mirror
/
M17_Implementations
kopia lustrzana https://github.com/M17-Project/M17_Implementations
1
0
Forkuj 0
Kod Zgłoszenia Packages Projekty Wydania Wiki Aktywność

WIP librarify 

doesn't build yet
librarify
Wojciech Kaczmarski 2023-12-29 16:14:41 +01:00
rodzic 7a43152c88
commit cc72c63cce
19 zmienionych plików z 1403 dodań i 1065 usunięć
Pokaż statystyki Ściągnij plik aktualizacji Ściągnij plik porównania Expand all files Collapse all files

1
.gitignore vendored
Wyświetl plik

@ -3,3 +3,4 @@
/SP5WWP/m17-decoder/m17-decoder-sym
/SP5WWP/m17-packet/m17-packet-encode
/SP5WWP/m17-packet/m17-packet-decode
/SP5WWP/.vscode

113
SP5WWP/lib/m17call.c 100644
Wyświetl plik

@ -0,0 +1,113 @@
//--------------------------------------------------------------------
// M17 C library - m17call.c
//
// This file contains:
// - callsign encoder and decoders
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <stdio.h>
#include <string.h>
#include "m17call.h"
//decodes a 6-byte long array to a callsign
void decode_callsign_bytes(uint8_t *outp, const uint8_t *inp)
{
uint64_t encoded=0;
//repack the data to a uint64_t
for(uint8_t i=0; i<6; i++)
encoded|=(uint64_t)inp[5-i]<<(8*i);
//check if the value is reserved (not a callsign)
if(encoded>=262144000000000ULL)
{
if(encoded==0xFFFFFFFFFFFF) //broadcast
{
sprintf((char*)outp, "#BCAST");
}
else
{
outp[0]=0;
}
return;
}
//decode the callsign
uint8_t i=0;
while(encoded>0)
{
outp[i]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/."[encoded%40];
encoded/=40;
i++;
}
outp[i]=0;
}
//decodes a 48-bit value to a callsign
void decode_callsign_value(uint8_t *outp, const uint64_t inp)
{
uint64_t encoded=inp;
//check if the value is reserved (not a callsign)
if(encoded>=262144000000000ULL)
{
if(encoded==0xFFFFFFFFFFFF) //broadcast
{
sprintf((char*)outp, "#BCAST");
}
else
{
outp[0]=0;
}
return;
}
//decode the callsign
uint8_t i=0;
while(encoded>0)
{
outp[i]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/."[encoded%40];
encoded/=40;
i++;
}
outp[i]=0;
}
//encode callsign into a 48-bit value
uint8_t encode_callsign(uint64_t* out, const uint8_t* inp)
{
//assert inp length
if(strlen((const char*)inp)>9)
{
return -1;
}
const uint8_t charMap[40]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/.";
uint64_t tmp=0;
if(strcmp((const char*)inp, "ALL")==0)
{
*out=0xFFFFFFFFFFFF;
return 0;
}
for(int8_t i=strlen((const char*)inp)-1; i>=0; i--)
{
for(uint8_t j=0; j<40; j++)
{
if(inp[i]==charMap[j])
{
tmp=tmp*40+j;
break;
}
}
}
*out=tmp;
return 0;
}

26
SP5WWP/lib/m17call.h 100644
Wyświetl plik

@ -0,0 +1,26 @@
//--------------------------------------------------------------------
// M17 C library - m17call.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_CALL_LIB
#define M17_CALL_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
//vars
;
//functions
void decode_callsign_bytes(uint8_t *outp, const uint8_t *inp);
void decode_callsign_value(uint8_t *outp, const uint64_t inp);
uint8_t encode_callsign(uint64_t* out, const uint8_t* inp);
#ifdef __cplusplus
}
#endif
#endif

135
SP5WWP/lib/m17consts.c 100644
Wyświetl plik

@ -0,0 +1,135 @@
//--------------------------------------------------------------------
// M17 C library - m17consts.c
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include "m17consts.h"
//syncword patterns (RX) TODO:Compute those at runtime from the consts below
const int8_t lsf_sync[8]={+3, +3, +3, +3, -3, -3, +3, -3};
const int8_t str_sync[8]={-3, -3, -3, -3, +3, +3, -3, +3};
const int8_t pkt_sync[8]={+3, -3, +3, +3, -3, -3, -3, -3};
//symbol levels (RX)
const float symbs[4]={-3.0, -1.0, +1.0, +3.0};
//dibits-symbols map (TX)
const int8_t symbol_map[4]={+1, +3, -1, -3};
//syncwords
const uint16_t SYNC_LSF = 0x55F7;
const uint16_t SYNC_STR = 0xFF5D;
const uint16_t SYNC_PKT = 0x75FF;
const uint16_t SYNC_BER = 0xDF55;
const uint16_t EOT_MRKR = 0x555D;
//puncturing pattern P_1
const uint8_t P_1[61]={1,1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1};
//puncturing pattern P_2
const uint8_t P_2[12]={1,1,1,1,1,1,1,1,1,1,1,0};
//puncturing pattern P_3
const uint8_t P_3[8]={1,1,1,1,1,1,1,0};
//M17 CRC polynomial
const uint16_t M17_CRC_POLY = 0x5935;
//sample RRC filter for 48kHz sample rate
//alpha=0.5, span=8, sps=10, gain=sqrt(sps)
const float taps_10[8*10+1]=
{
-0.003195702904062073f, -0.002930279157647190f, -0.001940667871554463f,
-0.000356087678023658f, 0.001547011339077758f, 0.003389554791179751f,
0.004761898604225673f, 0.005310860846138910f, 0.004824746306020221f,
0.003297923526848786f, 0.000958710871218619f, -0.001749908029791816f,
-0.004238694106631223f, -0.005881783042101693f, -0.006150256456781309f,
-0.004745376707651645f, -0.001704189656473565f, 0.002547854551539951f,
0.007215575568844704f, 0.011231038205363532f, 0.013421952197060707f,
0.012730475385624438f, 0.008449554307303753f, 0.000436744366018287f,
-0.010735380379191660f, -0.023726883538258272f, -0.036498030780605324f,
-0.046500883189991064f, -0.050979050575999614f, -0.047340680079891187f,
-0.033554880492651755f, -0.008513823955725943f, 0.027696543159614194f,
0.073664520037517042f, 0.126689053778116234f, 0.182990955139333916f,
0.238080025892859704f, 0.287235637987091563f, 0.326040247765297220f,
0.350895727088112619f, 0.359452932027607974f, 0.350895727088112619f,
0.326040247765297220f, 0.287235637987091563f, 0.238080025892859704f,
0.182990955139333916f, 0.126689053778116234f, 0.073664520037517042f,
0.027696543159614194f, -0.008513823955725943f, -0.033554880492651755f,
-0.047340680079891187f, -0.050979050575999614f, -0.046500883189991064f,
-0.036498030780605324f, -0.023726883538258272f, -0.010735380379191660f,
0.000436744366018287f, 0.008449554307303753f, 0.012730475385624438f,
0.013421952197060707f, 0.011231038205363532f, 0.007215575568844704f,
0.002547854551539951f, -0.001704189656473565f, -0.004745376707651645f,
-0.006150256456781309f, -0.005881783042101693f, -0.004238694106631223f,
-0.001749908029791816f, 0.000958710871218619f, 0.003297923526848786f,
0.004824746306020221f, 0.005310860846138910f, 0.004761898604225673f,
0.003389554791179751f, 0.001547011339077758f, -0.000356087678023658f,
-0.001940667871554463f, -0.002930279157647190f, -0.003195702904062073f
};
//sample RRC filter for 24kHz sample rate
//alpha=0.5, span=8, sps=5, gain=sqrt(sps)
const float taps_5[8*5+1]=
{
-0.004519384154389f, -0.002744505321971f,
0.002187793653660f, 0.006734308458208f,
0.006823188093192f, 0.001355815246317f,
-0.005994389201970f, -0.008697733303330f,
-0.002410076268276f, 0.010204314627992f,
0.018981413448435f, 0.011949415510291f,
-0.015182045838927f, -0.051615756197679f,
-0.072094910038768f, -0.047453533621088f,
0.039168634270669f, 0.179164496628150f,
0.336694345124862f, 0.461088271869920f,
0.508340710642860f, 0.461088271869920f,
0.336694345124862f, 0.179164496628150f,
0.039168634270669f, -0.047453533621088f,
-0.072094910038768f, -0.051615756197679f,
-0.015182045838927f, 0.011949415510291f,
0.018981413448435f, 0.010204314627992f,
-0.002410076268276f, -0.008697733303330f,
-0.005994389201970f, 0.001355815246317f,
0.006823188093192f, 0.006734308458208f,
0.002187793653660f, -0.002744505321971f,
-0.004519384154389f
};
//randomizing pattern
const uint8_t rand_seq[46]=
{
0xD6, 0xB5, 0xE2, 0x30, 0x82, 0xFF, 0x84, 0x62, 0xBA, 0x4E, 0x96, 0x90, 0xD8, 0x98, 0xDD, 0x5D, 0x0C, 0xC8, 0x52, 0x43, 0x91, 0x1D, 0xF8,
0x6E, 0x68, 0x2F, 0x35, 0xDA, 0x14, 0xEA, 0xCD, 0x76, 0x19, 0x8D, 0xD5, 0x80, 0xD1, 0x33, 0x87, 0x13, 0x57, 0x18, 0x2D, 0x29, 0x78, 0xC3
};
//interleaver pattern
const uint16_t intrl_seq[368]=
{
0, 137, 90, 227, 180, 317, 270, 39, 360, 129, 82, 219, 172, 309, 262, 31,
352, 121, 74, 211, 164, 301, 254, 23, 344, 113, 66, 203, 156, 293, 246, 15,
336, 105, 58, 195, 148, 285, 238, 7, 328, 97, 50, 187, 140, 277, 230, 367,
320, 89, 42, 179, 132, 269, 222, 359, 312, 81, 34, 171, 124, 261, 214, 351,
304, 73, 26, 163, 116, 253, 206, 343, 296, 65, 18, 155, 108, 245, 198, 335,
288, 57, 10, 147, 100, 237, 190, 327, 280, 49, 2, 139, 92, 229, 182, 319,
272, 41, 362, 131, 84, 221, 174, 311, 264, 33, 354, 123, 76, 213, 166, 303,
256, 25, 346, 115, 68, 205, 158, 295, 248, 17, 338, 107, 60, 197, 150, 287,
240, 9, 330, 99, 52, 189, 142, 279, 232, 1, 322, 91, 44, 181, 134, 271,
224, 361, 314, 83, 36, 173, 126, 263, 216, 353, 306, 75, 28, 165, 118, 255,
208, 345, 298, 67, 20, 157, 110, 247, 200, 337, 290, 59, 12, 149, 102, 239,
192, 329, 282, 51, 4, 141, 94, 231, 184, 321, 274, 43, 364, 133, 86, 223,
176, 313, 266, 35, 356, 125, 78, 215, 168, 305, 258, 27, 348, 117, 70, 207,
160, 297, 250, 19, 340, 109, 62, 199, 152, 289, 242, 11, 332, 101, 54, 191,
144, 281, 234, 3, 324, 93, 46, 183, 136, 273, 226, 363, 316, 85, 38, 175,
128, 265, 218, 355, 308, 77, 30, 167, 120, 257, 210, 347, 300, 69, 22, 159,
112, 249, 202, 339, 292, 61, 14, 151, 104, 241, 194, 331, 284, 53, 6, 143,
96, 233, 186, 323, 276, 45, 366, 135, 88, 225, 178, 315, 268, 37, 358, 127,
80, 217, 170, 307, 260, 29, 350, 119, 72, 209, 162, 299, 252, 21, 342, 111,
64, 201, 154, 291, 244, 13, 334, 103, 56, 193, 146, 283, 236, 5, 326, 95,
48, 185, 138, 275, 228, 365, 318, 87, 40, 177, 130, 267, 220, 357, 310, 79,
32, 169, 122, 259, 212, 349, 302, 71, 24, 161, 114, 251, 204, 341, 294, 63,
16, 153, 106, 243, 196, 333, 286, 55, 8, 145, 98, 235, 188, 325, 278, 47
};

