-M0 */
-  q15_t real, imag;                              /* Temporary variables to store real and imaginary values */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  while (numSamples > 0U)
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+
+  while (blkCnt > 0U)
   {
-    /* out = ((real * real) + (imag * imag)) */
+    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
+
+#if defined (ARM_MATH_DSP)
+    in = read_q15x2_ia ((q15_t **) &pSrc);
+    acc0 = __SMUAD(in, in);
+
+    /* store result in 3.13 format in destination buffer. */
+    *pDst++ = (q15_t) (acc0 >> 17);
+#else
     real = *pSrc++;
     imag = *pSrc++;
-    acc0 = (real * real);
-    acc1 = (imag * imag);
-    /* store the result in 3.13 format in the destination buffer. */
+    acc0 = ((q31_t) real * real);
+    acc1 = ((q31_t) imag * imag);
+
+    /* store result in 3.13 format in destination buffer. */
     *pDst++ = (q15_t) (((q63_t) acc0 + acc1) >> 17);
+#endif
 
-    /* Decrement the loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of cmplx_mag_squared group
+  @} end of cmplx_mag_squared group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q31.c
index c2b2c50..54863ef 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mag_squared_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mag_squared_q31.c
  * Description:  Q31 complex magnitude squared
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,49 +29,44 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup cmplx_mag_squared
- * @{
+  @addtogroup cmplx_mag_squared
+  @{
  */
 
-
 /**
- * @brief  Q31 complex magnitude squared
- * @param  *pSrc points to the complex input vector
- * @param  *pDst points to the real output vector
- * @param  numSamples number of complex samples in the input vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.31 by 1.31 multiplications and finally output is converted into 3.29 format.
- * Input down scaling is not required.
+  @brief         Q31 complex magnitude squared.
+  @param[in]     pSrc        points to input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.31 by 1.31 multiplications and finally output is converted into 3.29 format.
+                   Input down scaling is not required.
  */
 
 void arm_cmplx_mag_squared_q31(
-  q31_t * pSrc,
-  q31_t * pDst,
-  uint32_t numSamples)
+  const q31_t * pSrc,
+        q31_t * pDst,
+        uint32_t numSamples)
 {
-  q31_t real, imag;                              /* Temporary variables to store real and imaginary values */
-  q31_t acc0, acc1;                              /* Accumulators */
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t real, imag;                              /* Temporary input variables */
+        q31_t acc0, acc1;                              /* Accumulators */
 
-#if defined (ARM_MATH_DSP)
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counter */
-
-  /* loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
     /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
+
     real = *pSrc++;
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
@@ -83,67 +78,52 @@ void arm_cmplx_mag_squared_q31(
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
     acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 3.29 format in the destination buffer. */
     *pDst++ = acc0 + acc1;
 
     real = *pSrc++;
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
     acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 3.29 format in the destination buffer. */
     *pDst++ = acc0 + acc1;
 
     real = *pSrc++;
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
     acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 3.29 format in the destination buffer. */
     *pDst++ = acc0 + acc1;
 
-    /* Decrement the loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
-  while (blkCnt > 0U)
-  {
-    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
-    real = *pSrc++;
-    imag = *pSrc++;
-    acc0 = (q31_t) (((q63_t) real * real) >> 33);
-    acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 3.29 format in the destination buffer. */
-    *pDst++ = acc0 + acc1;
-
-    /* Decrement the loop counter */
-    blkCnt--;
-  }
-
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
+
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-  while (numSamples > 0U)
+  while (blkCnt > 0U)
   {
-    /* out = ((real * real) + (imag * imag)) */
+    /* C[0] = (A[0] * A[0] + A[1] * A[1]) */
+
     real = *pSrc++;
     imag = *pSrc++;
     acc0 = (q31_t) (((q63_t) real * real) >> 33);
     acc1 = (q31_t) (((q63_t) imag * imag) >> 33);
-    /* store the result in 3.29 format in the destination buffer. */
+
+    /* store result in 3.29 format in destination buffer. */
     *pDst++ = acc0 + acc1;
 
-    /* Decrement the loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of cmplx_mag_squared group
+  @} end of cmplx_mag_squared group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_f32.c
index 3717591..8d14821 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_cmplx_f32.c
  * Description:  Floating-point complex-by-complex multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,168 +29,166 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup CmplxByCmplxMult Complex-by-Complex Multiplication
- *
- * Multiplies a complex vector by another complex vector and generates a complex result.
- * The data in the complex arrays is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * The parameter <code>numSamples</code> represents the number of complex
- * samples processed.  The complex arrays have a total of <code>2*numSamples</code>
- * real values.
- *
- * The underlying algorithm is used:
- *
- * <pre>
- * for(n=0; n<numSamples; n++) {
- *     pDst[(2*n)+0] = pSrcA[(2*n)+0] * pSrcB[(2*n)+0] - pSrcA[(2*n)+1] * pSrcB[(2*n)+1];
- *     pDst[(2*n)+1] = pSrcA[(2*n)+0] * pSrcB[(2*n)+1] + pSrcA[(2*n)+1] * pSrcB[(2*n)+0];
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup CmplxByCmplxMult Complex-by-Complex Multiplication
+
+  Multiplies a complex vector by another complex vector and generates a complex result.
+  The data in the complex arrays is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  The parameter <code>numSamples</code> represents the number of complex
+  samples processed.  The complex arrays have a total of <code>2*numSamples</code>
+  real values.
+
+  The underlying algorithm is used:
+
+  <pre>
+  for (n = 0; n < numSamples; n++) {
+      pDst[(2*n)+0] = pSrcA[(2*n)+0] * pSrcB[(2*n)+0] - pSrcA[(2*n)+1] * pSrcB[(2*n)+1];
+      pDst[(2*n)+1] = pSrcA[(2*n)+0] * pSrcB[(2*n)+1] + pSrcA[(2*n)+1] * pSrcB[(2*n)+0];
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup CmplxByCmplxMult
- * @{
+  @addtogroup CmplxByCmplxMult
+  @{
  */
 
