Block coding scheme for fractional-bit transmission

Claim 2

Method of encoding data by converting a bit stream representing the data into symbols of an expanded two-dimensional symbol set, each symbol representing a discrete value of of a finite number of discrete values; said method comprising the steps of: - subdividing said symbol set into a plurality t of disjunct symbol subsets, each comprising a number of symbols which is not an integer power of two, and each comprising outer symbols located at the periphery of the symbol set and inner symbols; - separating said data bit stream into fixed-length blocks each comprising a number of bits which is not an integer multiple of a given number m; each such block to be represented by w of said symbols; - expanding by trellis encoding each one of w separate p-bit subgroups derived from a first portion (s) of each block, and using each expanded bit group for specifying one of the t symbol subsets; - block encoding, including expansion, the remaining second portion (r) of each block, for deriving w separate q-bit subgroups, each for selecting one of the symbols of a specified symbol set; and - selecting outer symbols less frequently than inner symbols.

Claim 3

Apparatus for encoding data by converting a bit stream representing the data into symbols of a two-dimensional expanded symbol set. each symbol representing a discrete value of of a finite number of discrete values, said apparatus comprising: - buffer storage means (13) for receiving a block of bits of the data stream to be encoded, the predetermined number of bits of said block being not an integer multiple of a given number m; - symbol selection means (14) for selecting one symbol of said two-dimensional expanded symbol set in reponse to two bit groups furnished to two groups of inputs of said symbol selection means; a bit group furnished to a first group (15) of inputs specifying one of t disjunct subsets of symbols, each subset comprising a number of symbols which is not an integer power of two; and a bit group furnished to a second group (16) of inputs selecting one particular symbol of a specified subset; - first scanning means (18) for selecting, in successive time intervals, each one of w separate p-bit subgroups of a first portion (17; s) of the bits stored in said buffer storage means; - convolutional encoding means (20) for receiving the bits of each p-bit subgroup (19) selected by said first scanning means, and for furnishing a respective expanded bit group from its output to the first group of inputs of said symbol selection means; - block encoding means (22) connected to said buffer storage means for receiving a second portion (21; r) of the bits stored in said buffer storage means, and for furnishing a resulting expanded bit group (23) in w separate q-bit subgroups at its outputs, and - second scanning means (24) for selecting, in successive time intervals, each one of said w q-bit output bit groups of said block encoding means, and for furnishing the respective bit subgroup (25) to the second group of inputs of said symbol selection means.

Claim 4

Apparatus for encoding data by converting a bit stream representing the data into symbols of a two-dimensional expanded symbol set. each symbol representing a discrete value of of a finite number of discrete values, said apparatus comprising: - buffer storage means (13) for receiving a block of bits of the data stream to be encoded, the predetermined number of bits of said block being not an integer multiple of a given number m; - symbol selection means (14) for selecting one symbol of said two-dimensional expanded symbol set in reponse to two bit groups furnished to two groups of inputs of said symbol selection means; a bit group furnished to a first group (15) of inputs specifying one of t disjunct subsets of symbols, each subset comprising a number of symbols which is not an integer power of two; and a bit group furnished to a second group (16) of inputs selecting one particular symbol of a specified subset; - first scanning means (18) for selecting, in successive time intervals, each one of w separate p-bit subgroups of a first portion (17; s) of the bits stored in said buffer storage means; - convolutional encoding means (20) for receiving the bits of each p-bit subgroup (19) selected by said first scanning means, and for furnishing a respective expanded bit group from its output to the first group of inputs of said symbol selection means; - block encoding means (22) connected to said buffer storage means for receiving a second portion (21; r) of the bits stored in said buffer storage means, and for furnishing a resulting expanded bit group (23) in w separate q-bit subgroups at its outputs, and - second scanning means (24) for selecting, in successive time intervals, each one of said w q-bit output bit groups of said block encoding means, and for furnishing the respective bit subgroup (25) to the second group of inputs of said symbol selection means.

Claim 5

A method according to Claim 1, for encoding data by symbols from a 144-symbol set subdivided into eight subsets; in which method each said block comprises 49 bits and is represented by w = 8 symbols; said first portion (s) comprising sixteen bits, and said second portion (r) comprising 33 bits; said trellis encoding step effecting a rate 2/3 expansion; and said block encoding step effecting a rate 33/40 expansion.

Claim 6

A method according to Claim 3, in which in said block encoding step, eight labels are generated each comprising five bits for selecting one symbol of any 18-symbol subset comprising sixteen inner symbols and two outer symbols; one of said labels selecting an outer symbol, and the remaining labels selecting inner symbols.

Claim 7

A method according to Claim 1, for encoding data by symbols from an 80-symbol set subdivided into eight subsets; in which method each said block comprises 21 bits and is represented by w = 4 symbols; said first portion (s) comprising eight bits, and said second portion (r) comprising thirteen bits; said trellis encoding step effecting a rate 2/3 expansion; and said block encoding step effecting a rate 13/16 expansion.

Claim 8

A method according to Claim 5, in which in said block encoding step, four labels are generated each comprising four bits for selecting one symbol of any 10-symbol subset comprising eight inner symbols and two outer symbols; one of said labels selecting an outer symbol, and the remaining labels selecting inner symbols.

Claim 9

A method in accordance with Claim 1, in which said block encoding step includes: encoding said second portion (r) to generate w separate q-bit subgroups such that one subgroup selects an outer symbol, and (w-1) subgroups select (w-1) inner symbols from a specified subset.

Claim 10

Apparatus according to Claim 2, in which - said buffer storage means (13) has a capacity to store blocks of 49 bits each; - said symbol selection means (14) providing an alphabet of 144 symbols subdivided into t = 8 subsets; - said first scanning means (18) selecting eight separate bit groups of two bits each from a first portion (s) comprising sixteen bits; - said convolutional encoding means (20) effecting a rate 2/3 expansion; and - said block encoding means (22) receiving a second portion (r) comprising 33 bits, for expanding it into eight groups of five bits each.

Claim 11

Apparatus according to Claim 8, in which said block encoding means (22) receives a second portion (r) comprising 33 bits (X 1 , X2, .... X 33 ) and provides eight output bit groups of five bits each according to the following relation:

Claim 12

Apparatus according to Claim 2, in which - said buffer storage means (13A) has a capacity to store blocks of 21 bits each; - said symbol selection means (14A) providing an alphabet of 80 symbols subdivided into t = 8 subsets; - said first scanning means (18A) selecting four separate bit groups of two bits each from a first portion (s) comprising eight bits; - said convolutional encoding means (20) effecting a rate 2/3 expansion; and - said block encoding means (22A) receiving a second portion (r) comprising 13 bits, for expanding it into four groups of four bits each.

Claim 13

Apparatus according to Claim 10, in which said block encoding means (22A) receives a second portion (r) comprising 13 bits (X 1 , X 2 , ... , X 13 ) and provides four output bit groups of four bits each according to the following relation: