/* Global definitions for Reed-Solomon encoder/decoder * Phil Karn KA9Q, August 1997 * * The parameters MM and KK specify the Reed-Solomon code parameters. * * Set MM to be the size of each code symbol in bits. The Reed-Solomon * block size will then be NN = 2**M - 1 symbols. Supported values are * defined in rs.c. * * Set KK to be the number of data symbols in each block, which must be * less than the block size. The code will then be able to correct up * to NN-KK erasures or (NN-KK)/2 errors, or combinations thereof with * each error counting as two erasures. */ #define MM 8 /* RS code over GF(256) */ #define KK 223 /* KK = number of information symbols */ #define NN ((1 << MM) - 1) #if (MM <= 8) typedef unsigned char dtype; #else typedef unsigned int dtype; #endif /* This defines the type used to store an element of the Galois Field * used by the code. Make sure this is something larger than a char if * if anything larger than GF(256) is used. * * Note: unsigned char will work up to GF(256) but int seems to run * faster on the Pentium. */ typedef int gf; /* Initialization function */ void init_rs(void); /* Reed-Solomon encoding * data[] is the input block, parity symbols are placed in bb[] * bb[] may lie past the end of the data, e.g., for (255,223): * encode_rs(&data[0],&data[223]); */ int rs32e(dtype data[KK], dtype bb[NN-KK]); /* Reed-Solomon erasures-and-errors decoding * The received block goes into data[], and a list of zero-origin * erasure positions, if any, goes in eras_pos[] with a count in no_eras. * * The decoder corrects the symbols in place, if possible and returns * the number of corrected symbols. If the codeword is illegal or * uncorrectible, the data array is unchanged and -1 is returned */ int rs32d(dtype data[NN], int eras_pos[], int no_eras);