-
 /**
- * @brief  Floating-point complex-by-complex multiplication
- * @param[in]  *pSrcA points to the first input vector
- * @param[in]  *pSrcB points to the second input vector
- * @param[out]  *pDst  points to the output vector
- * @param[in]  numSamples number of complex samples in each vector
- * @return none.
+  @brief         Floating-point complex-by-complex multiplication.
+  @param[in]     pSrcA       points to first input vector
+  @param[in]     pSrcB       points to second input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
  */
 
 void arm_cmplx_mult_cmplx_f32(
-  float32_t * pSrcA,
-  float32_t * pSrcB,
-  float32_t * pDst,
-  uint32_t numSamples)
+  const float32_t * pSrcA,
+  const float32_t * pSrcB,
+        float32_t * pDst,
+        uint32_t numSamples)
 {
-  float32_t a1, b1, c1, d1;                      /* Temporary variables to store real and imaginary values */
-  uint32_t blkCnt;                               /* loop counters */
-
-#if defined (ARM_MATH_DSP)
-
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  float32_t a2, b2, c2, d2;                      /* Temporary variables to store real and imaginary values */
-  float32_t acc1, acc2, acc3, acc4;
-
-
-  /* loop Unrolling */
-  blkCnt = numSamples >> 2U;
-
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
-    a1 = *pSrcA;                /* A[2 * i] */
-    c1 = *pSrcB;                /* B[2 * i] */
+    uint32_t blkCnt;                               /* Loop counter */
+    float32_t a, b, c, d;  /* Temporary variables to store real and imaginary values */
 
-    b1 = *(pSrcA + 1);          /* A[2 * i + 1] */
-    acc1 = a1 * c1;             /* acc1 = A[2 * i] * B[2 * i] */
+#if defined(ARM_MATH_NEON)
+    float32x4x2_t va, vb;
+    float32x4_t real, imag;
+    float32x4x2_t outCplx;
 
-    a2 = *(pSrcA + 2);          /* A[2 * i + 2] */
-    acc2 = (b1 * c1);           /* acc2 = A[2 * i + 1] * B[2 * i] */
+    /* Compute 4 outputs at a time */
+    blkCnt = numSamples >> 2U;
 
-    d1 = *(pSrcB + 1);          /* B[2 * i + 1] */
-    c2 = *(pSrcB + 2);          /* B[2 * i + 2] */
-    acc1 -= b1 * d1;            /* acc1 =      A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1] */
+    while (blkCnt > 0U)
+    {
+        va = vld2q_f32(pSrcA);  // load & separate real/imag pSrcA (de-interleave 2)
+        vb = vld2q_f32(pSrcB);  // load & separate real/imag pSrcB
 
-    d2 = *(pSrcB + 3);          /* B[2 * i + 3] */
-    acc3 = a2 * c2;             /* acc3 =       A[2 * i + 2] * B[2 * i + 2] */
+	/* Increment pointers */
+        pSrcA += 8;
+        pSrcB += 8;
+	
+	/* Re{C} = Re{A}*Re{B} - Im{A}*Im{B} */
+        outCplx.val[0] = vmulq_f32(va.val[0], vb.val[0]);
+        outCplx.val[0] = vmlsq_f32(outCplx.val[0], va.val[1], vb.val[1]);
 
-    b2 = *(pSrcA + 3);          /* A[2 * i + 3] */
-    acc2 += (a1 * d1);          /* acc2 =      A[2 * i + 1] * B[2 * i] + A[2 * i] * B[2 * i + 1] */
+	/* Im{C} = Re{A}*Im{B} + Im{A}*Re{B} */
+        outCplx.val[1] = vmulq_f32(va.val[0], vb.val[1]);
+        outCplx.val[1] = vmlaq_f32(outCplx.val[1], va.val[1], vb.val[0]);
 
-    a1 = *(pSrcA + 4);          /* A[2 * i + 4] */
-    acc4 = (a2 * d2);           /* acc4 =   A[2 * i + 2] * B[2 * i + 3] */
+        vst2q_f32(pDst, outCplx);
 
-    c1 = *(pSrcB + 4);          /* B[2 * i + 4] */
-    acc3 -= (b2 * d2);          /* acc3 =       A[2 * i + 2] * B[2 * i + 2] - A[2 * i + 3] * B[2 * i + 3] */
-    *pDst = acc1;               /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1] */
+	/* Increment pointer */
+        pDst += 8;
 
-    b1 = *(pSrcA + 5);          /* A[2 * i + 5] */
-    acc4 += b2 * c2;            /* acc4 =   A[2 * i + 2] * B[2 * i + 3] + A[2 * i + 3] * B[2 * i + 2] */
+	/* Decrement the loop counter */
+        blkCnt--;
+    }
 
-    *(pDst + 1) = acc2;         /* C[2 * i + 1] = A[2 * i + 1] * B[2 * i] + A[2 * i] * B[2 * i + 1]  */
-    acc1 = (a1 * c1);
+    /* Tail */
+    blkCnt = numSamples & 3;
 
-    d1 = *(pSrcB + 5);
-    acc2 = (b1 * c1);
-
-    *(pDst + 2) = acc3;
-    *(pDst + 3) = acc4;
-
-    a2 = *(pSrcA + 6);
-    acc1 -= (b1 * d1);
-
-    c2 = *(pSrcB + 6);
-    acc2 += (a1 * d1);
-
-    b2 = *(pSrcA + 7);
-    acc3 = (a2 * c2);
-
-    d2 = *(pSrcB + 7);
-    acc4 = (b2 * c2);
-
-    *(pDst + 4) = acc1;
-    pSrcA += 8U;
-
-    acc3 -= (b2 * d2);
-    acc4 += (a2 * d2);
-
-    *(pDst + 5) = acc2;
-    pSrcB += 8U;
-
-    *(pDst + 6) = acc3;
-    *(pDst + 7) = acc4;
+#else
+#if defined (ARM_MATH_LOOPUNROLL)
 
-    pDst += 8U;
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
-    /* Decrement the numSamples loop counter */
+  while (blkCnt > 0U)
+  {
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
+
+    a = *pSrcA++;
+    b = *pSrcA++;
+    c = *pSrcB++;
+    d = *pSrcB++;
+    /* store result in destination buffer. */
+    *pDst++ = (a * c) - (b * d);
+    *pDst++ = (a * d) + (b * c);
+
+    a = *pSrcA++;
+    b = *pSrcA++;
+    c = *pSrcB++;
+    d = *pSrcB++;
+    *pDst++ = (a * c) - (b * d);
+    *pDst++ = (a * d) + (b * c);
+
+    a = *pSrcA++;
+    b = *pSrcA++;
+    c = *pSrcB++;
+    d = *pSrcB++;
+    *pDst++ = (a * c) - (b * d);
+    *pDst++ = (a * d) + (b * c);
+
+    a = *pSrcA++;
+    b = *pSrcA++;
+    c = *pSrcB++;
+    d = *pSrcB++;
+    *pDst++ = (a * c) - (b * d);
+    *pDst++ = (a * d) + (b * c);
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined(ARM_MATH_NEON) */
 
