398 lines
16 KiB
C
398 lines
16 KiB
C
#include "include/signal_path.h"
|
||
#define BLOCK_LEN 1
|
||
|
||
/* Global variables decleration*/
|
||
static int counter=0;
|
||
static int mu;
|
||
|
||
static int leak=2147462173; //0.999 // (1 ? <20>?)
|
||
|
||
int chess_storage(DMB) fir_lms_delay_line[MAX_FIR_COEFFS]; //Int-Array für Acc-Sensors Samples (Delay Line) anlegen
|
||
int chess_storage(DMA % (sizeof(long long))) fir_lms_coeffs[MAX_FIR_COEFFS]; //Int-Array für Filterkoeffizienten anlegen
|
||
|
||
BufferPtrDMB chess_storage(DMB) ptr_fir_lms_delay_line;
|
||
BufferPtr ptr_fir_lms_coeffs;
|
||
|
||
|
||
|
||
#ifdef PLATFORM_GENERIC
|
||
// lpdsp32 functionallity moddeling functions
|
||
accum_t fract_mult(int a, int b){
|
||
long int a_long = a;
|
||
long int b_long = b;
|
||
return (b_long * a_long);
|
||
}
|
||
accum_t to_accum(int a){
|
||
long int a_long = (long int) a;
|
||
return a_long << 31;
|
||
}
|
||
int rnd_saturate(accum_t a){
|
||
return a >> 31;
|
||
}
|
||
int extract_high(accum_t a){
|
||
return a >> 31;
|
||
}
|
||
void lldecompose(unsigned long long l, int* int1, int* int2){
|
||
*int2 = (int)(l >> 32);
|
||
*int1 = (int)(l);
|
||
}
|
||
uint64_t llcompose(int a, int b) {
|
||
uint64_t result = (uint64_t)b; // Assign b to the higher 32 bits of the result
|
||
result <<= 32; // Shift the higher 32 bits to the left
|
||
result |= (uint32_t)a; // Bitwise OR operation with the lower 32 bits of a
|
||
return result;
|
||
}
|
||
// unsigned long long llcompose(int a, int b){
|
||
// unsigned long long l;
|
||
// l = a << 32;
|
||
// l |= b;
|
||
// return l;
|
||
//}
|
||
int* cyclic_add(int *ptr, int i_pp, int *ptr_start, int buffer_len){
|
||
int *p_ptr=ptr;
|
||
for (int i=0; i < abs(i_pp); i+=1){ // end of buffer wraparound
|
||
if (i_pp > 0){
|
||
p_ptr ++;
|
||
if (p_ptr >= ptr_start + buffer_len){
|
||
p_ptr=ptr_start;
|
||
}
|
||
}
|
||
else{ // start of buffer wraparound
|
||
p_ptr--;
|
||
if (p_ptr < ptr_start){
|
||
p_ptr=ptr_start + (buffer_len -1);
|
||
}
|
||
}
|
||
}
|
||
return p_ptr;
|
||
}
|
||
#endif
|
||
|
||
int sig_init_buffer(BufferPtr *buffer, int *buffer_start_add, int length, int max_buffer_len) {
|
||
buffer->buffer_len = length;
|
||
buffer->ptr_start = buffer_start_add;
|
||
buffer->ptr_current = buffer_start_add;
|
||
// initialize delay line with 0
|
||
for (int i = 0; i < length; i++) {
|
||
buffer_start_add[i] = 0;
|
||
}
|
||
if (length<max_buffer_len){
|
||
return 0;
|
||
}
|
||
else{
|
||
return 1;
|
||
}
|
||
}
|
||
|
||
int sig_init_buffer_DMB(BufferPtrDMB chess_storage(DMB) *buffer, int chess_storage(DMB) *buffer_start_add, int length, int max_buffer_len){
|
||
buffer->buffer_len = length;
|
||
buffer->ptr_start = buffer_start_add;
|
||
buffer->ptr_current = buffer_start_add;