90
SP5WWP/lib/m17consts.h 100644
Wyświetl plik

@ -0,0 +1,90 @@
//--------------------------------------------------------------------
// M17 C library - m17consts.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_CONSTS
#define M17_CONSTS
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
#define BSB_SPS 10 //samples per symbol
#define FLT_SPAN 8 //baseband RRC filter span in symbols
#define SYM_PER_SWD 8 //symbols per syncword
#define SW_LEN (BSB_SPS*SYM_PER_SWD) //syncword detector length
#define SYM_PER_PLD 184 //symbols per payload in a frame
#define SYM_PER_FRA 192 //symbols per whole 40 ms frame
#define RRC_DEV 7168.0f //.rrc file deviation for +1.0 symbol
//L2 metric threshold
#define DIST_THRESH 2.0f //threshold for distance (syncword detection)
//Viterbi
#define K 5 //constraint length
#define NUM_STATES (1 << (K - 1)) //number of states
//syncword patterns (RX) TODO:Compute those at runtime from the consts below
extern const int8_t lsf_sync[8];
extern const int8_t str_sync[8];
extern const int8_t pkt_sync[8];
//symbol levels (RX)
extern const float symbs[4];
//dibits-symbols map (TX)
extern const int8_t symbol_map[4];
//syncwords
extern const uint16_t SYNC_LSF;
extern const uint16_t SYNC_STR;
extern const uint16_t SYNC_PKT;
extern const uint16_t SYNC_BER;
extern const uint16_t EOT_MRKR;
//puncturing pattern P_1
extern const uint8_t P_1[61];
//puncturing pattern P_2
extern const uint8_t P_2[12];
//puncturing pattern P_3
extern const uint8_t P_3[8];
//M17 CRC polynomial
extern const uint16_t M17_CRC_POLY;
//sample RRC filter for 48kHz sample rate
//alpha=0.5, span=8, sps=10, gain=sqrt(sps)
extern const float taps_10[8*10+1];
//sample RRC filter for 24kHz sample rate
//alpha=0.5, span=8, sps=5, gain=sqrt(sps)
extern const float taps_5[8*5+1];
//randomizing pattern
extern const uint8_t rand_seq[46];
//interleaver pattern
extern const uint16_t intrl_seq[368];
/**
* @brief Structure holding Link Setup Frame data.
*
*/
typedef struct LSF
{
uint8_t dst[6];
uint8_t src[6];
uint8_t type[2];
uint8_t meta[112/8];
uint8_t crc[2];
};
#ifdef __cplusplus
}
#endif
#endif

227
SP5WWP/lib/m17convol.c 100644
Wyświetl plik

@ -0,0 +1,227 @@
//--------------------------------------------------------------------
// M17 C library - m17convol.c
//
// This file contains:
// - convolutional encoders for the LSF, stream, and packet frames
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <string.h>
#include "m17convol.h"
/**
* @brief Encode M17 stream frame using convolutional encoder with puncturing.
*
* @param out Output array, unpacked.
* @param in Input - packed array of uint8_t, 144 type-1 bits.
* @param fn Input - 16-bit frame number.
*/
void conv_encode_stream_frame(uint8_t* out, const uint8_t* in, const uint16_t fn)
{
uint8_t pp_len = sizeof(P_2);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[144+4+4]; //unpacked data
memset(ud, 0, 144+4+4);
//unpack frame number
for(uint8_t i=0; i<16; i++)
{
ud[4+i]=(fn>>(15-i))&1;
}
//unpack data
for(uint8_t i=0; i<16; i++)
{
for(uint8_t j=0; j<8; j++)
{
ud[4+16+i*8+j]=(in[i]>>(7-j))&1;
}
}
//encode
for(uint8_t i=0; i<144+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//printf("%d%d", G1, G2);
if(P_2[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_2[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//printf("pb=%d\n", pb);
}
/**
* @brief Encode M17 packet frame using convolutional encoder with puncturing.
*
* @param out Output array, unpacked.
* @param in Input - packed array of uint8_t, 206 type-1 bits.
*/
void conv_encode_packet_frame(uint8_t* out, const uint8_t* in)
{
uint8_t pp_len = sizeof(P_3);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[206+4+4]; //unpacked data
memset(ud, 0, 206+4+4);
//unpack data
for(uint8_t i=0; i<26; i++)
{
for(uint8_t j=0; j<8; j++)
{
if(i<=24 || j<=5)
ud[4+i*8+j]=(in[i]>>(7-j))&1;
}
}
//encode
for(uint8_t i=0; i<206+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//fprintf(stderr, "%d%d", G1, G2);
if(P_3[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_3[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//fprintf(stderr, "pb=%d\n", pb);
}
/**
* @brief Encode M17 stream frame using convolutional encoder with puncturing.
*
* @param out Output array, unpacked.
* @param in Input - pointer to a struct holding the Link Setup Frame.
*/
void conv_encode_LSF(uint8_t* out, const struct LSF *in)
{
uint8_t pp_len = sizeof(P_1);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[240+4+4]; //unpacked data
memset(ud, 0, 240+4+4);
//unpack DST
for(uint8_t i=0; i<8; i++)
{
ud[4+i] =((in->dst[0])>>(7-i))&1;
ud[4+i+8] =((in->dst[1])>>(7-i))&1;
ud[4+i+16]=((in->dst[2])>>(7-i))&1;
ud[4+i+24]=((in->dst[3])>>(7-i))&1;
ud[4+i+32]=((in->dst[4])>>(7-i))&1;
ud[4+i+40]=((in->dst[5])>>(7-i))&1;
}
//unpack SRC
for(uint8_t i=0; i<8; i++)
{
ud[4+i+48]=((in->src[0])>>(7-i))&1;
ud[4+i+56]=((in->src[1])>>(7-i))&1;
ud[4+i+64]=((in->src[2])>>(7-i))&1;
ud[4+i+72]=((in->src[3])>>(7-i))&1;
ud[4+i+80]=((in->src[4])>>(7-i))&1;
ud[4+i+88]=((in->src[5])>>(7-i))&1;
}
//unpack TYPE
for(uint8_t i=0; i<8; i++)
{
ud[4+i+96] =((in->type[0])>>(7-i))&1;
ud[4+i+104]=((in->type[1])>>(7-i))&1;
}
//unpack META
for(uint8_t i=0; i<8; i++)
{
ud[4+i+112]=((in->meta[0])>>(7-i))&1;
ud[4+i+120]=((in->meta[1])>>(7-i))&1;
ud[4+i+128]=((in->meta[2])>>(7-i))&1;
ud[4+i+136]=((in->meta[3])>>(7-i))&1;
ud[4+i+144]=((in->meta[4])>>(7-i))&1;
ud[4+i+152]=((in->meta[5])>>(7-i))&1;
ud[4+i+160]=((in->meta[6])>>(7-i))&1;
ud[4+i+168]=((in->meta[7])>>(7-i))&1;
ud[4+i+176]=((in->meta[8])>>(7-i))&1;
ud[4+i+184]=((in->meta[9])>>(7-i))&1;
ud[4+i+192]=((in->meta[10])>>(7-i))&1;
ud[4+i+200]=((in->meta[11])>>(7-i))&1;
ud[4+i+208]=((in->meta[12])>>(7-i))&1;
ud[4+i+216]=((in->meta[13])>>(7-i))&1;
}
//unpack CRC
for(uint8_t i=0; i<8; i++)
{
ud[4+i+224]=((in->crc[0])>>(7-i))&1;
ud[4+i+232]=((in->crc[1])>>(7-i))&1;
}
//encode
for(uint8_t i=0; i<240+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//printf("%d%d", G1, G2);
if(P_1[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_1[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//printf("pb=%d\n", pb);
}

