Accelerated method of detection by spheres
AssigneeMitsubishi Denki Kabushiki Kaisha
InventorLoic Brunel
Life of this patent
Estimating the term…
Loading patent…
AssigneeMitsubishi Denki Kabushiki Kaisha
InventorLoic Brunel
Estimating the term…
A method of detecting a plurality of symbols (dk(i)) transmitted by or for a plurality K of users, each symbol belonging to a modulation constellation and being the subject of a spectral spreading by means of a spreading sequence, the method including a filtering step adapted for supplying a real vector (z) characteristic of the received signal, at least the closest neighbor of the vector being sought within a lattice of points (Ξ) generated by the modulation constellations, the search being limited to candidate vectors (x), each of whose components (bi) has a value lying within a search interval defined by a lower bound (Bi−)and an upper bound (Bi+), the bounds being chosen so that each of the intervals comprises only component values of points situated within a sphere with a predetermined radius (√{square root over (C)}) and belonging to a modulation constellation.
Method of detecting a plurality of symbols (dk(i)) transmitted by or for a plurality K of users, each symbol belonging to a modulation constellation and being the subject of a spectral spreading by means of a spreading sequence, the method comprising a filtering step adapted for supplying a real vector (z) characteristic of a received signal, wherein at least the closest neighbour of the vector is sought within a lattice of points (Ξ) generated by said modulation constellations, the search being limited to candidate vectors (x), each of whose components (bi) has a value lying within a search interval defined by a lower bound (Bi−) and an upper bound (Bi+), the bounds being chosen so that each of the intervals comprises only component values of points situated within a sphere with a predetermined radius (√{square root over (C)}) and belonging to a modulation constellation.
Detection method according to any of the preceding claims, wherein the search for the closest neighbour is effected by scanning one by one the search intervals for the different component indices (i) and choosing a component value (bi) in each of the intervals, the bounds of each interval being determined according to the width of the interval with the preceding index (i+1) and the component value (bi+1) chosen in this same interval.
Dependencies are inferred from the claim text.