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
-    a1 = *pSrcA++;
-    b1 = *pSrcA++;
-    c1 = *pSrcB++;
-    d1 = *pSrcB++;
-
-    /* store the result in the destination buffer. */
-    *pDst++ = (a1 * c1) - (b1 * d1);
-    *pDst++ = (a1 * d1) + (b1 * c1);
-
-    /* Decrement the numSamples loop counter */
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
+
+    a = *pSrcA++;
+    b = *pSrcA++;
+    c = *pSrcB++;
+    d = *pSrcB++;
+
+    /* store result in destination buffer. */
+    *pDst++ = (a * c) - (b * d);
+    *pDst++ = (a * d) + (b * c);
+
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of CmplxByCmplxMult group
+  @} end of CmplxByCmplxMult group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q15.c
index 2869837..6659427 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_cmplx_q15.c
  * Description:  Q15 complex-by-complex multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,153 +29,108 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup CmplxByCmplxMult
- * @{
+  @addtogroup CmplxByCmplxMult
+  @{
  */
 
 /**
- * @brief  Q15 complex-by-complex multiplication
- * @param[in]  *pSrcA points to the first input vector
- * @param[in]  *pSrcB points to the second input vector
- * @param[out]  *pDst  points to the output vector
- * @param[in]  numSamples number of complex samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.15 by 1.15 multiplications and finally output is converted into 3.13 format.
+  @brief         Q15 complex-by-complex multiplication.
+  @param[in]     pSrcA       points to first input vector
+  @param[in]     pSrcB       points to second input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.15 by 1.15 multiplications and finally output is converted into 3.13 format.
  */
 
 void arm_cmplx_mult_cmplx_q15(
-  q15_t * pSrcA,
-  q15_t * pSrcB,
-  q15_t * pDst,
-  uint32_t numSamples)
+  const q15_t * pSrcA,
+  const q15_t * pSrcB,
+        q15_t * pDst,
+        uint32_t numSamples)
 {
-  q15_t a, b, c, d;                              /* Temporary variables to store real and imaginary values */
-
-#if defined (ARM_MATH_DSP)
+        uint32_t blkCnt;                               /* Loop counter */
+        q15_t a, b, c, d;                              /* Temporary variables */
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counters */
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /* loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
+
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
+    /* store result in 3.13 format in destination buffer. */
+    *pDst++ = (q15_t) ( (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17) );
+    *pDst++ = (q15_t) ( (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
+    *pDst++ = (q15_t) ( (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17) );
+    *pDst++ = (q15_t) ( (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
+    *pDst++ = (q15_t) ( (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17) );
+    *pDst++ = (q15_t) ( (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
+    *pDst++ = (q15_t) ( (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17) );
+    *pDst++ = (q15_t) ( (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17) );
 
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
-
-    /* Decrement the blockSize loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
-    a = *pSrcA++;
-    b = *pSrcA++;
-    c = *pSrcB++;
-    d = *pSrcB++;
-
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
-
-    /* Decrement the blockSize loop counter */
-    blkCnt--;
-  }
-
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  while (numSamples > 0U)
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+
+  while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
+
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
 
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17);
-    /* store the result in 3.13 format in the destination buffer. */
-    *pDst++ =
-      (q15_t) (q31_t) (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17);
+    /* store result in 3.13 format in destination buffer. */
+    *pDst++ = (q15_t) ( (((q31_t) a * c) >> 17) - (((q31_t) b * d) >> 17) );
+    *pDst++ = (q15_t) ( (((q31_t) a * d) >> 17) + (((q31_t) b * c) >> 17) );
 
-    /* Decrement the blockSize loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of CmplxByCmplxMult group
+  @} end of CmplxByCmplxMult group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q31.c
index b01c4f6..f6d6dc6 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_cmplx_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_cmplx_q31.c
  * Description:  Q31 complex-by-complex multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,286 +29,109 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup CmplxByCmplxMult
- * @{
+  @addtogroup CmplxByCmplxMult
+  @{
  */
 
-
 /**
- * @brief  Q31 complex-by-complex multiplication
- * @param[in]  *pSrcA points to the first input vector
- * @param[in]  *pSrcB points to the second input vector
- * @param[out]  *pDst  points to the output vector
- * @param[in]  numSamples number of complex samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function implements 1.31 by 1.31 multiplications and finally output is converted into 3.29 format.
- * Input down scaling is not required.
+  @brief         Q31 complex-by-complex multiplication.
+  @param[in]     pSrcA       points to first input vector
+  @param[in]     pSrcB       points to second input vector
+  @param[out]    pDst        points to output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function implements 1.31 by 1.31 multiplications and finally output is converted into 3.29 format.
+                   Input down scaling is not required.
  */
 
 void arm_cmplx_mult_cmplx_q31(
-  q31_t * pSrcA,
-  q31_t * pSrcB,
-  q31_t * pDst,
-  uint32_t numSamples)
+  const q31_t * pSrcA,
+  const q31_t * pSrcB,
+        q31_t * pDst,
+        uint32_t numSamples)
 {
-  q31_t a, b, c, d;                              /* Temporary variables to store real and imaginary values */
-  uint32_t blkCnt;                               /* loop counters */
-  q31_t mul1, mul2, mul3, mul4;
-  q31_t out1, out2;
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t a, b, c, d;                              /* Temporary variables */
 
-#if defined (ARM_MATH_DSP)
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-
-  /* loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
+
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
+    /* store result in 3.29 format in destination buffer. */
+    *pDst++ = (q31_t) ( (((q63_t) a * c) >> 33) - (((q63_t) b * d) >> 33) );
+    *pDst++ = (q31_t) ( (((q63_t) a * d) >> 33) + (((q63_t) b * c) >> 33) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
+    *pDst++ = (q31_t) ( (((q63_t) a * c) >> 33) - (((q63_t) b * d) >> 33) );
+    *pDst++ = (q31_t) ( (((q63_t) a * d) >> 33) + (((q63_t) b * c) >> 33) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
+    *pDst++ = (q31_t) ( (((q63_t) a * c) >> 33) - (((q63_t) b * d) >> 33) );
+    *pDst++ = (q31_t) ( (((q63_t) a * d) >> 33) + (((q63_t) b * c) >> 33) );
 
