CMSIS-DSP  Verison 1.1.0
CMSIS DSP Software Library
Complex-by-Complex Multiplication

Functions

void arm_cmplx_mult_cmplx_f32 (float32_t *pSrcA, float32_t *pSrcB, float32_t *pDst, uint32_t numSamples)
 Floating-point complex-by-complex multiplication.
void arm_cmplx_mult_cmplx_q15 (q15_t *pSrcA, q15_t *pSrcB, q15_t *pDst, uint32_t numSamples)
 Q15 complex-by-complex multiplication.
void arm_cmplx_mult_cmplx_q31 (q31_t *pSrcA, q31_t *pSrcB, q31_t *pDst, uint32_t numSamples)
 Q31 complex-by-complex multiplication.

Description

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 numSamples represents the number of complex samples processed. The complex arrays have a total of 2*numSamples real values.

The underlying algorithm is used:

        
 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];        
 }        
 

There are separate functions for floating-point, Q15, and Q31 data types.


Function Documentation

void arm_cmplx_mult_cmplx_f32 ( float32_t pSrcA,
float32_t pSrcB,
float32_t pDst,
uint32_t  numSamples 
)
Parameters:
[in]*pSrcApoints to the first input vector
[in]*pSrcBpoints to the second input vector
[out]*pDstpoints to the output vector
[in]numSamplesnumber of complex samples in each vector
Returns:
none.
Examples:
arm_convolution_example_f32.c.

Referenced by main().

void arm_cmplx_mult_cmplx_q15 ( q15_t pSrcA,
q15_t pSrcB,
q15_t pDst,
uint32_t  numSamples 
)
Parameters:
[in]*pSrcApoints to the first input vector
[in]*pSrcBpoints to the second input vector
[out]*pDstpoints to the output vector
[in]numSamplesnumber of complex samples in each vector
Returns:
none.

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_q31 ( q31_t pSrcA,
q31_t pSrcB,
q31_t pDst,
uint32_t  numSamples 
)
Parameters:
[in]*pSrcApoints to the first input vector
[in]*pSrcBpoints to the second input vector
[out]*pDstpoints to the output vector
[in]numSamplesnumber of complex samples in each vector
Returns:
none.

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.

Referenced by arm_dct4_q31().