To provide a variable length encoding method that can detect a 1-bit error without fail wherever the error occurs in a code sequence, and decode the code sequence by the same algorithm in both forward and backward directions and to provide its decoding method.
Let the symbol value and the order number that are data to be coded be (n) and (m) respectively, then n modulo m is expressed by a binary number of k bits (k=log2 m) in step S1. A unary part in ([n/m]+1)-bits is generated in the succeeding step S2, where [n/m] indicates an integer part resulting from dividing (n) by (m). The bits of the unary part are used respectively for even number parity in k-bits for a bit just before them. Bit strings generated in steps S1, S2 are concatenated and the first (head) bit and the final bit of the bit string obtained through the concatenation are inverted.
MATSUMOTO SHUICHI
KAKU BOKURYO
KATO SHIGEO
Next Patent: ENCODING AND DECODING DEVICE AND ENCODING AND DECODING METHOD