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
+    *pDst++ = (q31_t) ( (((q63_t) a * c) >> 33) - (((q63_t) b * d) >> 33) );
+    *pDst++ = (q31_t) ( (((q63_t) a * d) >> 33) + (((q63_t) b * c) >> 33) );
 
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
-
-    /* Decrement the blockSize loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
-    a = *pSrcA++;
-    b = *pSrcA++;
-    c = *pSrcB++;
-    d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
-
-    /* Decrement the blockSize loop counter */
-    blkCnt--;
-  }
-
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
 
-  /* loop Unrolling */
-  blkCnt = numSamples >> 1U;
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-  /* First part of the processing with loop unrolling.  Compute 2 outputs at a time.
-   ** a second loop below computes the remaining 1 sample. */
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
-    a = *pSrcA++;
-    b = *pSrcA++;
-    c = *pSrcB++;
-    d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
-
-    a = *pSrcA++;
-    b = *pSrcA++;
-    c = *pSrcB++;
-    d = *pSrcB++;
-
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
-
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
-
-    /* Decrement the blockSize loop counter */
-    blkCnt--;
-  }
+    /* C[2 * i    ] = A[2 * i] * B[2 * i    ] - A[2 * i + 1] * B[2 * i + 1]. */
+    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i    ]. */
 
-  /* If the blockSize is not a multiple of 2, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x2U;
-
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[2 * i] - A[2 * i + 1] * B[2 * i + 1].  */
-    /* C[2 * i + 1] = A[2 * i] * B[2 * i + 1] + A[2 * i + 1] * B[2 * i].  */
     a = *pSrcA++;
     b = *pSrcA++;
     c = *pSrcB++;
     d = *pSrcB++;
 
-    mul1 = (q31_t) (((q63_t) a * c) >> 32);
-    mul2 = (q31_t) (((q63_t) b * d) >> 32);
-    mul3 = (q31_t) (((q63_t) a * d) >> 32);
-    mul4 = (q31_t) (((q63_t) b * c) >> 32);
-
-    mul1 = (mul1 >> 1);
-    mul2 = (mul2 >> 1);
-    mul3 = (mul3 >> 1);
-    mul4 = (mul4 >> 1);
+    /* store result in 3.29 format in destination buffer. */
+    *pDst++ = (q31_t) ( (((q63_t) a * c) >> 33) - (((q63_t) b * d) >> 33) );
+    *pDst++ = (q31_t) ( (((q63_t) a * d) >> 33) + (((q63_t) b * c) >> 33) );
 
-    out1 = mul1 - mul2;
-    out2 = mul3 + mul4;
-
-    /* store the real result in 3.29 format in the destination buffer. */
-    *pDst++ = out1;
-    /* store the imag result in 3.29 format in the destination buffer. */
-    *pDst++ = out2;
-
-    /* Decrement the blockSize loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of CmplxByCmplxMult group
+  @} end of CmplxByCmplxMult group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_f32.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_f32.c
index 8c7ca31..9651999 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_f32.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_f32.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_real_f32.c
  * Description:  Floating-point complex by real multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,185 +29,141 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @defgroup CmplxByRealMult Complex-by-Real Multiplication
- *
- * Multiplies a complex vector by a real vector and generates a complex result.
- * The data in the complex arrays is stored in an interleaved fashion
- * (real, imag, real, imag, ...).
- * The parameter <code>numSamples</code> represents the number of complex
- * samples processed.  The complex arrays have a total of <code>2*numSamples</code>
- * real values while the real array has a total of <code>numSamples</code>
- * real values.
- *
- * The underlying algorithm is used:
- *
- * <pre>
- * for(n=0; n<numSamples; n++) {
- *     pCmplxDst[(2*n)+0] = pSrcCmplx[(2*n)+0] * pSrcReal[n];
- *     pCmplxDst[(2*n)+1] = pSrcCmplx[(2*n)+1] * pSrcReal[n];
- * }
- * </pre>
- *
- * There are separate functions for floating-point, Q15, and Q31 data types.
+  @defgroup CmplxByRealMult Complex-by-Real Multiplication
+
+  Multiplies a complex vector by a real vector and generates a complex result.
+  The data in the complex arrays is stored in an interleaved fashion
+  (real, imag, real, imag, ...).
+  The parameter <code>numSamples</code> represents the number of complex
+  samples processed.  The complex arrays have a total of <code>2*numSamples</code>
+  real values while the real array has a total of <code>numSamples</code>
+  real values.
+
+  The underlying algorithm is used:
+
+  <pre>
+  for (n = 0; n < numSamples; n++) {
+      pCmplxDst[(2*n)+0] = pSrcCmplx[(2*n)+0] * pSrcReal[n];
+      pCmplxDst[(2*n)+1] = pSrcCmplx[(2*n)+1] * pSrcReal[n];
+  }
+  </pre>
+
+  There are separate functions for floating-point, Q15, and Q31 data types.
  */
 
 /**
- * @addtogroup CmplxByRealMult
- * @{
+  @addtogroup CmplxByRealMult
+  @{
  */
 
