Method for defining a search sequence for soft-decision sphere decoding algorithm
View Patent ↗A low complexity method for determining a search sequence of nodes for an efficient soft-decision sphere decoding algorithm for use in receivers for quadrature amplitude modulation (QAM) communication signals is achieved by determining a first member of the search sequence by rounding a received symbol (z) to a first constellation symbol (x c ) of the QAM constellation, classifying the remaining constellation symbols (x i ) of the QAM constellation into a plurality of sub-sets of constellation symbols having the same distance metric relative to said first constellation point (x c ) according to a metric d sequ(n) =2a·n=max{|real(x c −x i )|,|imag(x c −x i )|}, a being a scaling factor of the constellation grid, and ordering said sub-sets of constellation symbols in ascending order of their distance metric relative to the first constellation symbol (x c ), and ordering the members of each sub-set of constellation symbols that are defined by the same distance metric according to their Euclidean distances.
1. A method for defining a search sequence for a soft-decision sphere decoding algorithm for use in receivers for quadrature amplitude modulation (QAM) communication signals, comprising:
determining a first member of the search sequence by rounding a received symbol (z) to a first constellation symbol (x c ) of a plurality of constellation symbols (x i ) which form a QAM constellation, said first constellation symbol being the constellation symbol closest to the received symbol (z); and
determining the remaining members of the search sequence by:
classifying, with a processing unit, the remaining constellation symbols of the QAM constellation into a plurality of sub-sets of constellation symbols having the same distance metric relative to said first constellation symbol (x c ) according to a metric d sequ(n) =2a·n=max{|real(x c −x i )|,|imag(x c −x i )|}, wherein a is a non-zero scaling factor of the constellation grid, the constellation symbols being spaced by 2a in both horizontal and vertical directions, and n is a non-zero integer, and
ordering said sub-sets of constellation symbols in ascending order of their distance metric relative to the first constellation symbol (x c ), and
ordering the members of each sub-set of constellation symbols that are defined by the same distance metric according to their Euclidean distances.
2. The method of claim 1 , wherein said ordering step of the members of each subset according to their Euclidean distances comprises ordering the member symbols into sets having the same Euclidean distance relative to the first constellation symbol (x c ), with ascending order of the Euclidean distance of the sets.
3. The method of claim 2 , wherein said ordering step of the members of each subset according to their Euclidean distances comprises ordering the symbols among a set of symbols having the same Euclidean distance relative to the first constellation symbol (x c ) by predefining a succession of these symbols and maintaining said succession for any set throughout the algorithm.