|
||
// initialize delay line with 0
|
||
for (int i = 0; i < length; i++) {
|
||
buffer_start_add[i] = 0;
|
||
}
|
||
if (length<max_buffer_len){
|
||
return 0;
|
||
}
|
||
else{
|
||
return 1;
|
||
}
|
||
}
|
||
|
||
void sig_cirular_buffer_ptr_increment(BufferPtr *buffer, int i_incr){
|
||
buffer->ptr_current = cyclic_add(buffer->ptr_current, i_incr, buffer->ptr_start, buffer->buffer_len);
|
||
}
|
||
|
||
void sig_cirular_buffer_ptr_increment_DMB(BufferPtrDMB *buffer, int i_incr){
|
||
buffer->ptr_current = cyclic_add(buffer->ptr_current, i_incr, buffer->ptr_start, buffer->buffer_len);
|
||
}
|
||
|
||
void sig_cirular_buffer_ptr_put_sample(BufferPtr *buffer, int sample){
|
||
*buffer->ptr_current = sample;
|
||
buffer->ptr_current = cyclic_add(buffer->ptr_current, 1, buffer->ptr_start, buffer->buffer_len);
|
||
}
|
||
|
||
void sig_cirular_buffer_ptr_put_sample_DMB(BufferPtrDMB chess_storage(DMB) *buffer, int sample){
|
||
*buffer->ptr_current = sample; //Sample des Acc-Sensors wird in Adresse geschrieben, auf die der Pointer zeigt
|
||
buffer->ptr_current = cyclic_add(buffer->ptr_current, 1, buffer->ptr_start, buffer->buffer_len); //Pointer wird inkrementiert
|
||
}
|
||
|
||
void static inline sig_circular_buffer_ptr_put_block(BufferPtr *buffer, int* block){
|
||
// increment pointer to oldest block
|
||
//buffer->ptr_current = cyclic_add(buffer->ptr_current, BLOCK_LEN, buffer->ptr_start, buffer->buffer_len);
|
||
// load the next block
|
||
for (int i=0; i<BLOCK_LEN; i+=2){
|
||
buffer->ptr_current[0] = block[i]; // TODO: use llcompose
|
||
buffer->ptr_current[1] = block[i+1];
|
||
buffer->ptr_current = cyclic_add(buffer->ptr_current, 2, buffer->ptr_start, buffer->buffer_len);
|
||
}
|
||
}
|
||
|
||
//Initialisierungsfunktion f<>r Biquad Filter Koeffizienten
|
||
void sig_init_preemph_coef(SingleSignalPath *signal, double b0, double b1, double b2, double a1, double a2, int scale_bits) {
|
||
// Wenn b0=1 und Rest 0 -> kein Filter weil effektiv 1*Xn
|
||
if (b0 == 1. && b1 == 0. && b2 == 0. && a1 == 0. && a2 == 0.) {
|
||
signal->preemph_activated = 0;
|
||
}
|
||
else{
|
||
signal->preemph_activated = 1; // Schreibe Eintrag in Struct
|
||
signal->_preemph_scale_nbits = scale_bits; // Schreibe Eintrag in Struct - wieviel Bits wird skaliert
|
||
int scale = pow(2, scale_bits) - 1; //2^n -1 Skalierung
|
||
// Skaliere Koeffizienten zu Interger und schreibe Eintrag in Struct
|
||
signal->b_preemph[0] = b0 * scale;
|
||
signal->b_preemph[1] = b1 * scale;
|
||
signal->b_preemph[2] = b2 * scale;
|
||
signal->b_preemph[3] = a1 * scale;
|
||
signal->b_preemph[4] = a2 * scale;
|
||
}
|
||
}
|
||
|
||
/*Initialization functions - make sure all of them were called to ensure functionality*/
|
||
int sig_init_delay(SingleSignalPath *signal, int n_delay) {