-
 /**
- * @brief  Floating-point complex-by-real multiplication
- * @param[in]  *pSrcCmplx points to the complex input vector
- * @param[in]  *pSrcReal points to the real input vector
- * @param[out]  *pCmplxDst points to the complex output vector
- * @param[in]  numSamples number of samples in each vector
- * @return none.
+  @brief         Floating-point complex-by-real multiplication.
+  @param[in]     pSrcCmplx   points to complex input vector
+  @param[in]     pSrcReal    points to real input vector
+  @param[out]    pCmplxDst   points to complex output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
  */
 
 void arm_cmplx_mult_real_f32(
-  float32_t * pSrcCmplx,
-  float32_t * pSrcReal,
-  float32_t * pCmplxDst,
-  uint32_t numSamples)
+  const float32_t * pSrcCmplx,
+  const float32_t * pSrcReal,
+        float32_t * pCmplxDst,
+        uint32_t numSamples)
 {
-  float32_t in;                                  /* Temporary variable to store input value */
-  uint32_t blkCnt;                               /* loop counters */
-
-#if defined (ARM_MATH_DSP)
-
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  float32_t inA1, inA2, inA3, inA4;              /* Temporary variables to hold input data */
-  float32_t inA5, inA6, inA7, inA8;              /* Temporary variables to hold input data */
-  float32_t inB1, inB2, inB3, inB4;              /* Temporary variables to hold input data */
-  float32_t out1, out2, out3, out4;              /* Temporary variables to hold output data */
-  float32_t out5, out6, out7, out8;              /* Temporary variables to hold output data */
-
-  /* loop Unrolling */
-  blkCnt = numSamples >> 2U;
-
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
-    /* read input from complex input buffer */
-    inA1 = pSrcCmplx[0];
-    inA2 = pSrcCmplx[1];
-    /* read input from real input buffer */
-    inB1 = pSrcReal[0];
-
-    /* read input from complex input buffer */
-    inA3 = pSrcCmplx[2];
-
-    /* multiply complex buffer real input with real buffer input */
-    out1 = inA1 * inB1;
-
-    /* read input from complex input buffer */
-    inA4 = pSrcCmplx[3];
+        uint32_t blkCnt;                               /* Loop counter */
+        float32_t in;                                  /* Temporary variable */
 
-    /* multiply complex buffer imaginary input with real buffer input */
-    out2 = inA2 * inB1;
+#if defined(ARM_MATH_NEON)
+    float32x4_t r;
+    float32x4x2_t ab,outCplx;
 
-    /* read input from real input buffer */
-    inB2 = pSrcReal[1];
-    /* read input from complex input buffer */
-    inA5 = pSrcCmplx[4];
+    /* Compute 4 outputs at a time */
+    blkCnt = numSamples >> 2U;
 
-    /* multiply complex buffer real input with real buffer input */
-    out3 = inA3 * inB2;
+    while (blkCnt > 0U)
+    {
+        ab = vld2q_f32(pSrcCmplx);  // load & separate real/imag pSrcA (de-interleave 2)
+        r = vld1q_f32(pSrcReal);  // load & separate real/imag pSrcB
 
-    /* read input from complex input buffer */
-    inA6 = pSrcCmplx[5];
-    /* read input from real input buffer */
-    inB3 = pSrcReal[2];
+	/* Increment pointers */
+        pSrcCmplx += 8;
+        pSrcReal += 4;
 
-    /* multiply complex buffer imaginary input with real buffer input */
-    out4 = inA4 * inB2;
+        outCplx.val[0] = vmulq_f32(ab.val[0], r);
+        outCplx.val[1] = vmulq_f32(ab.val[1], r);
 
-    /* read input from complex input buffer */
-    inA7 = pSrcCmplx[6];
+        vst2q_f32(pCmplxDst, outCplx);
+        pCmplxDst += 8;
 
-    /* multiply complex buffer real input with real buffer input */
-    out5 = inA5 * inB3;
+        blkCnt--;
+    }
 
-    /* read input from complex input buffer */
-    inA8 = pSrcCmplx[7];
-
-    /* multiply complex buffer imaginary input with real buffer input */
-    out6 = inA6 * inB3;
-
-    /* read input from real input buffer */
-    inB4 = pSrcReal[3];
-
-    /* store result to destination bufer */
-    pCmplxDst[0] = out1;
-
-    /* multiply complex buffer real input with real buffer input */
-    out7 = inA7 * inB4;
-
-    /* store result to destination bufer */
-    pCmplxDst[1] = out2;
-
-    /* multiply complex buffer imaginary input with real buffer input */
-    out8 = inA8 * inB4;
+    /* Tail */
+    blkCnt = numSamples & 3;
+#else
+#if defined (ARM_MATH_LOOPUNROLL)
 
-    /* store result to destination bufer */
-    pCmplxDst[2] = out3;
-    pCmplxDst[3] = out4;
-    pCmplxDst[4] = out5;
+  /* Loop unrolling: Compute 4 outputs at a time */
+  blkCnt = numSamples >> 2U;
 
-    /* incremnet complex input buffer by 8 to process next samples */
-    pSrcCmplx += 8U;
+  while (blkCnt > 0U)
+  {
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
 
-    /* store result to destination bufer */
-    pCmplxDst[5] = out6;
+    in = *pSrcReal++;
+    /* store result in destination buffer. */
+    *pCmplxDst++ = *pSrcCmplx++ * in;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
 
-    /* increment real input buffer by 4 to process next samples */
-    pSrcReal += 4U;
+    in = *pSrcReal++;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
 
-    /* store result to destination bufer */
-    pCmplxDst[6] = out7;
-    pCmplxDst[7] = out8;
+    in = *pSrcReal++;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
 
-    /* increment destination buffer by 8 to process next sampels */
-    pCmplxDst += 8U;
+    in = *pSrcReal++;
+    *pCmplxDst++ = *pSrcCmplx++* in;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
 
-    /* Decrement the numSamples loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
   blkCnt = numSamples;
 
-#endif /* #if defined (ARM_MATH_DSP) */
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+#endif /* #if defined(ARM_MATH_NEON) */
 
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
+
     in = *pSrcReal++;
-    /* store the result in the destination buffer. */
-    *pCmplxDst++ = (*pSrcCmplx++) * (in);
-    *pCmplxDst++ = (*pSrcCmplx++) * (in);
+    /* store result in destination buffer. */
+    *pCmplxDst++ = *pSrcCmplx++ * in;
+    *pCmplxDst++ = *pSrcCmplx++ * in;
 
-    /* Decrement the numSamples loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
+
 }
 
 /**
- * @} end of CmplxByRealMult group
+  @} end of CmplxByRealMult group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q15.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q15.c
index 340d852..4877d20 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q15.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q15.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_real_q15.c
  * Description:  Q15 complex by real multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,77 +29,71 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup CmplxByRealMult
- * @{
+  @addtogroup CmplxByRealMult
+  @{
  */
 