27
SP5WWP/lib/m17convol.h 100644
Wyświetl plik

@ -0,0 +1,27 @@
//--------------------------------------------------------------------
// M17 C library - m17convol.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_CONVOL_LIB
#define M17_CONVOL_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
#include "m17consts.h"
//vars
;
//functions
void conv_encode_stream_frame(uint8_t* out, const uint8_t* in, const uint16_t fn);
void conv_encode_packet_frame(uint8_t* out, const uint8_t* in);
void conv_encode_LSF(uint8_t* out, const struct LSF *in);
#ifdef __cplusplus
}
#endif
#endif

55
SP5WWP/lib/m17crc.c 100644
Wyświetl plik

@ -0,0 +1,55 @@
//--------------------------------------------------------------------
// M17 C library - m17crc.c
//
// This file contains:
// - CRC calculating functions (arbitrary length)
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <string.h>
#include "m17crc.h"
#include "m17consts.h"
/**
* @brief Calculate CRC value.
*
* @param in Pointer to the input byte array.
* @param len Input's length.
* @return uint16_t CRC value.
*/
uint16_t CRC_M17(const uint8_t *in, const uint16_t len)
{
uint32_t crc=0xFFFF; //init val
for(uint16_t i=0; i<len; i++)
{
crc^=in[i]<<8;
for(uint8_t j=0; j<8; j++)
{
crc<<=1;
if(crc&0x10000)
crc=(crc^M17_CRC_POLY)&0xFFFF;
}
}
return crc&(0xFFFF);
}
/**
* @brief Calculate CRC value for the Link Setup Frame.
*
* @param in Pointer to an LSF struct.
* @return uint16_t CRC value.
*/
uint16_t LSF_CRC(const struct LSF_t *in)
{
uint8_t d[28];
memcpy(&d[0], in->dst, 6);
memcpy(&d[6], in->src, 6);
memcpy(&d[12], in->type, 2);
memcpy(&d[14], in->meta, 14);
return CRC_M17(d, 28);
}

26
SP5WWP/lib/m17crc.h 100644
Wyświetl plik

@ -0,0 +1,26 @@
//--------------------------------------------------------------------
// M17 C library - m17crc.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_CRC_LIB
#define M17_CRC_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
//vars
;
//functions
uint16_t CRC_M17(const uint8_t *in, const uint16_t len);
uint16_t LSF_CRC(const struct LSF *in);
#ifdef __cplusplus
}
#endif
#endif

261
SP5WWP/lib/m17golay.c 100644
Wyświetl plik

@ -0,0 +1,261 @@
//--------------------------------------------------------------------
// M17 C library - m17golay.c
//
// This file contains:
// - Golay(24, 12) encoder
// - Golay(24, 12) soft decoder with accompanying utility functions
// - Link Information Channel (LICH) decoder
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <string.h>
#include "m17golay.h"
/**
* @brief Precomputed encoding matrix for Golay(24, 12).
*
*/
const uint16_t encode_matrix[12]=
{
0x8eb, 0x93e, 0xa97, 0xdc6, 0x367, 0x6cd,
0xd99, 0x3da, 0x7b4, 0xf68, 0x63b, 0xc75
};
/**
* @brief Precomputed decoding matrix for Golay(24, 12).
*
*/
const uint16_t decode_matrix[12]=
{
0xc75, 0x49f, 0x93e, 0x6e3, 0xdc6, 0xf13,
0xab9, 0x1ed, 0x3da, 0x7b4, 0xf68, 0xa4f
};
/**
* @brief Encode a 12-bit value with Golay(24, 12).
*
* @param data 12-bit input value (right justified).
* @return uint32_t 24-bit Golay codeword.
*/
uint32_t golay24_encode(const uint16_t data)
{
uint16_t checksum=0;
for(uint8_t i=0; i<12; i++)
{
if(data&(1<<i))
{
checksum ^= encode_matrix[i];
}
}
return (data<<12) | checksum;
}
/**
* @brief Soft-valued equivalent of `popcount()`
*
* @param in Pointer to an array holding soft logic vector.
* @param siz Vector's size.
* @return uint32_t Sum of all values.
*/
uint32_t s_popcount(const uint16_t* in, uint8_t siz)
{
uint32_t tmp=0;
for(uint8_t i=0; i<siz; i++)
tmp+=in[i];
return tmp;
}
/**
* @brief
*
* @param out
* @param value
*/
void s_calc_checksum(uint16_t* out, const uint16_t* value)
{
uint16_t checksum[12];
uint16_t soft_em[12]; //soft valued encoded matrix entry
for(uint8_t i=0; i<12; i++)
checksum[i]=0;
for(uint8_t i=0; i<12; i++)
{
int_to_soft(soft_em, encode_matrix[i], 12);
if(value[i]>0x7FFF)
{
soft_XOR(checksum, checksum, soft_em, 12);
}
}
memcpy((uint8_t*)out, (uint8_t*)checksum, 12*2);
}
/**
* @brief Detect errors in a soft-valued Golay(24, 12) codeword.
*
* @param codeword Input 24-bit soft codeword.
* @return uint32_t Detected errors vector.
*/
uint32_t s_detect_errors(const uint16_t* codeword)
{
uint16_t data[12];
uint16_t parity[12];
uint16_t cksum[12];
uint16_t syndrome[12];
uint32_t weight; //for soft popcount
memcpy((uint8_t*)data, (uint8_t*)&codeword[12], 2*12);
memcpy((uint8_t*)parity, (uint8_t*)&codeword[0], 2*12);
s_calc_checksum(cksum, data);
soft_XOR(syndrome, parity, cksum, 12);
weight=s_popcount(syndrome, 12);
//all (less than 4) errors in the parity part
if(weight < 4*0xFFFE)
{
//printf("1: %1.2f\n", (float)weight/0xFFFF);
return soft_to_int(syndrome, 12);
}
//one of the errors in data part, up to 3 in parity
for(uint8_t i = 0; i<12; i++)
{
uint16_t e = 1<<i;
uint16_t coded_error = encode_matrix[i];
uint16_t scoded_error[12]; //soft coded_error
uint16_t sc[12]; //syndrome^coded_error
int_to_soft(scoded_error, coded_error, 12);
soft_XOR(sc, syndrome, scoded_error, 12);
weight=s_popcount(sc, 12);
if(weight < 3*0xFFFE)
{
//printf("2: %1.2f\n", (float)weight/0xFFFF+1);
uint16_t s=soft_to_int(syndrome, 12);
return (e << 12) | (s ^ coded_error);
}
}
//two of the errors in data part and up to 2 in parity
for(uint8_t i = 0; i<11; i++)
{
for(uint8_t j = i+1; j<12; j++)
{
uint16_t e = (1<<i) | (1<<j);
uint16_t coded_error = encode_matrix[i]^encode_matrix[j];
uint16_t scoded_error[12]; //soft coded_error
uint16_t sc[12]; //syndrome^coded_error
int_to_soft(scoded_error, coded_error, 12);
soft_XOR(sc, syndrome, scoded_error, 12);
weight=s_popcount(sc, 12);
if(weight < 2*0xFFFF)
{
//printf("3: %1.2f\n", (float)weight/0xFFFF+2);
uint16_t s=soft_to_int(syndrome, 12);
return (e << 12) | (s ^ coded_error);
}
}
}
//algebraic decoding magic
uint16_t inv_syndrome[12]={0,0,0,0,0,0,0,0,0,0,0,0};
uint16_t dm[12]; //soft decode matrix
for(uint8_t i=0; i<12; i++)
{
if(syndrome[i] > 0x7FFF)
{
int_to_soft(dm, decode_matrix[i], 12);
soft_XOR(inv_syndrome, inv_syndrome, dm, 12);
}
}
//all (less than 4) errors in the data part
weight=s_popcount(inv_syndrome, 12);
if(weight < 4*0xFFFF)
{
//printf("4: %1.2f\n", (float)weight/0xFFFF);
return soft_to_int(inv_syndrome, 12) << 12;
}
//one error in parity bits, up to 3 in data - this part has some quirks, the reason remains unknown
for(uint8_t i=0; i<12; i++)
{
uint16_t e = 1<<i;
uint16_t coding_error = decode_matrix[i];
uint16_t ce[12]; //soft coding error
uint16_t tmp[12];
int_to_soft(ce, coding_error, 12);
soft_XOR(tmp, inv_syndrome, ce, 12);
weight=s_popcount(tmp, 12);
if(weight < 3*(0xFFFF+2))
{
//printf("5: %1.2f\n", (float)weight/0xFFFF+1);
return ((soft_to_int(inv_syndrome, 12) ^ coding_error) << 12) | e;
}
}
return 0xFFFFFFFFUL;
}
/**
* @brief Soft decode Golay(24, 12) codeword.
*
* @param codeword Pointer to a 24-element soft-valued (fixed-point) bit codeword.
* @return uint16_t Decoded data.
*/
uint16_t golay24_sdecode(const uint16_t codeword[24])
{
//match the bit order in M17
uint16_t cw[24]; //local copy
for(uint8_t i=0; i<24; i++)
cw[i]=codeword[23-i];
uint32_t errors = s_detect_errors(cw);
if(errors == 0xFFFFFFFF)
return 0xFFFF;
return (((soft_to_int(&cw[0], 16) | (soft_to_int(&cw[16], 8) << 16)) ^ errors) >> 12) & 0x0FFF;
}
/**
* @brief Soft decode LICH into a 6-byte array.
*
* @param outp An array of packed, decoded bits.
* @param inp Pointer to an array of 96 soft bits.
*/
void decode_LICH(uint8_t* outp, const uint16_t* inp)
{
uint16_t tmp;
memset(outp, 0, 5);
tmp=golay24_sdecode(&inp[0]);
outp[0]=(tmp>>4)&0xFF;
outp[1]|=(tmp&0xF)<<4;
tmp=golay24_sdecode(&inp[1*24]);
outp[1]|=(tmp>>8)&0xF;
outp[2]=tmp&0xFF;
tmp=golay24_sdecode(&inp[2*24]);
outp[3]=(tmp>>4)&0xFF;
outp[4]|=(tmp&0xF)<<4;
tmp=golay24_sdecode(&inp[3*24]);
outp[4]|=(tmp>>8)&0xF;
outp[5]=tmp&0xFF;
}