|
||
return sig_init_buffer(&signal->delay_buffer, signal->_delay_buffer, n_delay, MAX_DELAY_SAMPS);
|
||
}
|
||
|
||
//Initialisierungsfunktion f<>r Gewichtung
|
||
void sig_init_weight(SingleSignalPath *signal, double weight, int scale_nbits) {
|
||
// Wenn Gewichtung 1 -> kein Effekt
|
||
if (weight == 1.) {
|
||
signal->weight_actived = 0;
|
||
}
|
||
// Wenn Gewichtung != 1 -> Zu Integer skalieren und Eintrag in Struct schreiben
|
||
else{
|
||
signal->weight_actived = 1;
|
||
int scale = pow(2, scale_nbits) - 1;
|
||
signal->weight = weight * scale;
|
||
signal->_weight_scale_nbits = scale_nbits;
|
||
}
|
||
}
|
||
|
||
/*Calculator functions for the given signal path*/
|
||
/*Calculate one biquad filter element*/
|
||
int sig_calc_biquad(SingleSignalPath *signal, int x) {
|
||
if (signal->preemph_activated == 0) {
|
||
return x;
|
||
}
|
||
accum_t sum =
|
||
fract_mult(x, signal->b_preemph[0]) + fract_mult(signal->_xd[0], signal->b_preemph[1]) +
|
||
fract_mult(signal->_xd[1], signal->b_preemph[2]) + fract_mult(signal->_yd[0], signal->b_preemph[3]) +
|
||
fract_mult(signal->_yd[1],signal->b_preemph[4]);
|
||
int y = rnd_saturate(sum << 1);
|
||
|
||
|
||
signal->_xd[1] = signal->_xd[0];
|
||
signal->_xd[0] = x;
|
||
signal->_yd[1] = signal->_yd[0];
|
||
signal->_yd[0] = y;
|
||
return y;
|
||
}
|
||
int inline sig_get_delayed_sample(SingleSignalPath *signal) {
|
||
return *signal->delay_buffer.ptr_current;
|
||
}
|
||
|
||
int sig_delay_buffer_load_and_get(SingleSignalPath *signal, int x) {
|
||
if (signal->delay_buffer.buffer_len == 0) {
|
||
return x;
|
||
}
|
||
int out = *signal->delay_buffer.ptr_current;
|
||
*signal->delay_buffer.ptr_current = x;
|
||
sig_cirular_buffer_ptr_increment(&signal->delay_buffer, 1);
|
||
return out;
|
||
}
|
||
|
||
int sig_calc_weight(SingleSignalPath *signal, int x) {
|
||
if (signal->weight_actived == 0) {
|
||
return x;
|
||
}
|
||
accum_t acc = fract_mult(x, signal->weight);
|
||
|
||
return rnd_saturate(acc);
|
||
}
|
||
|
||
int inline sig_calc_fir_lpdsp32_single(BufferPtrDMB chess_storage(DMB) *ptr_fir_lms_delay_line, BufferPtr *ptr_fir_lms_coeffs){
|
||
// Filterkoeffizienten mit Acc-Sensor Samples multiplizieren und aufsummieren um Akkumulator Output des adaptiven Filters zu erhalten
|
||
|
||
//Pointer für Koeffizienten und Delay Line Samples anlegen
|
||
int chess_storage(DMB) *p_x0 = ptr_fir_lms_delay_line->ptr_current;
|
||
int chess_storage(DMB) *px_start = ptr_fir_lms_delay_line->ptr_start;
|
||
int *p_h = ptr_fir_lms_coeffs->ptr_current;
|
||
int delay_line_len = ptr_fir_lms_delay_line->buffer_len;
|
||
int n_coeff = ptr_fir_lms_coeffs->buffer_len;
|
||
|
||
//Variablen und Akkumulatoren (72-Bit) anlegen
|
||
int d0,d1,h0,h1;
|
||
accum_t acc1_A = to_accum(0);