-
 /**
- * @brief  Q15 complex-by-real multiplication
- * @param[in]  *pSrcCmplx points to the complex input vector
- * @param[in]  *pSrcReal points to the real input vector
- * @param[out]  *pCmplxDst points to the complex output vector
- * @param[in]  numSamples number of samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function uses saturating arithmetic.
- * Results outside of the allowable Q15 range [0x8000 0x7FFF] will be saturated.
+  @brief         Q15 complex-by-real multiplication.
+  @param[in]     pSrcCmplx   points to complex input vector
+  @param[in]     pSrcReal    points to real input vector
+  @param[out]    pCmplxDst   points to complex output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function uses saturating arithmetic.
+                   Results outside of the allowable Q15 range [0x8000 0x7FFF] are saturated.
  */
 
 void arm_cmplx_mult_real_q15(
-  q15_t * pSrcCmplx,
-  q15_t * pSrcReal,
-  q15_t * pCmplxDst,
-  uint32_t numSamples)
+  const q15_t * pSrcCmplx,
+  const q15_t * pSrcReal,
+        q15_t * pCmplxDst,
+        uint32_t numSamples)
 {
-  q15_t in;                                      /* Temporary variable to store input value */
+        uint32_t blkCnt;                               /* Loop counter */
+        q15_t in;                                      /* Temporary variable */
 
-#if defined (ARM_MATH_DSP)
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counters */
-  q31_t inA1, inA2;                              /* Temporary variables to hold input data */
-  q31_t inB1;                                    /* Temporary variables to hold input data */
-  q15_t out1, out2, out3, out4;                  /* Temporary variables to hold output data */
-  q31_t mul1, mul2, mul3, mul4;                  /* Temporary variables to hold intermediate data */
+#if defined (ARM_MATH_DSP)
+        q31_t inA1, inA2;                              /* Temporary variables to hold input data */
+        q31_t inB1;                                    /* Temporary variables to hold input data */
+        q15_t out1, out2, out3, out4;                  /* Temporary variables to hold output data */
+        q31_t mul1, mul2, mul3, mul4;                  /* Temporary variables to hold intermediate data */
+#endif
 
-  /* loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
-    /* read complex number both real and imaginary from complex input buffer */
-    inA1 = *__SIMD32(pSrcCmplx)++;
-    /* read two real values at a time from real input buffer */
-    inB1 = *__SIMD32(pSrcReal)++;
-    /* read complex number both real and imaginary from complex input buffer */
-    inA2 = *__SIMD32(pSrcCmplx)++;
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
+
+#if defined (ARM_MATH_DSP)
+    /* read 2 complex numbers both real and imaginary from complex input buffer */
+    inA1 = read_q15x2_ia ((q15_t **) &pSrcCmplx);
+    inA2 = read_q15x2_ia ((q15_t **) &pSrcCmplx);
+    /* read 2 real values at a time from real input buffer */
+    inB1 = read_q15x2_ia ((q15_t **) &pSrcReal);
 
     /* multiply complex number with real numbers */
 #ifndef ARM_MATH_BIG_ENDIAN
-
-    mul1 = (q31_t) ((q15_t) (inA1) * (q15_t) (inB1));
+    mul1 = (q31_t) ((q15_t) (inA1)       * (q15_t) (inB1));
     mul2 = (q31_t) ((q15_t) (inA1 >> 16) * (q15_t) (inB1));
-    mul3 = (q31_t) ((q15_t) (inA2) * (q15_t) (inB1 >> 16));
+    mul3 = (q31_t) ((q15_t) (inA2)       * (q15_t) (inB1 >> 16));
     mul4 = (q31_t) ((q15_t) (inA2 >> 16) * (q15_t) (inB1 >> 16));
-
 #else
-
     mul2 = (q31_t) ((q15_t) (inA1 >> 16) * (q15_t) (inB1 >> 16));
-    mul1 = (q31_t) ((q15_t) inA1 * (q15_t) (inB1 >> 16));
+    mul1 = (q31_t) ((q15_t) inA1         * (q15_t) (inB1 >> 16));
     mul4 = (q31_t) ((q15_t) (inA2 >> 16) * (q15_t) inB1);
-    mul3 = (q31_t) ((q15_t) inA2 * (q15_t) inB1);
-
+    mul3 = (q31_t) ((q15_t) inA2         * (q15_t) inB1);
 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
 
     /* saturate the result */
@@ -109,27 +103,23 @@ void arm_cmplx_mult_real_q15(
     out4 = (q15_t) __SSAT(mul4 >> 15U, 16);
 
     /* pack real and imaginary outputs and store them to destination */
-    *__SIMD32(pCmplxDst)++ = __PKHBT(out1, out2, 16);
-    *__SIMD32(pCmplxDst)++ = __PKHBT(out3, out4, 16);
+    write_q15x2_ia (&pCmplxDst, __PKHBT(out1, out2, 16));
+    write_q15x2_ia (&pCmplxDst, __PKHBT(out3, out4, 16));
 
-    inA1 = *__SIMD32(pSrcCmplx)++;
-    inB1 = *__SIMD32(pSrcReal)++;
-    inA2 = *__SIMD32(pSrcCmplx)++;
+    inA1 = read_q15x2_ia ((q15_t **) &pSrcCmplx);
+    inA2 = read_q15x2_ia ((q15_t **) &pSrcCmplx);
+    inB1 = read_q15x2_ia ((q15_t **) &pSrcReal);
 
 #ifndef ARM_MATH_BIG_ENDIAN
-
-    mul1 = (q31_t) ((q15_t) (inA1) * (q15_t) (inB1));
+    mul1 = (q31_t) ((q15_t) (inA1)       * (q15_t) (inB1));
     mul2 = (q31_t) ((q15_t) (inA1 >> 16) * (q15_t) (inB1));
-    mul3 = (q31_t) ((q15_t) (inA2) * (q15_t) (inB1 >> 16));
+    mul3 = (q31_t) ((q15_t) (inA2)       * (q15_t) (inB1 >> 16));
     mul4 = (q31_t) ((q15_t) (inA2 >> 16) * (q15_t) (inB1 >> 16));
-
 #else
-
     mul2 = (q31_t) ((q15_t) (inA1 >> 16) * (q15_t) (inB1 >> 16));
-    mul1 = (q31_t) ((q15_t) inA1 * (q15_t) (inB1 >> 16));
+    mul1 = (q31_t) ((q15_t) inA1         * (q15_t) (inB1 >> 16));
     mul4 = (q31_t) ((q15_t) (inA2 >> 16) * (q15_t) inB1);
     mul3 = (q31_t) ((q15_t) inA2 * (q15_t) inB1);
-
 #endif /* #ifndef ARM_MATH_BIG_ENDIAN */
 
     out1 = (q15_t) __SSAT(mul1 >> 15U, 16);
@@ -137,55 +127,56 @@ void arm_cmplx_mult_real_q15(
     out3 = (q15_t) __SSAT(mul3 >> 15U, 16);
     out4 = (q15_t) __SSAT(mul4 >> 15U, 16);
 