27
SP5WWP/lib/m17golay.h 100644
Wyświetl plik

@ -0,0 +1,27 @@
//--------------------------------------------------------------------
// M17 C library - m17golay.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_GOLAY_LIB
#define M17_GOLAY_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
//consts
extern const uint16_t encode_matrix[12];
extern const uint16_t decode_matrix[12];
//functions
uint32_t golay24_encode(const uint16_t data);
uint16_t golay24_sdecode(const uint16_t codeword[24]);
void decode_LICH(uint8_t* outp, const uint16_t* inp);
#ifdef __cplusplus
}
#endif
#endif

959
SP5WWP/lib/m17lib.c
Wyświetl plik

@ -1,610 +1,14 @@
//-------------------------------
//--------------------------------------------------------------------
// M17 C library - m17lib.c
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 28 December 2023
//-------------------------------
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <stdio.h>
#include <string.h>
#include <stdint.h>
#include <math.h>
#include "m17consts.h"
#include "m17lib.h"
//syncword patterns (RX) TODO:Compute those at runtime from the consts below
const int8_t lsf_sync[8]={+3, +3, +3, +3, -3, -3, +3, -3};
const int8_t str_sync[8]={-3, -3, -3, -3, +3, +3, -3, +3};
const int8_t pkt_sync[8]={+3, -3, +3, +3, -3, -3, -3, -3};
//symbol levels (RX)
const float symbs[4]={-3.0, -1.0, +1.0, +3.0};
//dibits-symbols map (TX)
const int8_t symbol_map[4]={+1, +3, -1, -3};
//syncwords
const uint16_t SYNC_LSF = 0x55F7;
const uint16_t SYNC_STR = 0xFF5D;
const uint16_t SYNC_PKT = 0x75FF;
const uint16_t SYNC_BER = 0xDF55;
const uint16_t EOT_MRKR = 0x555D;
//puncturing pattern P_1
const uint8_t P_1[61]={1,1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1,1,0,1,1,
1,0,1,1,1,0,1,1,1,0,1,1};
//puncturing pattern P_2
const uint8_t P_2[12]={1,1,1,1,1,1,1,1,1,1,1,0};
//puncturing pattern P_3
const uint8_t P_3[8]={1,1,1,1,1,1,1,0};
//M17 CRC polynomial
const uint16_t M17_CRC_POLY = 0x5935;
//sample RRC filter for 48kHz sample rate
//alpha=0.5, span=8, sps=10, gain=sqrt(sps)
const float taps_10[8*10+1]=
{
-0.003195702904062073f, -0.002930279157647190f, -0.001940667871554463f,
-0.000356087678023658f, 0.001547011339077758f, 0.003389554791179751f,
0.004761898604225673f, 0.005310860846138910f, 0.004824746306020221f,
0.003297923526848786f, 0.000958710871218619f, -0.001749908029791816f,
-0.004238694106631223f, -0.005881783042101693f, -0.006150256456781309f,
-0.004745376707651645f, -0.001704189656473565f, 0.002547854551539951f,
0.007215575568844704f, 0.011231038205363532f, 0.013421952197060707f,
0.012730475385624438f, 0.008449554307303753f, 0.000436744366018287f,
-0.010735380379191660f, -0.023726883538258272f, -0.036498030780605324f,
-0.046500883189991064f, -0.050979050575999614f, -0.047340680079891187f,
-0.033554880492651755f, -0.008513823955725943f, 0.027696543159614194f,
0.073664520037517042f, 0.126689053778116234f, 0.182990955139333916f,
0.238080025892859704f, 0.287235637987091563f, 0.326040247765297220f,
0.350895727088112619f, 0.359452932027607974f, 0.350895727088112619f,
0.326040247765297220f, 0.287235637987091563f, 0.238080025892859704f,
0.182990955139333916f, 0.126689053778116234f, 0.073664520037517042f,
0.027696543159614194f, -0.008513823955725943f, -0.033554880492651755f,
-0.047340680079891187f, -0.050979050575999614f, -0.046500883189991064f,
-0.036498030780605324f, -0.023726883538258272f, -0.010735380379191660f,
0.000436744366018287f, 0.008449554307303753f, 0.012730475385624438f,
0.013421952197060707f, 0.011231038205363532f, 0.007215575568844704f,
0.002547854551539951f, -0.001704189656473565f, -0.004745376707651645f,
-0.006150256456781309f, -0.005881783042101693f, -0.004238694106631223f,
-0.001749908029791816f, 0.000958710871218619f, 0.003297923526848786f,
0.004824746306020221f, 0.005310860846138910f, 0.004761898604225673f,
0.003389554791179751f, 0.001547011339077758f, -0.000356087678023658f,
-0.001940667871554463f, -0.002930279157647190f, -0.003195702904062073f
};
//sample RRC filter for 24kHz sample rate
//alpha=0.5, span=8, sps=5, gain=sqrt(sps)
const float taps_5[8*5+1]=
{
-0.004519384154389f, -0.002744505321971f,
0.002187793653660f, 0.006734308458208f,
0.006823188093192f, 0.001355815246317f,
-0.005994389201970f, -0.008697733303330f,
-0.002410076268276f, 0.010204314627992f,
0.018981413448435f, 0.011949415510291f,
-0.015182045838927f, -0.051615756197679f,
-0.072094910038768f, -0.047453533621088f,
0.039168634270669f, 0.179164496628150f,
0.336694345124862f, 0.461088271869920f,
0.508340710642860f, 0.461088271869920f,
0.336694345124862f, 0.179164496628150f,
0.039168634270669f, -0.047453533621088f,
-0.072094910038768f, -0.051615756197679f,
-0.015182045838927f, 0.011949415510291f,
0.018981413448435f, 0.010204314627992f,
-0.002410076268276f, -0.008697733303330f,
-0.005994389201970f, 0.001355815246317f,
0.006823188093192f, 0.006734308458208f,
0.002187793653660f, -0.002744505321971f,
-0.004519384154389f
};
//Golay coding
const uint16_t encode_matrix[12]=
{
0x8eb, 0x93e, 0xa97, 0xdc6, 0x367, 0x6cd,
0xd99, 0x3da, 0x7b4, 0xf68, 0x63b, 0xc75
};
const uint16_t decode_matrix[12]=
{
0xc75, 0x49f, 0x93e, 0x6e3, 0xdc6, 0xf13,
0xab9, 0x1ed, 0x3da, 0x7b4, 0xf68, 0xa4f
};
//randomizing pattern
const uint8_t rand_seq[46]=
{
0xD6, 0xB5, 0xE2, 0x30, 0x82, 0xFF, 0x84, 0x62, 0xBA, 0x4E, 0x96, 0x90, 0xD8, 0x98, 0xDD, 0x5D, 0x0C, 0xC8, 0x52, 0x43, 0x91, 0x1D, 0xF8,
0x6E, 0x68, 0x2F, 0x35, 0xDA, 0x14, 0xEA, 0xCD, 0x76, 0x19, 0x8D, 0xD5, 0x80, 0xD1, 0x33, 0x87, 0x13, 0x57, 0x18, 0x2D, 0x29, 0x78, 0xC3
};
//interleaver pattern
const uint16_t intrl_seq[368]=
{
0, 137, 90, 227, 180, 317, 270, 39, 360, 129, 82, 219, 172, 309, 262, 31,
352, 121, 74, 211, 164, 301, 254, 23, 344, 113, 66, 203, 156, 293, 246, 15,
336, 105, 58, 195, 148, 285, 238, 7, 328, 97, 50, 187, 140, 277, 230, 367,
320, 89, 42, 179, 132, 269, 222, 359, 312, 81, 34, 171, 124, 261, 214, 351,
304, 73, 26, 163, 116, 253, 206, 343, 296, 65, 18, 155, 108, 245, 198, 335,
288, 57, 10, 147, 100, 237, 190, 327, 280, 49, 2, 139, 92, 229, 182, 319,
272, 41, 362, 131, 84, 221, 174, 311, 264, 33, 354, 123, 76, 213, 166, 303,
256, 25, 346, 115, 68, 205, 158, 295, 248, 17, 338, 107, 60, 197, 150, 287,
240, 9, 330, 99, 52, 189, 142, 279, 232, 1, 322, 91, 44, 181, 134, 271,
224, 361, 314, 83, 36, 173, 126, 263, 216, 353, 306, 75, 28, 165, 118, 255,
208, 345, 298, 67, 20, 157, 110, 247, 200, 337, 290, 59, 12, 149, 102, 239,
192, 329, 282, 51, 4, 141, 94, 231, 184, 321, 274, 43, 364, 133, 86, 223,
176, 313, 266, 35, 356, 125, 78, 215, 168, 305, 258, 27, 348, 117, 70, 207,
160, 297, 250, 19, 340, 109, 62, 199, 152, 289, 242, 11, 332, 101, 54, 191,
144, 281, 234, 3, 324, 93, 46, 183, 136, 273, 226, 363, 316, 85, 38, 175,
128, 265, 218, 355, 308, 77, 30, 167, 120, 257, 210, 347, 300, 69, 22, 159,
112, 249, 202, 339, 292, 61, 14, 151, 104, 241, 194, 331, 284, 53, 6, 143,
96, 233, 186, 323, 276, 45, 366, 135, 88, 225, 178, 315, 268, 37, 358, 127,
80, 217, 170, 307, 260, 29, 350, 119, 72, 209, 162, 299, 252, 21, 342, 111,
64, 201, 154, 291, 244, 13, 334, 103, 56, 193, 146, 283, 236, 5, 326, 95,
48, 185, 138, 275, 228, 365, 318, 87, 40, 177, 130, 267, 220, 357, 310, 79,
32, 169, 122, 259, 212, 349, 302, 71, 24, 161, 114, 251, 204, 341, 294, 63,
16, 153, 106, 243, 196, 333, 286, 55, 8, 145, 98, 235, 188, 325, 278, 47
};
//Viterbi vars
uint32_t prevMetrics[NUM_STATES];
uint32_t currMetrics[NUM_STATES];
uint32_t prevMetricsData[NUM_STATES];
