Normalized LMS Filters

Functions
void	arm_lms_norm_f32 (arm_lms_norm_instance_f32 *S, float32_t *pSrc, float32_t *pRef, float32_t *pOut, float32_t *pErr, uint32_t blockSize)
	Processing function for floating-point normalized LMS filter. More...
void	arm_lms_norm_init_f32 (arm_lms_norm_instance_f32 *S, uint16_t numTaps, float32_t *pCoeffs, float32_t *pState, float32_t mu, uint32_t blockSize)
	Initialization function for floating-point normalized LMS filter. More...
void	arm_lms_norm_init_q15 (arm_lms_norm_instance_q15 *S, uint16_t numTaps, q15_t *pCoeffs, q15_t *pState, q15_t mu, uint32_t blockSize, uint8_t postShift)
	Initialization function for Q15 normalized LMS filter. More...
void	arm_lms_norm_init_q31 (arm_lms_norm_instance_q31 *S, uint16_t numTaps, q31_t *pCoeffs, q31_t *pState, q31_t mu, uint32_t blockSize, uint8_t postShift)
	Initialization function for Q31 normalized LMS filter. More...
void	arm_lms_norm_q15 (arm_lms_norm_instance_q15 *S, q15_t *pSrc, q15_t *pRef, q15_t *pOut, q15_t *pErr, uint32_t blockSize)
	Processing function for Q15 normalized LMS filter. More...
void	arm_lms_norm_q31 (arm_lms_norm_instance_q31 *S, q31_t *pSrc, q31_t *pRef, q31_t *pOut, q31_t *pErr, uint32_t blockSize)
	Processing function for Q31 normalized LMS filter. More...

Description

This set of functions implements a commonly used adaptive filter. It is related to the Least Mean Square (LMS) adaptive filter and includes an additional normalization factor which increases the adaptation rate of the filter. The CMSIS DSP Library contains normalized LMS filter functions that operate on Q15, Q31, and floating-point data types.

A normalized least mean square (NLMS) filter consists of two components as shown below. The first component is a standard transversal or FIR filter. The second component is a coefficient update mechanism. The NLMS filter has two input signals. The "input" feeds the FIR filter while the "reference input" corresponds to the desired output of the FIR filter. That is, the FIR filter coefficients are updated so that the output of the FIR filter matches the reference input. The filter coefficient update mechanism is based on the difference between the FIR filter output and the reference input. This "error signal" tends towards zero as the filter adapts. The NLMS processing functions accept the input and reference input signals and generate the filter output and error signal.

Internal structure of the NLMS adaptive filter

The functions operate on blocks of data and each call to the function processes blockSize samples through the filter. pSrc points to input signal, pRef points to reference signal, pOut points to output signal and pErr points to error signal. All arrays contain blockSize values.

The functions operate on a block-by-block basis. Internally, the filter coefficients b[n] are updated on a sample-by-sample basis. The convergence of the LMS filter is slower compared to the normalized LMS algorithm.

Algorithm:

The output signal y[n] is computed by a standard FIR filter:

     y[n] = b[0] * x[n] + b[1] * x[n-1] + b[2] * x[n-2] + ...+ b[numTaps-1] * x[n-numTaps+1]
 

The error signal equals the difference between the reference signal d[n] and the filter output:

     e[n] = d[n] - y[n].
 

After each sample of the error signal is computed the instanteous energy of the filter state variables is calculated:

    E = x[n]^2 + x[n-1]^2 + ... + x[n-numTaps+1]^2.
 

The filter coefficients b[k] are then updated on a sample-by-sample basis:

     b[k] = b[k] + e[n] * (mu/E) * x[n-k],  for k=0, 1, ..., numTaps-1
 

where mu is the step size and controls the rate of coefficient convergence.

In the APIs, pCoeffs points to a coefficient array of size numTaps. Coefficients are stored in time reversed order.

    {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}
 

pState points to a state array of size numTaps + blockSize - 1. Samples in the state buffer are stored in the order:

    {x[n-numTaps+1], x[n-numTaps], x[n-numTaps-1], x[n-numTaps-2]....x[0], x[1], ..., x[blockSize-1]}
 

Note that the length of the state buffer exceeds the length of the coefficient array by blockSize-1 samples. The increased state buffer length allows circular addressing, which is traditionally used in FIR filters, to be avoided and yields a significant speed improvement. The state variables are updated after each block of data is processed.

Instance Structure

The coefficients and state variables for a filter are stored together in an instance data structure. A separate instance structure must be defined for each filter and coefficient and state arrays cannot be shared among instances. There are separate instance structure declarations for each of the 3 supported data types.