-    *__SIMD32(pCmplxDst)++ = __PKHBT(out1, out2, 16);
-    *__SIMD32(pCmplxDst)++ = __PKHBT(out3, out4, 16);
+    write_q15x2_ia (&pCmplxDst, __PKHBT(out1, out2, 16));
+    write_q15x2_ia (&pCmplxDst, __PKHBT(out3, out4, 16));
+#else
+    in = *pSrcReal++;
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
 
-    /* Decrement the numSamples loop counter */
-    blkCnt--;
-  }
+    in = *pSrcReal++;
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
-  blkCnt = numSamples % 0x4U;
+    in = *pSrcReal++;
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
 
-  while (blkCnt > 0U)
-  {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
     in = *pSrcReal++;
-    /* store the result in the destination buffer. */
-    *pCmplxDst++ =
-      (q15_t) __SSAT((((q31_t) (*pSrcCmplx++) * (in)) >> 15), 16);
-    *pCmplxDst++ =
-      (q15_t) __SSAT((((q31_t) (*pSrcCmplx++) * (in)) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+#endif
 
-    /* Decrement the numSamples loop counter */
+    /* Decrement loop counter */
     blkCnt--;
   }
 
+  /* Loop unrolling: Compute remaining outputs */
+  blkCnt = numSamples % 0x4U;
+
 #else
 
-  /* Run the below code for Cortex-M0 */
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
+
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
 
-  while (numSamples > 0U)
+  while (blkCnt > 0U)
   {
-    /* realOut = realA * realB.            */
-    /* imagOut = imagA * realB.                */
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
+
     in = *pSrcReal++;
     /* store the result in the destination buffer. */
-    *pCmplxDst++ =
-      (q15_t) __SSAT((((q31_t) (*pSrcCmplx++) * (in)) >> 15), 16);
-    *pCmplxDst++ =
-      (q15_t) __SSAT((((q31_t) (*pSrcCmplx++) * (in)) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
+    *pCmplxDst++ = (q15_t) __SSAT((((q31_t) *pSrcCmplx++ * in) >> 15), 16);
 
-    /* Decrement the numSamples loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of CmplxByRealMult group
+  @} end of CmplxByRealMult group
  */
diff --git a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q31.c b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q31.c
index 19fc55b..906410f 100644
--- a/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q31.c
+++ b/DSP/Source/ComplexMathFunctions/arm_cmplx_mult_real_q31.c
@@ -3,13 +3,13 @@
  * Title:        arm_cmplx_mult_real_q31.c
  * Description:  Q31 complex by real multiplication
  *
- * $Date:        27. January 2017
- * $Revision:    V.1.5.1
+ * $Date:        18. March 2019
+ * $Revision:    V1.6.0
  *
  * Target Processor: Cortex-M cores
  * -------------------------------------------------------------------- */
 /*
- * Copyright (C) 2010-2017 ARM Limited or its affiliates. All rights reserved.
+ * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
  *
  * SPDX-License-Identifier: Apache-2.0
  *
@@ -29,183 +29,120 @@
 #include "arm_math.h"
 
 /**
- * @ingroup groupCmplxMath
+  @ingroup groupCmplxMath
  */
 
 /**
- * @addtogroup CmplxByRealMult
- * @{
+  @addtogroup CmplxByRealMult
+  @{
  */
 
-
 /**
- * @brief  Q31 complex-by-real multiplication
- * @param[in]  *pSrcCmplx points to the complex input vector
- * @param[in]  *pSrcReal points to the real input vector
- * @param[out]  *pCmplxDst points to the complex output vector
- * @param[in]  numSamples number of samples in each vector
- * @return none.
- *
- * <b>Scaling and Overflow Behavior:</b>
- * \par
- * The function uses saturating arithmetic.
- * Results outside of the allowable Q31 range[0x80000000 0x7FFFFFFF] will be saturated.
+  @brief         Q31 complex-by-real multiplication.
+  @param[in]     pSrcCmplx   points to complex input vector
+  @param[in]     pSrcReal    points to real input vector
+  @param[out]    pCmplxDst   points to complex output vector
+  @param[in]     numSamples  number of samples in each vector
+  @return        none
+
+  @par           Scaling and Overflow Behavior
+                   The function uses saturating arithmetic.
+                   Results outside of the allowable Q31 range[0x80000000 0x7FFFFFFF] are saturated.
  */
 
 void arm_cmplx_mult_real_q31(
-  q31_t * pSrcCmplx,
-  q31_t * pSrcReal,
-  q31_t * pCmplxDst,
-  uint32_t numSamples)
+  const q31_t * pSrcCmplx,
+  const q31_t * pSrcReal,
+        q31_t * pCmplxDst,
+        uint32_t numSamples)
 {
-  q31_t inA1;                                    /* Temporary variable to store input value */
-
-#if defined (ARM_MATH_DSP)
+        uint32_t blkCnt;                               /* Loop counter */
+        q31_t in;                                      /* Temporary variable */
 
-  /* Run the below code for Cortex-M4 and Cortex-M3 */
-  uint32_t blkCnt;                               /* loop counters */
-  q31_t inA2, inA3, inA4;                        /* Temporary variables to hold input data */
-  q31_t inB1, inB2;                              /* Temporary variabels to hold input data */
-  q31_t out1, out2, out3, out4;                  /* Temporary variables to hold output data */
+#if defined (ARM_MATH_LOOPUNROLL)
 
-  /* loop Unrolling */
+  /* Loop unrolling: Compute 4 outputs at a time */
   blkCnt = numSamples >> 2U;
 