uint32_t currMetricsData[NUM_STATES];
uint16_t viterbi_history[244];
//CRC
uint16_t CRC_M17(const uint8_t *in, const uint16_t len)
{
uint32_t crc=0xFFFF; //init val
for(uint16_t i=0; i<len; i++)
{
crc^=in[i]<<8;
for(uint8_t j=0; j<8; j++)
{
crc<<=1;
if(crc&0x10000)
crc=(crc^M17_CRC_POLY)&0xFFFF;
}
}
return crc&(0xFFFF);
}
//Golay coding
uint32_t golay24_encode(const uint16_t data)
{
uint16_t checksum=0;
for(uint8_t i=0; i<12; i++)
{
if(data&(1<<i))
{
checksum ^= encode_matrix[i];
}
}
return (data<<12) | checksum;
}
//0 index - LSB
void int_to_soft(uint16_t* out, const uint16_t in, uint8_t len)
{
for(uint8_t i=0; i<len; i++)
{
(in>>i)&1 ? (out[i]=0xFFFF) : (out[i]=0);
}
}
uint16_t soft_to_int(const uint16_t* in, uint8_t len)
{
uint16_t tmp=0;
for(uint8_t i=0; i<len; i++)
{
if(in[i]>0x7FFF)
tmp|=(1<<i);
}
return tmp;
}
//Quadrant I fixed point division with saturation
//result=a/b
uint16_t div16(uint16_t a, uint16_t b)
{
uint32_t aa=(uint32_t)a<<16;
uint32_t r=aa/b;
if(r<=0xFFFF)
return r;
else
return 0xFFFF;
}
//Quadrant I fixed point multiplication
//result=a/b
uint16_t mul16(uint16_t a, uint16_t b)
{
return (uint16_t)(((uint32_t)a*b)>>16);
}
//use bilinear interpolation for XOR
uint16_t soft_bit_XOR(const uint16_t a, const uint16_t b)
{
return mul16(div16(0xFFFF-b, 0xFFFF), div16(a, 0xFFFF)) + mul16(div16(b, 0xFFFF), div16(0xFFFF-a, 0xFFFF));
}
//soft XOR
void soft_XOR(uint16_t* out, const uint16_t* a, const uint16_t* b, uint8_t len)
{
for(uint8_t i=0; i<len; i++)
out[i]=soft_bit_XOR(a[i], b[i]);
}
//soft equivalent of popcount
uint32_t s_popcount(const uint16_t* in, uint8_t siz)
{
uint32_t tmp=0;
for(uint8_t i=0; i<siz; i++)
tmp+=in[i];
return tmp;
}
void s_calc_checksum(uint16_t* out, const uint16_t* value)
{
uint16_t checksum[12];
uint16_t soft_em[12]; //soft valued encoded matrix entry
for(uint8_t i=0; i<12; i++)
checksum[i]=0;
for(uint8_t i=0; i<12; i++)
{
int_to_soft(soft_em, encode_matrix[i], 12);
if(value[i]>0x7FFF)
{
soft_XOR(checksum, checksum, soft_em, 12);
}
}
memcpy((uint8_t*)out, (uint8_t*)checksum, 12*2);
}
uint32_t s_detect_errors(const uint16_t* codeword)
{
uint16_t data[12];
uint16_t parity[12];
uint16_t cksum[12];
uint16_t syndrome[12];
uint32_t weight; //for soft popcount
memcpy((uint8_t*)data, (uint8_t*)&codeword[12], 2*12);
memcpy((uint8_t*)parity, (uint8_t*)&codeword[0], 2*12);
s_calc_checksum(cksum, data);
soft_XOR(syndrome, parity, cksum, 12);
weight=s_popcount(syndrome, 12);
//all (less than 4) errors in the parity part
if(weight < 4*0xFFFE)
{
//printf("1: %1.2f\n", (float)weight/0xFFFF);
return soft_to_int(syndrome, 12);
}
//one of the errors in data part, up to 3 in parity
for(uint8_t i = 0; i<12; i++)
{
uint16_t e = 1<<i;
uint16_t coded_error = encode_matrix[i];
uint16_t scoded_error[12]; //soft coded_error
uint16_t sc[12]; //syndrome^coded_error
int_to_soft(scoded_error, coded_error, 12);
soft_XOR(sc, syndrome, scoded_error, 12);
weight=s_popcount(sc, 12);
if(weight < 3*0xFFFE)
{
//printf("2: %1.2f\n", (float)weight/0xFFFF+1);
uint16_t s=soft_to_int(syndrome, 12);
return (e << 12) | (s ^ coded_error);
}
}
//two of the errors in data part and up to 2 in parity
for(uint8_t i = 0; i<11; i++)
{
for(uint8_t j = i+1; j<12; j++)
{
uint16_t e = (1<<i) | (1<<j);
uint16_t coded_error = encode_matrix[i]^encode_matrix[j];
uint16_t scoded_error[12]; //soft coded_error
uint16_t sc[12]; //syndrome^coded_error
int_to_soft(scoded_error, coded_error, 12);
soft_XOR(sc, syndrome, scoded_error, 12);
weight=s_popcount(sc, 12);
if(weight < 2*0xFFFF)
{
//printf("3: %1.2f\n", (float)weight/0xFFFF+2);
uint16_t s=soft_to_int(syndrome, 12);
return (e << 12) | (s ^ coded_error);
}
}
}
//algebraic decoding magic
uint16_t inv_syndrome[12]={0,0,0,0,0,0,0,0,0,0,0,0};
uint16_t dm[12]; //soft decode matrix
for(uint8_t i=0; i<12; i++)
{
if(syndrome[i] > 0x7FFF)
{
int_to_soft(dm, decode_matrix[i], 12);
soft_XOR(inv_syndrome, inv_syndrome, dm, 12);
}
}
//all (less than 4) errors in the data part
weight=s_popcount(inv_syndrome, 12);
if(weight < 4*0xFFFF)
{
//printf("4: %1.2f\n", (float)weight/0xFFFF);
return soft_to_int(inv_syndrome, 12) << 12;
}
//one error in parity bits, up to 3 in data - this part has some quirks, the reason remains unknown
for(uint8_t i=0; i<12; i++)
{
uint16_t e = 1<<i;
uint16_t coding_error = decode_matrix[i];
uint16_t ce[12]; //soft coding error
uint16_t tmp[12];
int_to_soft(ce, coding_error, 12);
soft_XOR(tmp, inv_syndrome, ce, 12);
weight=s_popcount(tmp, 12);
if(weight < 3*(0xFFFF+2))
{
//printf("5: %1.2f\n", (float)weight/0xFFFF+1);
return ((soft_to_int(inv_syndrome, 12) ^ coding_error) << 12) | e;
}
}
return 0xFFFFFFFF;
}
//soft decode
uint16_t golay24_sdecode(const uint16_t* codeword)
{
//match the bit order in M17
uint16_t cw[24];
for(uint8_t i=0; i<24; i++)
cw[i]=codeword[23-i];
uint32_t errors = s_detect_errors(cw);
if(errors == 0xFFFFFFFF)
return 0xFFFF;
return (((soft_to_int(&cw[0], 16) | (soft_to_int(&cw[16], 8) << 16)) ^ errors) >> 12) & 0x0FFF;
}
//Viterbi
/**
* Decode unpunctured convolutionally encoded data.
*
* @param out: destination array where decoded data is written.
* @param in: input data.
* @param len: input length in bits
* @return number of bit errors corrected.
*/
uint32_t viterbi_decode(uint8_t* out, const uint16_t* in, uint16_t len)
{
if(len > 244*2)
fprintf(stderr, "Input size exceeds max history\n");
viterbi_reset();
size_t pos = 0;
for(size_t i = 0; i < len; i += 2)
{
uint16_t s0 = in[i];
uint16_t s1 = in[i + 1];
viterbi_decode_bit(s0, s1, pos);
pos++;
}
return viterbi_chainback(out, pos, len/2);
}
/**
* Decode punctured convolutionally encoded data.
*
* @param out: destination array where decoded data is written.
* @param in: input data.
* @param punct: puncturing matrix.
* @param in_len: input data length.
* @param p_len: puncturing matrix length (entries).
* @return number of bit errors corrected.
*/
uint32_t viterbi_decode_punctured(uint8_t* out, const uint16_t* in, const uint8_t* punct, const uint16_t in_len, const uint16_t p_len)
{
if(in_len > 244*2)
fprintf(stderr, "Input size exceeds max history\n");
uint16_t umsg[244*2]; //unpunctured message
uint8_t p=0; //puncturer matrix entry
uint16_t u=0; //bits count - unpunctured message
uint16_t i=0; //bits read from the input message
while(i<in_len)
{
if(punct[p])
{
umsg[u]=in[i];
i++;
}
else
{
umsg[u]=0x7FFF;
}
u++;
p++;
p%=p_len;
}
return viterbi_decode(out, umsg, u) - (u-in_len)*0x7FFF;
}
/**
* Decode one bit and update trellis.
*
* @param s0: cost of the first symbol.
* @param s1: cost of the second symbol.
* @param pos: bit position in history.
*/
void viterbi_decode_bit(uint16_t s0, uint16_t s1, size_t pos)
{
static const uint16_t COST_TABLE_0[] = {0, 0, 0, 0, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF};
static const uint16_t COST_TABLE_1[] = {0, 0xFFFF, 0xFFFF, 0, 0, 0xFFFF, 0xFFFF, 0};
for(uint8_t i = 0; i < NUM_STATES/2; i++)
{
uint32_t metric = q_abs_diff(COST_TABLE_0[i], s0)
+ q_abs_diff(COST_TABLE_1[i], s1);
uint32_t m0 = prevMetrics[i] + metric;
uint32_t m1 = prevMetrics[i + NUM_STATES/2] + (0x1FFFE - metric);
uint32_t m2 = prevMetrics[i] + (0x1FFFE - metric);
uint32_t m3 = prevMetrics[i + NUM_STATES/2] + metric;
uint8_t i0 = 2 * i;
uint8_t i1 = i0 + 1;
if(m0 >= m1)
{
viterbi_history[pos]|=(1<<i0);
currMetrics[i0] = m1;
}
else
{
viterbi_history[pos]&=~(1<<i0);
currMetrics[i0] = m0;
}
if(m2 >= m3)
{
viterbi_history[pos]|=(1<<i1);
currMetrics[i1] = m3;
}
else
{
viterbi_history[pos]&=~(1<<i1);
