Multirate digital transmission method and device for implementing said method

A method for digitally transmitting a speech signal at a given transmission rate chosen from amongst a group of predetermined different transmission rates, said method including the steps of:

• sampling and digitally coding said speech signal;
• receiving the coded speech signal and generating partial auto correlation (PARCOR) coefficients;
• recoding said PARCOR coefficients;
• extracting from said coded speech signal and said PARCOR coefficients a residual signal;
• filtering said residual signal and providing a residual base band signal and an information signal indicative of the energy in the high frequency band of said residual signal;
• recoding said filtered residual base band signal and said high frequency band energy signal;
• splitting said filtered residual base band signal into a plurality of sub-bands;
• regrouping said sub-bands into subgroups, each of said sub-groups being assigned to one of said predetermined different transmission rates;
• quantizing the sub-band signals in each of said sub-groups by dynamically allocating the quantizing bits among the sub-bands in each of said sub-groups and quantizing each sub-group at a rate which is a function of the transmission rate assigned to each said sub-group;
• multiplexing said quantized sub-band signals by arranging the bits representing the energy of said high frequency band of said residual signal, the PARCOR bits, and the bits obtained from said quantizing into a multirate frame comprising a plurality of zones;
• ordering said plurality of zones from most significant to least significant, said most significant zone receiving those bits which are most significant to the reconstruction of said speech signal, said most significant zone corresponding to the lowest of said transmission rates, each successive remaining zone of said plurality of zones of successively lesser significant order and corresponding to successively higher of said transmission rates receiving those bits from said coding and said quantizing which are of successively lesser significance to the reconstruction of said speech signal;
• and,
• selecting for transmission all or a portion of said multirate frame, said portion being commenced with said most significant zone, said portion continuing, if at all, with the next zone of said successively lesser significant order, the number of said zones in said transmission being a function of said given transmission rate.

A method for digitally transmitting a limited frequency band speech signal at a given transmission rate chosen from amongst a group of predetermined different transmission rates, said method including the steps of:

• sampling and digitally coding said speech signal;
• selecting at least part of said frequency band represented by said digitally coded speech signal;
• dividing said selected frequency band into a fixed number of sub-bands;
• regrouping said sub-bands into subgroups, each of said sub-groups being assigned to one of said predetermined different transmission rates;
• quantizing the sub-band signals in each of said sub-groups by dynamically allocating the quantizing bits among the sub-bands in each of said sub-groups and quantizing each sub-group at a rate which is a function of the transmission rate assigned to each said sub-group;
• multiplexing said quantized sub-band signals by arranging the bits obtained from said coding and the bits obtained from said quantizing of said sub-band signals into a multirate frame comprising a plurality of zones wherein at least one of said zones includes sub-band signals from sub-bands separated in frequency from each other by one or more sub-bands not included in said zone;
• ordering said plurality of zones from most significant to least significant, said most significant zone receiving those bits which are most significant to the reconstruction of said speech signal, said most significant zone corresponding to the lowest of said transmission rates, each remaining successive zone of said plurality of zones of successively lesser significant order and corresponding to successively higher of said transmission rates receiving those bits from said coding and said quantizing which are of successively lesser significance to the reconstruction of said speech signal;
• and,
• selecting for transmission all or a portion of said multirate frame, said portion being commenced with said most significant zone, said portion continuing, it at all, with the next zone of said successively lesser significant order, the number of said zones in said transmission being a function of said given transmission rate.

A method for digitally transmitting a speech signal at a given transmission rate chosen from amongst a group of predetermined different transmission rates, said method including the steps of:

• sampling and digitally coding said speech signal;
• receiving the coded speech signal and generating partial auto correlation (PARCOR) coefficients;
• recoding said PARCOR coefficients;
• extracting from said coded speech signal and said PARCOR coefficients a residual signal;
• filtering said residual signal and providing a residual base band signal and an information signal indicative of the energy in the high frequency band of said residual signal;
• recoding said filtered residual base band signal and said high frequency band energy signal;
• splitting said filtered residual base band signal into a plurality of sub-bands;
• regrouping said sub-bands into subgroups, each of said sub-groups being assigned to one of said predetermined different transmission rates;
• quantizing the sub-band signals in each of said sub-groups by dynamically allocating the quantizing bits among the sub-bands in each of said sub-groups and quantizing each sub-group at a rate which is a function of the transmission rate assigned to each of said sub-group;
• multiplexing said quantized sub-band signals by arranging the bits representing the energy of said high frequency band of said residual signal, the PARCOR bits, and the bits obtained from said quantizing into a multirate frame comprising a plurality of zones wherein at least one of said zones includes sub-band signals from sub-bands separated in frequency from each other by one or more sub-bands not included in said zone;
• ordering said plurality of zones from most significant to least significant, said most significant zone receiving those bits which are most significant to the reconstruction of said speech signal, said most significant zone corresponding to the lowest of said transmission rates, each successive remaining zone of said plurality of zones of successively lesser significant order and corresponding to successively higher of said transmission rates receiving those bits from said coding and said quantizing which are of successively lesser significance to the reconstruction of said speech signal;
• and,
• selecting for transmission all or a portion of said multirate frame, said portion being commenced with said most significant zone, said portion continuing, if at all, with the next zone of said successively lesser significant order, the number of said zones in said transmission being a function of said given transmission rate.