-  /* First part of the processing with loop unrolling.  Compute 4 outputs at a time.
-   ** a second loop below computes the remaining 1 to 3 samples. */
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
-    /* read real input from complex input buffer */
-    inA1 = *pSrcCmplx++;
-    inA2 = *pSrcCmplx++;
-    /* read input from real input bufer */
-    inB1 = *pSrcReal++;
-    inB2 = *pSrcReal++;
-    /* read imaginary input from complex input buffer */
-    inA3 = *pSrcCmplx++;
-    inA4 = *pSrcCmplx++;
-
-    /* multiply complex input with real input */
-    out1 = ((q63_t) inA1 * inB1) >> 32;
-    out2 = ((q63_t) inA2 * inB1) >> 32;
-    out3 = ((q63_t) inA3 * inB2) >> 32;
-    out4 = ((q63_t) inA4 * inB2) >> 32;
-
-    /* sature the result */
-    out1 = __SSAT(out1, 31);
-    out2 = __SSAT(out2, 31);
-    out3 = __SSAT(out3, 31);
-    out4 = __SSAT(out4, 31);
-
-    /* get result in 1.31 format */
-    out1 = out1 << 1;
-    out2 = out2 << 1;
-    out3 = out3 << 1;
-    out4 = out4 << 1;
-
-    /* store the result to destination buffer */
-    *pCmplxDst++ = out1;
-    *pCmplxDst++ = out2;
-    *pCmplxDst++ = out3;
-    *pCmplxDst++ = out4;
-
-    /* read real input from complex input buffer */
-    inA1 = *pSrcCmplx++;
-    inA2 = *pSrcCmplx++;
-    /* read input from real input bufer */
-    inB1 = *pSrcReal++;
-    inB2 = *pSrcReal++;
-    /* read imaginary input from complex input buffer */
-    inA3 = *pSrcCmplx++;
-    inA4 = *pSrcCmplx++;
-
-    /* multiply complex input with real input */
-    out1 = ((q63_t) inA1 * inB1) >> 32;
-    out2 = ((q63_t) inA2 * inB1) >> 32;
-    out3 = ((q63_t) inA3 * inB2) >> 32;
-    out4 = ((q63_t) inA4 * inB2) >> 32;
-
-    /* sature the result */
-    out1 = __SSAT(out1, 31);
-    out2 = __SSAT(out2, 31);
-    out3 = __SSAT(out3, 31);
-    out4 = __SSAT(out4, 31);
-
-    /* get result in 1.31 format */
-    out1 = out1 << 1;
-    out2 = out2 << 1;
-    out3 = out3 << 1;
-    out4 = out4 << 1;
-
-    /* store the result to destination buffer */
-    *pCmplxDst++ = out1;
-    *pCmplxDst++ = out2;
-    *pCmplxDst++ = out3;
-    *pCmplxDst++ = out4;
-
-    /* Decrement the numSamples loop counter */
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
+
+    in = *pSrcReal++;
+#if defined (ARM_MATH_DSP)
+    /* store saturated result in 1.31 format to destination buffer */
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+#else
+    /* store result in destination buffer. */
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+#endif
+
+    in = *pSrcReal++;
+#if defined (ARM_MATH_DSP)
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+#else
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+#endif
+
+    in = *pSrcReal++;
+#if defined (ARM_MATH_DSP)
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+#else
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+#endif
+
+    in = *pSrcReal++;
+#if defined (ARM_MATH_DSP)
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+#else
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+#endif
+
+    /* Decrement loop counter */
     blkCnt--;
   }
 
-  /* If the numSamples is not a multiple of 4, compute any remaining output samples here.
-   ** No loop unrolling is used. */
+  /* Loop unrolling: Compute remaining outputs */
   blkCnt = numSamples % 0x4U;
 
+#else
+
+  /* Initialize blkCnt with number of samples */
+  blkCnt = numSamples;
+
+#endif /* #if defined (ARM_MATH_LOOPUNROLL) */
+
   while (blkCnt > 0U)
   {
-    /* C[2 * i] = A[2 * i] * B[i].            */
-    /* C[2 * i + 1] = A[2 * i + 1] * B[i].        */
-    /* read real input from complex input buffer */
-    inA1 = *pSrcCmplx++;
-    inA2 = *pSrcCmplx++;
-    /* read input from real input bufer */
-    inB1 = *pSrcReal++;
-
-    /* multiply complex input with real input */
-    out1 = ((q63_t) inA1 * inB1) >> 32;
-    out2 = ((q63_t) inA2 * inB1) >> 32;
-
-    /* sature the result */
-    out1 = __SSAT(out1, 31);
-    out2 = __SSAT(out2, 31);
-
-    /* get result in 1.31 format */
-    out1 = out1 << 1;
-    out2 = out2 << 1;
-
-    /* store the result to destination buffer */
-    *pCmplxDst++ = out1;
-    *pCmplxDst++ = out2;
-
-    /* Decrement the numSamples loop counter */
-    blkCnt--;
-  }
+    /* C[2 * i    ] = A[2 * i    ] * B[i]. */
+    /* C[2 * i + 1] = A[2 * i + 1] * B[i]. */
 
+    in = *pSrcReal++;
+#if defined (ARM_MATH_DSP)
+    /* store saturated result in 1.31 format to destination buffer */
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
+    *pCmplxDst++ = (__SSAT((q31_t) (((q63_t) *pSrcCmplx++ * in) >> 32), 31) << 1);
 #else
+    /* store result in destination buffer. */
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+    *pCmplxDst++ = (q31_t) clip_q63_to_q31(((q63_t) *pSrcCmplx++ * in) >> 31);
+#endif
 
-  /* Run the below code for Cortex-M0 */
-
-  while (numSamples > 0U)
-  {
-    /* realOut = realA * realB.            */
-    /* imagReal = imagA * realB.               */
-    inA1 = *pSrcReal++;
-    /* store the result in the destination buffer. */
-    *pCmplxDst++ =
-      (q31_t) clip_q63_to_q31(((q63_t) * pSrcCmplx++ * inA1) >> 31);
-    *pCmplxDst++ =
-      (q31_t) clip_q63_to_q31(((q63_t) * pSrcCmplx++ * inA1) >> 31);
-
-    /* Decrement the numSamples loop counter */
-    numSamples--;
+    /* Decrement loop counter */
+    blkCnt--;
   }
 
-#endif /* #if defined (ARM_MATH_DSP) */
-
 }
 
 /**
- * @} end of CmplxByRealMult group
+  @} end of CmplxByRealMult group
  */
-- 
cgit