currMetrics[i1] = m2;
}
}
//swap
uint32_t tmp[NUM_STATES];
for(uint8_t i=0; i<NUM_STATES; i++)
{
tmp[i]=currMetrics[i];
}
for(uint8_t i=0; i<NUM_STATES; i++)
{
currMetrics[i]=prevMetrics[i];
prevMetrics[i]=tmp[i];
}
}
/**
* History chainback to obtain final byte array.
*
* @param out: destination byte array for decoded data.
* @param pos: starting position for the chainback.
* @param len: length of the output in bits.
* @return minimum Viterbi cost at the end of the decode sequence.
*/
uint32_t viterbi_chainback(uint8_t* out, size_t pos, uint16_t len)
{
uint8_t state = 0;
size_t bitPos = len+4;
memset(out, 0, (len-1)/8+1);
while(pos > 0)
{
bitPos--;
pos--;
uint16_t bit = viterbi_history[pos]&((1<<(state>>4)));
state >>= 1;
if(bit)
{
state |= 0x80;
out[bitPos/8]|=1<<(7-(bitPos%8));
}
}
uint32_t cost = prevMetrics[0];
for(size_t i = 0; i < NUM_STATES; i++)
{
uint32_t m = prevMetrics[i];
if(m < cost) cost = m;
}
return cost;
}
/**
* Utility function to compute the absolute value of a difference between
* two fixed-point values.
*
* @param v1: first value
* @param v2: second value
* @return abs(v1-v2)
*/
uint16_t q_abs_diff(const uint16_t v1, const uint16_t v2)
{
if(v2 > v1) return v2 - v1;
return v1 - v2;
}
/**
* Reset the decoder state.
*
*/
void viterbi_reset(void)
{
memset((uint8_t*)viterbi_history, 0, 2*244);
memset((uint8_t*)currMetrics, 0, 4*NUM_STATES);
memset((uint8_t*)prevMetrics, 0, 4*NUM_STATES);
memset((uint8_t*)currMetricsData, 0, 4*NUM_STATES);
memset((uint8_t*)prevMetricsData, 0, 4*NUM_STATES);
}
//M17
//misc
void send_preamble(const uint8_t type)
{
float symb;
@ -661,356 +65,3 @@ void send_eot(void)
fwrite((uint8_t*)&symb, sizeof(float), 1, stdout);
}
}
//out - unpacked bits
//in - packed raw bits
//fn - frame number
void conv_encode_stream_frame(uint8_t* out, const uint8_t* in, const uint16_t fn)
{
uint8_t pp_len = sizeof(P_2);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[144+4+4]; //unpacked data
memset(ud, 0, 144+4+4);
//unpack frame number
for(uint8_t i=0; i<16; i++)
{
ud[4+i]=(fn>>(15-i))&1;
}
//unpack data
for(uint8_t i=0; i<16; i++)
{
for(uint8_t j=0; j<8; j++)
{
ud[4+16+i*8+j]=(in[i]>>(7-j))&1;
}
}
//encode
for(uint8_t i=0; i<144+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//printf("%d%d", G1, G2);
if(P_2[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_2[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//printf("pb=%d\n", pb);
}
//out - unpacked bits
//in - packed raw bits
void conv_encode_packet_frame(uint8_t* out, uint8_t* in)
{
uint8_t pp_len = sizeof(P_3);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[206+4+4]; //unpacked data
memset(ud, 0, 206+4+4);
//unpack data
for(uint8_t i=0; i<26; i++)
{
for(uint8_t j=0; j<8; j++)
{
if(i<=24 || j<=5)
ud[4+i*8+j]=(in[i]>>(7-j))&1;
}
}
//encode
for(uint8_t i=0; i<206+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//fprintf(stderr, "%d%d", G1, G2);
if(P_3[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_3[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//fprintf(stderr, "pb=%d\n", pb);
}
//out - unpacked bits
//in - packed raw bits (LSF struct)
void conv_encode_LSF(uint8_t* out, const struct LSF *in)
{
uint8_t pp_len = sizeof(P_1);
uint8_t p=0; //puncturing pattern index
uint16_t pb=0; //pushed punctured bits
uint8_t ud[240+4+4]; //unpacked data
memset(ud, 0, 240+4+4);
//unpack DST
for(uint8_t i=0; i<8; i++)
{
ud[4+i] =((in->dst[0])>>(7-i))&1;
ud[4+i+8] =((in->dst[1])>>(7-i))&1;
ud[4+i+16]=((in->dst[2])>>(7-i))&1;
ud[4+i+24]=((in->dst[3])>>(7-i))&1;
ud[4+i+32]=((in->dst[4])>>(7-i))&1;
ud[4+i+40]=((in->dst[5])>>(7-i))&1;
}
//unpack SRC
for(uint8_t i=0; i<8; i++)
{
ud[4+i+48]=((in->src[0])>>(7-i))&1;
ud[4+i+56]=((in->src[1])>>(7-i))&1;
ud[4+i+64]=((in->src[2])>>(7-i))&1;
ud[4+i+72]=((in->src[3])>>(7-i))&1;
ud[4+i+80]=((in->src[4])>>(7-i))&1;
ud[4+i+88]=((in->src[5])>>(7-i))&1;
}
//unpack TYPE
for(uint8_t i=0; i<8; i++)
{
ud[4+i+96] =((in->type[0])>>(7-i))&1;
ud[4+i+104]=((in->type[1])>>(7-i))&1;
}
//unpack META
for(uint8_t i=0; i<8; i++)
{
ud[4+i+112]=((in->meta[0])>>(7-i))&1;
ud[4+i+120]=((in->meta[1])>>(7-i))&1;
ud[4+i+128]=((in->meta[2])>>(7-i))&1;
ud[4+i+136]=((in->meta[3])>>(7-i))&1;
ud[4+i+144]=((in->meta[4])>>(7-i))&1;
ud[4+i+152]=((in->meta[5])>>(7-i))&1;
ud[4+i+160]=((in->meta[6])>>(7-i))&1;
ud[4+i+168]=((in->meta[7])>>(7-i))&1;
ud[4+i+176]=((in->meta[8])>>(7-i))&1;
ud[4+i+184]=((in->meta[9])>>(7-i))&1;
ud[4+i+192]=((in->meta[10])>>(7-i))&1;
ud[4+i+200]=((in->meta[11])>>(7-i))&1;
ud[4+i+208]=((in->meta[12])>>(7-i))&1;
ud[4+i+216]=((in->meta[13])>>(7-i))&1;
}
//unpack CRC
for(uint8_t i=0; i<8; i++)
{
ud[4+i+224]=((in->crc[0])>>(7-i))&1;
ud[4+i+232]=((in->crc[1])>>(7-i))&1;
}
//encode
for(uint8_t i=0; i<240+4; i++)
{
uint8_t G1=(ud[i+4] +ud[i+1]+ud[i+0])%2;
uint8_t G2=(ud[i+4]+ud[i+3]+ud[i+2] +ud[i+0])%2;
//printf("%d%d", G1, G2);
if(P_1[p])
{
out[pb]=G1;
pb++;
}
p++;
p%=pp_len;
if(P_1[p])
{
out[pb]=G2;
pb++;
}
p++;
p%=pp_len;
}
//printf("pb=%d\n", pb);
}
uint16_t LSF_CRC(const struct LSF *in)
{
uint8_t d[28];
memcpy(&d[0], in->dst, 6);
memcpy(&d[6], in->src, 6);
memcpy(&d[12], in->type, 2);
memcpy(&d[14], in->meta, 14);
return CRC_M17(d, 28);
}
//decodes a 6-byte long array to a callsign
void decode_callsign_bytes(uint8_t *outp, const uint8_t *inp)
{
uint64_t encoded=0;
//repack the data to a uint64_t
for(uint8_t i=0; i<6; i++)
encoded|=(uint64_t)inp[5-i]<<(8*i);
//check if the value is reserved (not a callsign)
if(encoded>=262144000000000ULL)
{
if(encoded==0xFFFFFFFFFFFF) //broadcast
{
sprintf((char*)outp, "#BCAST");
}
else
{
outp[0]=0;
}
return;
}
//decode the callsign
uint8_t i=0;
while(encoded>0)
{
outp[i]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/."[encoded%40];
encoded/=40;
i++;
}
outp[i]=0;
}
//decodes a 48-bit value to a callsign
void decode_callsign_value(uint8_t *outp, const uint64_t inp)
{
uint64_t encoded=inp;
//check if the value is reserved (not a callsign)
if(encoded>=262144000000000ULL)
{
if(encoded==0xFFFFFFFFFFFF) //broadcast
{
sprintf((char*)outp, "#BCAST");
}
else
{
outp[0]=0;
}
return;
}
//decode the callsign
uint8_t i=0;
while(encoded>0)
{
outp[i]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/."[encoded%40];
encoded/=40;
i++;
}
outp[i]=0;
}
//encode callsign into a 48-bit value
uint8_t encode_callsign(uint64_t* out, const uint8_t* inp)
{
//assert inp length
if(strlen((const char*)inp)>9)
{
return -1;
}
const uint8_t charMap[40]=" ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-/.";
uint64_t tmp=0;
if(strcmp((const char*)inp, "ALL")==0)
{
*out=0xFFFFFFFFFFFF;
return 0;
}
for(int8_t i=strlen((const char*)inp)-1; i>=0; i--)
{
for(uint8_t j=0; j<40; j++)
{
if(inp[i]==charMap[j])
{
tmp=tmp*40+j;
break;
}
}
}
*out=tmp;
return 0;
}
//soft decodes LICH into a 6-byte array
//input - soft bits
//output - an array of packed bits
void decode_LICH(uint8_t* outp, const uint16_t* inp)
{
uint16_t tmp;
memset(outp, 0, 5);
tmp=golay24_sdecode(&inp[0]);
outp[0]=(tmp>>4)&0xFF;
outp[1]|=(tmp&0xF)<<4;
tmp=golay24_sdecode(&inp[1*24]);
outp[1]|=(tmp>>8)&0xF;
outp[2]=tmp&0xFF;
tmp=golay24_sdecode(&inp[2*24]);
outp[3]=(tmp>>4)&0xFF;
outp[4]|=(tmp&0xF)<<4;
tmp=golay24_sdecode(&inp[3*24]);
outp[4]|=(tmp>>8)&0xF;
outp[5]=tmp&0xFF;
}
//calculate L2 norm between two len-dimensional vectors of floats
float eucl_norm(const float* in1, const int8_t* in2, uint8_t len)
{
float tmp = 0.0f;
for(uint8_t i=0; i<len; i++)
{
tmp += powf(in1[i]-(float)in2[i], 2.0f);
}
return sqrt(tmp);
}