Initialization Functions

There is also an associated initialization function for each data type. The initialization function performs the following operations:

• Sets the values of the internal structure fields.
• Zeros out the values in the state buffer. To do this manually without calling the init function, assign the follow subfields of the instance structure: numTaps, pCoeffs, mu, energy, x0, pState. Also set all of the values in pState to zero. For Q7, Q15, and Q31 the following fields must also be initialized; recipTable, postShift

Instance structure cannot be placed into a const data section and it is recommended to use the initialization function.

Fixed-Point Behavior:

Care must be taken when using the Q15 and Q31 versions of the normalised LMS filter. The following issues must be considered:

• Scaling of coefficients
• Overflow and saturation

Scaling of Coefficients:

Filter coefficients are represented as fractional values and coefficients are restricted to lie in the range [-1 +1). The fixed-point functions have an additional scaling parameter postShift. At the output of the filter's accumulator is a shift register which shifts the result by postShift bits. This essentially scales the filter coefficients by 2^postShift and allows the filter coefficients to exceed the range [+1 -1). The value of postShift is set by the user based on the expected gain through the system being modeled.

Overflow and Saturation:

Overflow and saturation behavior of the fixed-point Q15 and Q31 versions are described separately as part of the function specific documentation below.

Function Documentation

void arm_lms_norm_f32	(	arm_lms_norm_instance_f32 *	S,
		float32_t *	pSrc,
		float32_t *	pRef,
		float32_t *	pOut,
		float32_t *	pErr,
		uint32_t	blockSize
	)

Parameters

[in]	*S	points to an instance of the floating-point normalized LMS filter structure.
[in]	*pSrc	points to the block of input data.
[in]	*pRef	points to the block of reference data.
[out]	*pOut	points to the block of output data.
[out]	*pErr	points to the block of error data.
[in]	blockSize	number of samples to process.

Returns

none.

References blockSize, arm_lms_norm_instance_f32::energy, arm_lms_norm_instance_f32::mu, arm_lms_norm_instance_f32::numTaps, arm_lms_norm_instance_f32::pCoeffs, arm_lms_norm_instance_f32::pState, and arm_lms_norm_instance_f32::x0.

Referenced by main().

void arm_lms_norm_init_f32	(	arm_lms_norm_instance_f32 *	S,
		uint16_t	numTaps,
		float32_t *	pCoeffs,
		float32_t *	pState,
		float32_t	mu,
		uint32_t	blockSize
	)

Parameters

[in]	*S	points to an instance of the floating-point LMS filter structure.
[in]	numTaps	number of filter coefficients.
[in]	*pCoeffs	points to coefficient buffer.
[in]	*pState	points to state buffer.
[in]	mu	step size that controls filter coefficient updates.
[in]	blockSize	number of samples to process.

Returns

none.

Description:

pCoeffs points to the array of filter coefficients stored in time reversed order:

   {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}

The initial filter coefficients serve as a starting point for the adaptive filter. pState points to an array of length numTaps+blockSize-1 samples, where blockSize is the number of input samples processed by each call to arm_lms_norm_f32().

References arm_lms_norm_instance_f32::energy, arm_lms_norm_instance_f32::mu, arm_lms_norm_instance_f32::numTaps, arm_lms_norm_instance_f32::pCoeffs, arm_lms_norm_instance_f32::pState, and arm_lms_norm_instance_f32::x0.

Referenced by main().

void arm_lms_norm_init_q15	(	arm_lms_norm_instance_q15 *	S,
		uint16_t	numTaps,
		q15_t *	pCoeffs,
		q15_t *	pState,
		q15_t	mu,
		uint32_t	blockSize,
		uint8_t	postShift
	)

Parameters

[in]	*S	points to an instance of the Q15 normalized LMS filter structure.
[in]	numTaps	number of filter coefficients.
[in]	*pCoeffs	points to coefficient buffer.
[in]	*pState	points to state buffer.
[in]	mu	step size that controls filter coefficient updates.
[in]	blockSize	number of samples to process.
[in]	postShift	bit shift applied to coefficients.

Returns

none.

Description:

pCoeffs points to the array of filter coefficients stored in time reversed order:

   {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}

The initial filter coefficients serve as a starting point for the adaptive filter. pState points to the array of state variables and size of array is numTaps+blockSize-1 samples, where blockSize is the number of input samples processed by each call to arm_lms_norm_q15().