|
||
accum_t acc1_B = to_accum(0);
|
||
accum_t acc1_C;
|
||
|
||
// iterate over the coefficients to calculate the filter on x - the canceller
|
||
/* Abschaetzung cycles per 2coefficient:
|
||
dual - load : 1
|
||
dual mac and dual load: 1
|
||
-> 48/2 * 2 = 48 cycles for 48 coefficents
|
||
*/
|
||
// In 2er Schritten durch die Koeffizienten iterieren, immer 2 Samples und 2 Koeffizienten pro Schleifendurchlauf -> DUAL LOAD und DUAL MAC
|
||
for (int i=0; i < n_coeff; i+=2) chess_loop_range(1,){
|
||
d0 = *p_x0; //Sample 1 aus Delay Line
|
||
h0 = *p_h; //Koeffizient 1 aus Koeffizienten Array
|
||
p_h++; //Koeffizienten-Pointer inkrementieren
|
||
p_x0 = cyclic_add(p_x0, -1, px_start, delay_line_len); //Delay-Line-Pointer dekrementieren (rueckwaerts durch Delay Line)
|
||
|
||
d1 = *p_x0; //Sample 2 aus Delay Line
|
||
h1 = *p_h; //Koeffizient 2 aus Koeffizienten Array
|
||
p_h++; //Koeffizienten-Pointer inkrementieren
|
||
p_x0 = cyclic_add(p_x0, -1, px_start, delay_line_len); //Delay-Line-Pointer dekrementieren (rueckwaerts durch Delay Line)
|
||
|
||
acc1_A+=fract_mult(d0, h0); //Akkumulator 1 mit Sample 1 * Koeffizient 1 addieren
|
||
acc1_B+=fract_mult(d1, h1); //Akkumulator 2 mit Sample 2 * Koeffizient 2 addieren
|
||
}
|
||
// Akkumulatoren addieren um das Filterergebnis zu erhalten
|
||
acc1_C = acc1_A + acc1_B;
|
||
return rnd_saturate(acc1_C);
|
||
}
|
||
|
||
void static inline adapt_coeffs_lpdsp32_single_v1(BufferPtrDMB chess_storage(DMB) *ptr_fir_lms_delay_line, BufferPtr *ptr_fir_lms_coeffs, int out){
|
||
|
||
int chess_storage(DMA) *p_h0 = ptr_fir_lms_coeffs->ptr_start; //Pointer auf Filterkoeffizienten-Array
|
||
int chess_storage(DMB) *p_x0 = ptr_fir_lms_delay_line->ptr_current; //Current-Pointer 1 auf Delay-Line Array
|
||
int chess_storage(DMB) *p_x1 = ptr_fir_lms_delay_line->ptr_current; //Current-Pointer 2 auf Delay-Line Array
|
||
int chess_storage(DMB) *px_start = ptr_fir_lms_delay_line->ptr_start; //Start-Pointer auf Delay-Line Array
|
||
|
||
int delay_line_len = ptr_fir_lms_delay_line->buffer_len; // Länge des Delay-Line Arrays
|
||
int n_coeff = ptr_fir_lms_coeffs->buffer_len; // Anzahl der Filterkoeffizienten
|
||
int prod, x0, x1, h0, h1;
|
||
|
||
p_x1 = cyclic_add(p_x1, -1, ptr_fir_lms_delay_line->ptr_start, ptr_fir_lms_delay_line->buffer_len); //Current-Pointer 2 dekrementieren um 1
|
||
|
||
accum_t acc_A, acc_B;
|
||
|
||
accum_t acc_C = fract_mult(mu, out); //Korrektursignal * mu um Filterkoeffizienten anzupassen
|
||
prod = rnd_saturate(acc_C);
|
||
|
||
/* Abschätzung cycles per 2 coefficient:
|
||
dual load coeffs: 1
|
||
single load tab value: 2
|
||
dual mac: 1
|
||
dual rnd_sat - store: 1
|
||
load/store hazard nop: 1
|
||
*/
|
||
for (int i=0; i< n_coeff; i+=2) chess_loop_range(1,){