109
SP5WWP/lib/m17lib.h
Wyświetl plik

@ -4,128 +4,21 @@
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 28 December 2023
//-------------------------------
#ifndef M17_LIB
#define M17_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <inttypes.h>
#define BSB_SPS 10 //samples per symbol
#define FLT_SPAN 8 //baseband RRC filter span in symbols
#define SYM_PER_SWD 8 //symbols per syncword
#define SW_LEN (BSB_SPS*SYM_PER_SWD) //syncword detector length
#define SYM_PER_PLD 184 //symbols per payload in a frame
#define SYM_PER_FRA 192 //symbols per whole 40 ms frame
#define RRC_DEV 7168.0f //.rrc file deviation for +1.0 symbol
//L2 metric threshold
#define DIST_THRESH 2.0f //threshold for distance (syncword detection)
//Viterbi
#define K 5 //constraint length
#define NUM_STATES (1 << (K - 1)) //number of states
//type declarations
struct LSF
{
uint8_t dst[6];
uint8_t src[6];
uint8_t type[2];
uint8_t meta[112/8];
uint8_t crc[2];
};
//consts
//syncword patterns (RX) TODO:Compute those at runtime from the consts below
extern const int8_t lsf_sync[8];
extern const int8_t str_sync[8];
extern const int8_t pkt_sync[8];
//symbol levels (RX)
extern const float symbs[4];
//dibits-symbols map (TX)
extern const int8_t symbol_map[4];
//syncwords
extern const uint16_t SYNC_LSF;
extern const uint16_t SYNC_STR;
extern const uint16_t SYNC_PKT;
extern const uint16_t SYNC_BER;
extern const uint16_t EOT_MRKR;
//puncturing pattern P_1
extern const uint8_t P_1[61];
//puncturing pattern P_2
extern const uint8_t P_2[12];
//puncturing pattern P_3
extern const uint8_t P_3[8];
//M17 CRC polynomial
extern const uint16_t M17_CRC_POLY;
//sample RRC filter for 48kHz sample rate
//alpha=0.5, span=8, sps=10, gain=sqrt(sps)
extern const float taps_10[8*10+1];
//sample RRC filter for 24kHz sample rate
//alpha=0.5, span=8, sps=5, gain=sqrt(sps)
extern const float taps_5[8*5+1];
//Golay coding
extern const uint16_t encode_matrix[12];
extern const uint16_t decode_matrix[12];
//randomizing pattern
extern const uint8_t rand_seq[46];
//interleaver pattern
extern const uint16_t intrl_seq[368];
#include <stdint.h>
//functions
uint16_t CRC_M17(const uint8_t *in, const uint16_t len);
uint32_t golay24_encode(const uint16_t data);
void int_to_soft(uint16_t* out, const uint16_t in, uint8_t len);
uint16_t soft_to_int(const uint16_t* in, uint8_t len);
uint16_t div16(uint16_t a, uint16_t b);
uint16_t mul16(uint16_t a, uint16_t b);
uint16_t soft_bit_XOR(const uint16_t a, const uint16_t b);
void soft_XOR(uint16_t* out, const uint16_t* a, const uint16_t* b, uint8_t len);
uint32_t s_popcount(const uint16_t* in, uint8_t siz);
void s_calc_checksum(uint16_t* out, const uint16_t* value);
uint32_t s_detect_errors(const uint16_t* codeword);
uint16_t golay24_sdecode(const uint16_t* codeword);
uint32_t viterbi_decode(uint8_t* out, const uint16_t* in, uint16_t len);
uint32_t viterbi_decode_punctured(uint8_t* out, const uint16_t* in, const uint8_t* punct, const uint16_t in_len, const uint16_t p_len);
void viterbi_decode_bit(uint16_t s0, uint16_t s1, size_t pos);
uint32_t viterbi_chainback(uint8_t* out, size_t pos, uint16_t len);
uint16_t q_abs_diff(const uint16_t v1, const uint16_t v2);
void viterbi_reset(void);
void send_preamble(const uint8_t type);
void send_syncword(const uint16_t syncword);
void send_data(const uint8_t* in);
void send_eot(void);
void conv_encode_stream_frame(uint8_t* out, const uint8_t* in, const uint16_t fn);
void conv_encode_packet_frame(uint8_t* out, uint8_t* in);
void conv_encode_LSF(uint8_t* out, const struct LSF *in);
uint16_t LSF_CRC(const struct LSF *in);
void decode_callsign_bytes(uint8_t *outp, const uint8_t *inp);
void decode_callsign_value(uint8_t *outp, const uint64_t inp);
uint8_t encode_callsign(uint64_t* out, const uint8_t* inp);
void decode_LICH(uint8_t* outp, const uint16_t* inp);
float eucl_norm(const float* in1, const int8_t* in2, uint8_t len);
#ifdef __cplusplus
}
#endif
#endif

136
SP5WWP/lib/m17math.c 100644
Wyświetl plik

@ -0,0 +1,136 @@
//--------------------------------------------------------------------
// M17 C library - m17math.c
//
// This file contains:
// - absolute difference value
// - Euclidean norm (L2) calculation for n-dimensional vectors (float)
// - soft-valued arrays to integer conversion (and vice-versa)
// - fixed-valued multiplication and division
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <math.h>
#include "m17math.h"
/**
* @brief Utility function returning the absolute value of a difference between
* two fixed-point values.
*
* @param v1 First value.
* @param v2 Second value.
* @return abs(v1-v2) value.
*/
uint16_t q_abs_diff(const uint16_t v1, const uint16_t v2)
{
if(v2 > v1) return v2 - v1;
return v1 - v2;
}
/**
* @brief Calculate L2 norm between two n-dimensional vectors of floats.
*
* @param in1 Vector 1.
* @param in2 Vector 2.
* @param n Vectors' size.
* @return float L2 norm.
*/
float eucl_norm(const float* in1, const int8_t* in2, const uint8_t n)
{
float tmp = 0.0f;
for(uint8_t i=0; i<n; i++)
{
tmp += powf(in1[i]-(float)in2[i], 2.0f);
}
return sqrt(tmp);
}
/**
* @brief Convert an unsigned int into an array of soft, fixed-point values.
* Maximum length is 16. LSB is at index 0.
* @param out Pointer to an array of uint16_t.
* @param in Input value.
* @param len Input's bit length.
*/
void int_to_soft(uint16_t* out, const uint16_t in, const uint8_t len)
{
for(uint8_t i=0; i<len; i++)
{
(in>>i)&1 ? (out[i]=0xFFFF) : (out[i]=0);
}
}
/**
* @brief Convert an array of soft, fixed-point
* Maximum length is 16. LSB is at index 0.
* @param in Pointer to an array of uint16_t.
* @param len Input's length.
* @return uint16_t Return value.
*/
uint16_t soft_to_int(const uint16_t* in, const uint8_t len)
{
uint16_t tmp=0;
for(uint8_t i=0; i<len; i++)
{
if(in[i]>0x7FFFU)
tmp|=(1<<i);
}
return tmp;
}
/**
* @brief 1st quadrant fixed point division with saturation.
*
* @param a Dividend.
* @param b Divisor.
* @return uint16_t Quotient = a/b.
*/
uint16_t div16(const uint16_t a, const uint16_t b)
{
uint32_t aa=(uint32_t)a<<16;
uint32_t r=aa/b;
return r<=0xFFFFU ? r : 0xFFFFU;
}
/**
* @brief 1st quadrant fixed point multiplication.
*
* @param a Multiplicand.
* @param b Multiplier.
* @return uint16_t Product = a*b.
*/
uint16_t mul16(const uint16_t a, const uint16_t b)
{
return (uint16_t)(((uint32_t)a*b)>>16);
}
/**
* @brief Bilinear interpolation (soft-valued expansion) for XOR.
*
* @param a Input A.
* @param b Input B.
* @return uint16_t Output = A xor B.
*/
uint16_t soft_bit_XOR(const uint16_t a, const uint16_t b)
{
return mul16(div16(0xFFFF-b, 0xFFFF), div16(a, 0xFFFF)) + mul16(div16(b, 0xFFFF), div16(0xFFFF-a, 0xFFFF));
}
/**
* @brief XOR for vectors of soft-valued logic.
* Max length is 16.
* @param out Output vector = A xor B.
* @param a Input vector A.
* @param b Input vector B.
* @param len Vectors' size.
*/
void soft_XOR(uint16_t* out, const uint16_t* a, const uint16_t* b, const uint8_t len)
{
for(uint8_t i=0; i<len; i++)
out[i]=soft_bit_XOR(a[i], b[i]);
}