References armRecipTableQ15, arm_lms_norm_instance_q15::energy, arm_lms_norm_instance_q15::mu, arm_lms_norm_instance_q15::numTaps, arm_lms_norm_instance_q15::pCoeffs, arm_lms_norm_instance_q15::postShift, arm_lms_norm_instance_q15::pState, arm_lms_norm_instance_q15::recipTable, and arm_lms_norm_instance_q15::x0.

void arm_lms_norm_init_q31	(	arm_lms_norm_instance_q31 *	S,
		uint16_t	numTaps,
		q31_t *	pCoeffs,
		q31_t *	pState,
		q31_t	mu,
		uint32_t	blockSize,
		uint8_t	postShift
	)

Parameters

[in]	*S	points to an instance of the Q31 normalized LMS filter structure.
[in]	numTaps	number of filter coefficients.
[in]	*pCoeffs	points to coefficient buffer.
[in]	*pState	points to state buffer.
[in]	mu	step size that controls filter coefficient updates.
[in]	blockSize	number of samples to process.
[in]	postShift	bit shift applied to coefficients.

Returns

none.

Description:

pCoeffs points to the array of filter coefficients stored in time reversed order:

   {b[numTaps-1], b[numTaps-2], b[N-2], ..., b[1], b[0]}

The initial filter coefficients serve as a starting point for the adaptive filter. pState points to an array of length numTaps+blockSize-1 samples, where blockSize is the number of input samples processed by each call to arm_lms_norm_q31().

References armRecipTableQ31, arm_lms_norm_instance_q31::energy, arm_lms_norm_instance_q31::mu, arm_lms_norm_instance_q31::numTaps, arm_lms_norm_instance_q31::pCoeffs, arm_lms_norm_instance_q31::postShift, arm_lms_norm_instance_q31::pState, arm_lms_norm_instance_q31::recipTable, and arm_lms_norm_instance_q31::x0.

void arm_lms_norm_q15	(	arm_lms_norm_instance_q15 *	S,
		q15_t *	pSrc,
		q15_t *	pRef,
		q15_t *	pOut,
		q15_t *	pErr,
		uint32_t	blockSize
	)

Parameters

[in]	*S	points to an instance of the Q15 normalized LMS filter structure.
[in]	*pSrc	points to the block of input data.
[in]	*pRef	points to the block of reference data.
[out]	*pOut	points to the block of output data.
[out]	*pErr	points to the block of error data.
[in]	blockSize	number of samples to process.

Returns

none.

Scaling and Overflow Behavior:

The function is implemented using a 64-bit internal accumulator. Both coefficients and state variables are represented in 1.15 format and multiplications yield a 2.30 result. The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format. There is no risk of internal overflow with this approach and the full precision of intermediate multiplications is preserved. After all additions have been performed, the accumulator is truncated to 34.15 format by discarding low 15 bits. Lastly, the accumulator is saturated to yield a result in 1.15 format.

In this filter, filter coefficients are updated for each sample and the updation of filter cofficients are saturted.

References __SIMD32, __SMLALD(), arm_recip_q15(), blockSize, DELTA_Q15, arm_lms_norm_instance_q15::energy, arm_lms_norm_instance_q15::mu, arm_lms_norm_instance_q15::numTaps, arm_lms_norm_instance_q15::pCoeffs, arm_lms_norm_instance_q15::postShift, arm_lms_norm_instance_q15::pState, arm_lms_norm_instance_q15::recipTable, and arm_lms_norm_instance_q15::x0.

void arm_lms_norm_q31	(	arm_lms_norm_instance_q31 *	S,
		q31_t *	pSrc,
		q31_t *	pRef,
		q31_t *	pOut,
		q31_t *	pErr,
		uint32_t	blockSize
	)

Parameters

[in]	*S	points to an instance of the Q31 normalized LMS filter structure.
[in]	*pSrc	points to the block of input data.
[in]	*pRef	points to the block of reference data.
[out]	*pOut	points to the block of output data.
[out]	*pErr	points to the block of error data.
[in]	blockSize	number of samples to process.

Returns

none.

Scaling and Overflow Behavior:

The function is implemented using an internal 64-bit accumulator. The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit. Thus, if the accumulator result overflows it wraps around rather than clip. In order to avoid overflows completely the input signal must be scaled down by log2(numTaps) bits. The reference signal should not be scaled down. After all multiply-accumulates are performed, the 2.62 accumulator is shifted and saturated to 1.31 format to yield the final result. The output signal and error signal are in 1.31 format.

In this filter, filter coefficients are updated for each sample and the updation of filter cofficients are saturted.

References arm_recip_q31(), blockSize, clip_q63_to_q31(), DELTA_Q31, arm_lms_norm_instance_q31::energy, arm_lms_norm_instance_q31::mu, arm_lms_norm_instance_q31::numTaps, arm_lms_norm_instance_q31::pCoeffs, arm_lms_norm_instance_q31::postShift, arm_lms_norm_instance_q31::pState, arm_lms_norm_instance_q31::recipTable, and arm_lms_norm_instance_q31::x0.