|
||
// Calculate the coefficient wise adaption
|
||
#ifdef PLATFORM_GENERIC
|
||
lldecompose(*((long long *)p_h0), &h0, &h1);
|
||
#else
|
||
lldecompose(*((long long *)p_h0), h0, h1);
|
||
#endif
|
||
|
||
acc_A = to_accum(h0);
|
||
acc_B = to_accum(h1);
|
||
|
||
acc_A += fract_mult(prod, *p_x0);
|
||
acc_B += fract_mult(prod, *p_x1);
|
||
|
||
p_x0 = cyclic_add(p_x0, -2, px_start, delay_line_len);
|
||
p_x1 = cyclic_add(p_x1, -2, px_start, delay_line_len);
|
||
|
||
// Filterkoeffizienten updaten - dual sat; dual store
|
||
*((long long *)p_h0) = llcompose(rnd_saturate(acc_A), rnd_saturate(acc_B));//load/store hazard ! - 1 nop is needed
|
||
p_h0+=2;
|
||
}
|
||
}
|
||
|
||
void init(
|
||
SingleSignalPath *cSensorSignal,
|
||
SingleSignalPath *accSensorSignal,
|
||
double *b_c,
|
||
double *b_acc,
|
||
int delay_c,
|
||
int delay_acc,
|
||
double weight_c,
|
||
double weight_acc,
|
||
double lms_mu,
|
||
int lms_fir_num_coeffs
|
||
){
|
||
int scale_bits=31;
|
||
|
||
// C-Sensor Initialisierung: Biquad, Delay, Weight skalieren und in Struct schreiben
|
||
sig_init_preemph_coef(cSensorSignal, b_c[0], b_c[1], b_c[2], b_c[3], b_c[4], scale_bits);
|
||
sig_init_delay(cSensorSignal, delay_c);
|
||
sig_init_weight(cSensorSignal, weight_c, scale_bits);
|
||
|
||
// Acc-Sensor Initialisierung: Biquad, Delay, Weight skalieren und in Struct schreiben
|
||
sig_init_preemph_coef(accSensorSignal, b_acc[0], b_acc[1], b_acc[2], b_acc[3], b_acc[4], scale_bits);
|
||
sig_init_delay(accSensorSignal, delay_acc);
|
||
sig_init_weight(accSensorSignal, weight_acc, 31);
|
||
|
||
//Mu Skalierung und in globale Variable schreiben
|
||
int scale = pow(2, scale_bits) - 1;
|
||
mu = lms_mu * scale;
|
||
// Buffer Initialisierung (Delay Line und Koeffizienten)
|
||
sig_init_buffer_DMB(&ptr_fir_lms_delay_line, fir_lms_delay_line, lms_fir_num_coeffs, MAX_FIR_COEFFS);
|
||
sig_init_buffer(&ptr_fir_lms_coeffs, fir_lms_coeffs, lms_fir_num_coeffs, MAX_FIR_COEFFS);
|
||
|
||
// Einträge in Delay Line und Koeffizienten-Array auf 0 setzen
|
||
for (int i = 0; i < lms_fir_num_coeffs; i++) {
|
||
ptr_fir_lms_delay_line.ptr_start[i] = 0;
|
||
ptr_fir_lms_coeffs.ptr_start[i] = 0;
|
||
}
|
||
}
|
||
|
||
// Data d(cSensor) is signal + noise
|
||
// x (accSensor) is reference noise signal
|
||
void calc(
|
||
SingleSignalPath *cSensorSignal,
|
||
SingleSignalPath *accSensorSignal,
|
||
OutputMode output_mode,
|
||
int16_t volatile chess_storage(DMB) *cSensor, //Pointer auf Input-Port im Shared Memory
|
||
int16_t volatile chess_storage(DMB) *accSensor, //Pointer auf Input-Port im Shared Memory
|
||
int16_t volatile chess_storage(DMB) *out_16 //Pointer auf Output-Port im Shared Memory
|
||
|
||
){
|
||
//Speicherbereiche anlegen -> bei blockweiser Verarbeitung hat jedes Array nur den Eintrag [0]