31
SP5WWP/lib/m17math.h 100644
Wyświetl plik

@ -0,0 +1,31 @@
//--------------------------------------------------------------------
// M17 C library - m17math.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_MATH_LIB
#define M17_MATH_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
//vars
;
//functions
uint16_t q_abs_diff(const uint16_t v1, const uint16_t v2);
float eucl_norm(const float* in1, const int8_t* in2, const uint8_t n);
void int_to_soft(uint16_t* out, const uint16_t in, const uint8_t len);
uint16_t soft_to_int(const uint16_t* in, const uint8_t len);
uint16_t div16(const uint16_t a, const uint16_t b);
uint16_t mul16(const uint16_t a, const uint16_t b);
uint16_t soft_bit_XOR(const uint16_t a, const uint16_t b);
void soft_XOR(uint16_t* out, const uint16_t* a, const uint16_t* b, const uint8_t len);
#ifdef __cplusplus
}
#endif
#endif

202
SP5WWP/lib/m17viterbi.c 100644
Wyświetl plik

@ -0,0 +1,202 @@
//--------------------------------------------------------------------
// M17 C library - m17viterbi.c
//
// This file contains:
// - all the Viterbi decoder stuff
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#include <stdio.h>
#include <string.h>
#include "m17viterbi.h"
#include "m17consts.h"
static uint32_t prevMetrics[NUM_STATES];
static uint32_t currMetrics[NUM_STATES];
static uint32_t prevMetricsData[NUM_STATES];
static uint32_t currMetricsData[NUM_STATES];
static uint16_t viterbi_history[244];
/**
* @brief Decode unpunctured convolutionally encoded data.
*
* @param out Destination array where decoded data is written.
* @param in Input data.
* @param len Input length in bits.
* @return Number of bit errors corrected.
*/
uint32_t viterbi_decode(uint8_t* out, const uint16_t* in, const uint16_t len)
{
if(len > 244*2)
fprintf(stderr, "Input size exceeds max history\n");
viterbi_reset();
size_t pos = 0;
for(size_t i = 0; i < len; i += 2)
{
uint16_t s0 = in[i];
uint16_t s1 = in[i + 1];
viterbi_decode_bit(s0, s1, pos);
pos++;
}
return viterbi_chainback(out, pos, len/2);
}
/**
* @brief Decode punctured convolutionally encoded data.
*
* @param out Destination array where decoded data is written.
* @param in Input data.
* @param punct Puncturing matrix.
* @param in_len Input data length.
* @param p_len Puncturing matrix length (entries).
* @return Number of bit errors corrected.
*/
uint32_t viterbi_decode_punctured(uint8_t* out, const uint16_t* in, const uint8_t* punct, const uint16_t in_len, const uint16_t p_len)
{
if(in_len > 244*2)
fprintf(stderr, "Input size exceeds max history\n");
uint16_t umsg[244*2]; //unpunctured message
uint8_t p=0; //puncturer matrix entry
uint16_t u=0; //bits count - unpunctured message
uint16_t i=0; //bits read from the input message
while(i<in_len)
{
if(punct[p])
{
umsg[u]=in[i];
i++;
}
else
{
umsg[u]=0x7FFF;
}
u++;
p++;
p%=p_len;
}
return viterbi_decode(out, umsg, u) - (u-in_len)*0x7FFF;
}
/**
* @brief Decode one bit and update trellis.
*
* @param s0 Cost of the first symbol.
* @param s1 Cost of the second symbol.
* @param pos Bit position in history.
*/
void viterbi_decode_bit(uint16_t s0, uint16_t s1, const size_t pos)
{
static const uint16_t COST_TABLE_0[] = {0, 0, 0, 0, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF};
static const uint16_t COST_TABLE_1[] = {0, 0xFFFF, 0xFFFF, 0, 0, 0xFFFF, 0xFFFF, 0};
for(uint8_t i = 0; i < NUM_STATES/2; i++)
{
uint32_t metric = q_abs_diff(COST_TABLE_0[i], s0)
+ q_abs_diff(COST_TABLE_1[i], s1);
uint32_t m0 = prevMetrics[i] + metric;
uint32_t m1 = prevMetrics[i + NUM_STATES/2] + (0x1FFFE - metric);
uint32_t m2 = prevMetrics[i] + (0x1FFFE - metric);
uint32_t m3 = prevMetrics[i + NUM_STATES/2] + metric;
uint8_t i0 = 2 * i;
uint8_t i1 = i0 + 1;
if(m0 >= m1)
{
viterbi_history[pos]|=(1<<i0);
currMetrics[i0] = m1;
}
else
{
viterbi_history[pos]&=~(1<<i0);
currMetrics[i0] = m0;
}
if(m2 >= m3)
{
viterbi_history[pos]|=(1<<i1);
currMetrics[i1] = m3;
}
else
{
viterbi_history[pos]&=~(1<<i1);
currMetrics[i1] = m2;
}
}
//swap
uint32_t tmp[NUM_STATES];
for(uint8_t i=0; i<NUM_STATES; i++)
{
tmp[i]=currMetrics[i];
}
for(uint8_t i=0; i<NUM_STATES; i++)
{
currMetrics[i]=prevMetrics[i];
prevMetrics[i]=tmp[i];
}
}
/**
* @brief History chainback to obtain final byte array.
*
* @param out Destination byte array for decoded data.
* @param pos Starting position for the chainback.
* @param len Length of the output in bits.
* @return Minimum Viterbi cost at the end of the decode sequence.
*/
uint32_t viterbi_chainback(uint8_t* out, size_t pos, uint16_t len)
{
uint8_t state = 0;
size_t bitPos = len+4;
memset(out, 0, (len-1)/8+1);
while(pos > 0)
{
bitPos--;
pos--;
uint16_t bit = viterbi_history[pos]&((1<<(state>>4)));
state >>= 1;
if(bit)
{
state |= 0x80;
out[bitPos/8]|=1<<(7-(bitPos%8));
}
}
uint32_t cost = prevMetrics[0];
for(size_t i = 0; i < NUM_STATES; i++)
{
uint32_t m = prevMetrics[i];
if(m < cost) cost = m;
}
return cost;
}
/**
* @brief Reset the decoder state. No args.
*
*/
void viterbi_reset(void)
{
memset((uint8_t*)viterbi_history, 0, 2*244);
memset((uint8_t*)currMetrics, 0, 4*NUM_STATES);
memset((uint8_t*)prevMetrics, 0, 4*NUM_STATES);
memset((uint8_t*)currMetricsData, 0, 4*NUM_STATES);
memset((uint8_t*)prevMetricsData, 0, 4*NUM_STATES);
}

28
SP5WWP/lib/m17viterbi.h 100644
Wyświetl plik

@ -0,0 +1,28 @@
//--------------------------------------------------------------------
// M17 C library - m17viterbi.h
//
// Wojciech Kaczmarski, SP5WWP
// M17 Project, 29 December 2023
//--------------------------------------------------------------------
#ifndef M17_VITERBI_LIB
#define M17_VITERBI_LIB
#ifdef __cplusplus
extern "C" {
#endif
#include <stdint.h>
//vars
;
//functions
uint32_t viterbi_decode(uint8_t* out, const uint16_t* in, const uint16_t len);
uint32_t viterbi_decode_punctured(uint8_t* out, const uint16_t* in, const uint8_t* punct, const uint16_t in_len, const uint16_t p_len);
void viterbi_decode_bit(uint16_t s0, uint16_t s1, size_t pos);
uint32_t viterbi_chainback(uint8_t* out, size_t pos, const uint16_t len);
void viterbi_reset(void);
#ifdef __cplusplus
}
#endif
#endif

6
SP5WWP/m17-coder/Makefile
Wyświetl plik

@ -1,5 +1,7 @@
m17-coder-sym: m17-coder-sym.c ../lib/m17lib.c ../lib/m17lib.h
gcc -O2 -Wall m17-coder-sym.c ../lib/m17lib.c -o m17-coder-sym -lm
m17-coder-sym:
gcc -O2 -Wall m17-coder-sym.c \
../lib/*.c \
-o m17-coder-sym -lm
clean:
rm -f m17-coder-sym

9
SP5WWP/m17-coder/m17-coder-sym.c
Wyświetl plik

@ -2,7 +2,14 @@
#include <string.h>
#include <stdint.h>
#include "../lib/m17lib.h"
//#include "../lib/m17lib.h"
//#include "../lib/m17call.h"
#include "../lib/m17consts.h"
//#include "../lib/m17convol.h"
//#include "../lib/m17crc.h"
//#include "../lib/m17golay.h"
//#include "../lib/m17math.h"
//#include "../lib/m17viterbi.h"
//#define FN60_DEBUG