Hierarchical and adaptive multi-carrier digital modulation and demodulation

A method for blindly decoding a binary stream from multi-carrier data symbols, and comprising, at a receive side:

• grouping a plurality of carriers into a plurality of carrier groups, each of the plurality of carrier groups including a separate selection of the plurality of carriers;
• iteratively performing, for each carrier group of the plurality of carrier groups,
• for a given carrier group including N carriers, selecting a group constellation diagram for demodulation of the N carriers of the given carrier group by statistical characterization of the N carriers or part thereof, the selected group constellation diagram including constellation points defined as respective hierarchical sums of I super-imposed constellation vectors having decreasing powers with the P strongest constellation vectors and the Q weakest constellation vectors being omitted, P being a positive integer parameter, Q being a positive integer parameter, I being a positive integer value that determines a maximum constellation size;
• and
• de-mapping the respective N carriers of the given carrier group into N respective binary words based on the selected group constellation diagram, with the P and Q parameters remaining unchanged for the given carrier group and being re-evaluated per carrier group and per multi-carrier data symbol.

A transmitter comprising:

• a processor configured to,
• group a plurality of carriers into a plurality of carrier groups, each of the plurality of carrier groups including a separate selection of the plurality of carriers;
• iteratively perform, for each carrier group of the plurality of carrier groups,
• for a given carrier group including N carriers, select a group constellation diagram for modulation of the N carriers of the given carrier group, the selected group constellation diagram including constellation points defined as respective hierarchical sums of I super-imposed constellation vectors having decreasing powers with the P strongest constellation vectors and the Q weakest constellation vectors being omitted, P being a positive integer parameter, Q being a positive integer parameter, and I being a positive integer value that determines a maximum constellation size;
• and
• map N binary words for respective N carriers of the given carrier group onto N respective constellation points of the selected constellation diagram, with the P and Q parameters remaining unchanged for the given carrier group and being adjusted to their optimal value per carrier group and per multi-carrier data symbol.

A receiver comprising:

• a processor configured to,
• group a plurality of carriers into a plurality of carrier groups, each of the plurality of carrier groups including a separate selection of the plurality of carriers;
• iteratively perform, for each carrier group of the plurality of carrier groups,
• for a given carrier group including N carriers, select a group constellation diagram for demodulation of the N carriers of the given carrier group by statistical characterization of the N carriers or part thereof, the selected group constellation diagram including constellation points defined as respective hierarchical sums of I super-imposed constellation vectors having decreasing powers with the P strongest constellation vectors and the Q weakest constellation vectors being omitted, P being a positive integer parameter, Q being a positive integer parameter, I being a positive integer value that determines a maximum constellation size;
• and
• de-map the respective N carriers of the given carrier group into N respective binary words based on the selected group constellation diagram, with the P and Q parameters remaining unchanged for the given carrier group and being re-evaluated per carrier group and per multi-carrier data symbol.