|
||
static int chess_storage(DMA) c_block_pre[BLOCK_LEN]; //Speicherbereich für C-Sensor Preemphasis Input
|
||
static int chess_storage(DMA) acc_block_pre[BLOCK_LEN]; //Speicherbereich für Acc-Sensor Preemphasis Input
|
||
static int chess_storage(DMA) cSensor_32[BLOCK_LEN]; //Speicherbereich für 32-Bit C-Sensor Input
|
||
static int chess_storage(DMA) accSensor_32[BLOCK_LEN]; //Speicherbereich für 32-Bit Acc-Sensor Input
|
||
|
||
static int chess_storage(DMB) acc_block_filt[BLOCK_LEN]; //Speicherbereich für Akkumulator Output des adaptiven Filters
|
||
static int chess_storage(DMB) out_32[BLOCK_LEN]; //Speicherbereich für 32-Bit Output Signal
|
||
|
||
// Pointer auf die Arrays anlegen
|
||
static int chess_storage(DMA) *p_c_block_pre =c_block_pre;
|
||
static int chess_storage(DMA) *p_acc_block_filt =acc_block_pre;
|
||
static int chess_storage(DMB) *p_out_32=out_32;
|
||
|
||
// 16-Bit Eingangssignale auf 32-Bit konvertieren mit Bitshift, in neuem Speicherbereich ablegen
|
||
for (uint32_t i=0; i<BLOCK_LEN; i++) chess_loop_range(1,){
|
||
cSensor_32[i] = ((int) cSensor[i]) << BITSHIFT_16_TO_32;
|
||
accSensor_32[i] = ((int) accSensor[i]) << BITSHIFT_16_TO_32;
|
||
}
|
||
|
||
// Preemphasis Filter anwenden - wird hier aber nicht genutzt (nur Durchreichen), in neuen Speicherbereich ablegen
|
||
for (uint32_t i=0; i<BLOCK_LEN; i++) chess_loop_range(1,){
|
||
c_block_pre[i] = cSensor_32[i];
|
||
acc_block_pre[i] = accSensor_32[i];
|
||
}
|
||
|
||
// Adaptiven Filter auf C-Sensor Signal anwenden
|
||
|
||
//Aktuelles Sample des Acc-Sensors wird in aktuelle Speicheradresse des Pointers der Delay Line geschrieben, dann wird der Pointer inkrementiert -> Delay Line hat Länge der Filterkoeffizienten
|
||
sig_cirular_buffer_ptr_put_sample_DMB(&ptr_fir_lms_delay_line, acc_block_pre[0]);
|
||
// Filter auf Acc-Sensor Signal anwenden und Korrektursignal berechnen
|
||
// Sample des Acc-Sensors in der Delay-Line werden mit den Filterkoeffizienten multipliziert und aufsummiert -> Akkumulator Output des adaptiven Filters
|
||
acc_block_filt[0]= sig_calc_fir_lpdsp32_single(&ptr_fir_lms_delay_line, &ptr_fir_lms_coeffs);
|
||
// Output-Signal berechnen -> C-Sensor Sample - Akkumulator Output des adaptiven Filters
|
||
out_32[0] = c_block_pre[0] - acc_block_filt[0];
|
||
// Filterkoeffizienten adaptieren
|
||
adapt_coeffs_lpdsp32_single_v1(&ptr_fir_lms_delay_line, &ptr_fir_lms_coeffs, out_32[0]);
|
||
|
||
// Bitshift zurück auf 16-Bit und in Ausgangsarray schreiben
|
||
for (uint32_t i=0; i<BLOCK_LEN; i++) chess_flatten_loop
|
||
{
|
||
out_16[i] = rnd_saturate(to_accum(out_32[i]) >> BITSHIFT_16_TO_32); // 12 cycles for blocksize 4 //TODO: use rnd_saturate(out_32[i] >> input_nbit_bitshift)
|
||
}
|
||
|
||
}
|
||
|