Metric computation for lowering complexity of MIMO detection algorithms
View Patent ↗Included are embodiments of method for method for computing metrics. At least one embodiment includes searching a MIMO detection tree, the detection tree configuration being formed by a plurality of nodes and a plurality of leaves connected via a plurality of branches, the computational complexity associated with computing a node metric decreases with the node depth in the tree configuration and providing an estimate on a transmitted signal.
1. A method for computing metrics, comprising:
searching a MIMO detection tree, the detection tree configuration being formed by a plurality of nodes and a plurality of leaves connected via a plurality of branches, the searching including updating node metrics such that computational complexity associated with each node decreases with the node depth in the tree configuration; and
providing an estimate on a transmitted signal.
2. The method of claim 1 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a hard estimate on the transmitted signal, the hard estimate indicating a signal associated with the leaf with the smallest metric in the tree configuration.
3. The method of claim 1 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a soft estimate on the transmitted signal, the soft estimate indicating one or more bitwise soft-reliability information on the bits mapped to the transmitted signal, the bitwise soft-reliability information including a scaled difference of soft-bit metrics, the soft-bit metric indicating a computed node metric.
4. The method of claim 3 , wherein computation of a soft estimate is based on:
one or more leaf metrics; and
one or more metrics associated with nodes pruned at levels with an index smaller than a level threshold (N′), where N′ is smaller or equal to the total number N of levels in the tree configuration.
5. The method of claim 3 , further comprising saturating the at least one soft-bit metric to a saturation threshold R, wherein saturating the at least one soft-bit metric includes determining a value R of the saturation threshold.
6. The method of claim 4 , further comprising optimizing the level threshold (N′) under which the pruned nodes are utilized in the computation of the soft-output, wherein optimizing includes at least one of the following:
determining N′=1 for a 2×2 MIMO system utilizing Quadrature phase shift keying (QPSK) modulation;
determining N′=1 for a 2×2 MIMO system utilizing 16-signal quadrature amplitude modulation (QAM);
determining N′=3 for a 2×2 MIMO system utilizing 64-signal quadrature amplitude modulation;
determining N′=1 for a 4×4 MIMO system utilizing Quadrature phase shift keying modulation;
determining N′=1 for a 4×4 MIMO system utilizing 16-signal quadrature amplitude modulation; and
determining N′=1 for a 4×4 MIMO system utilizing 64-signal quadrature amplitude modulation.
7. The method of claim 5 , further comprising optimizing a value of the saturation threshold (R), wherein optimizing includes at least one of the following:
determining R=5.0 for a 2×2 MIMO system utilizing Quadrature phase shift keying (QPSK) modulation;
determining R=0.75 for a 2×2 MIMO system utilizing 16-signal quadrature amplitude modulation (QAM);
determining R=0.15 for a 2×2 MIMO system utilizing 64-signal quadrature amplitude modulation;
determining R=4.0 for a 4×4 MIMO system utilizing Quadrature phase shift keying modulation;
determining R=0.65 for a 4×4 MIMO system utilizing 16-signal quadrature amplitude modulation; and
determining R=0.1 for a 4×4 MIMO system utilizing 64-signal quadrature amplitude modulation.
8. A system for computing metrics, comprising:
a searching component configured to search a MIMO detection tree, the detection tree configuration being formed by a plurality of nodes and a plurality of leaves connected via a plurality of branches, the searching component including a node metrics updating component configured such that computational complexity associated with each node decreases with the node depth in the tree configuration; and
a providing component configured to provide an estimate on a transmitted signal.
9. The system of claim 8 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a hard estimate on the transmitted signal, the hard estimate indicating a signal associated with the leaf with the smallest metric in the tree configuration.
10. The system of claim 8 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a soft estimate on the transmitted signal, the soft estimate indicating one or more bitwise soft-reliability information on the bits mapped to the transmitted signal, the bitwise soft-reliability information including a scaled difference of soft-bit metrics, the soft-bit metric indicating a computed node metric.
11. The system of claim 10 , wherein computation of a soft estimate is based on:
one or more leaf metrics; and
one or more metrics associated with nodes pruned at levels with an index smaller than a level threshold (N′), where N′ is smaller or equal to the total number N of levels in the tree configuration.
12. The system of claim 10 , further comprising a saturating component configured to saturate the at least one soft-bit metric to a saturation threshold R, wherein saturating the at least one soft-bit metric includes determining a value R of the saturation threshold.
13. The system of claim 11 , further comprising an optimizing component configured to optimize the level threshold (N′) under which the pruned nodes are utilized in the computation of the soft-output, wherein optimizing includes at least one of the following:
determining N′=1 for a 2×2 MIMO system utilizing Quadrature phase shift keying (QPSK) modulation;
determining N′=1 for a 2×2 MIMO system utilizing 16-signal quadrature amplitude modulation (QAM);
determining N′=3 for a 2×2 MIMO system utilizing 64-signal quadrature amplitude modulation;
determining N′=1 for a 4×4 MIMO system utilizing Quadrature phase shift keying modulation;
determining N′=1 for a 4×4 MIMO system utilizing 16-signal quadrature amplitude modulation; and
determining N′=1 for a 4×4 MIMO system utilizing 64-signal quadrature amplitude modulation.
14. The system of claim 12 , further comprising an optimizing component configured to optimize a value of the saturation threshold (R), wherein optimizing includes at least one of the following:
determining R=5.0 for a 2×2 MIMO system utilizing Quadrature phase shift keying (QPSK) modulation;
determining R=0.75 for a 2×2 MIMO system utilizing 16-signal quadrature amplitude modulation (QAM);
determining R=0.15 for a 2×2 MIMO system utilizing 64-signal quadrature amplitude modulation;
determining R=4.0 for a 4×4 MIMO system utilizing Quadrature phase shift keying modulation;
determining R=0.65 for a 4×4 MIMO system utilizing 16-signal quadrature amplitude modulation; and
determining R=0.1 for a 4×4 MIMO system utilizing 64-signal quadrature amplitude modulation.
15. A system for computing metrics, comprising:
means for searching a MIMO detection tree, the detection tree configuration being formed by a plurality of nodes and a plurality of leaves connected via a plurality of branches, the means for searching including means for updating node metrics such that computational complexity associated with each node decreases with the node depth in the tree configuration; and
means for providing an estimate on a transmitted signal.
16. The system of claim 15 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a hard estimate on the transmitted signal, the hard estimate indicating a signal associated with the leaf with the smallest metric in the tree configuration.
17. The system of claim 15 , wherein the tree configuration is searched in a K-best manner, and wherein providing the estimate includes providing a soft estimate on the transmitted signal, the soft estimate indicating one or more bitwise soft-reliability information on the bits mapped to the transmitted signal, the bitwise soft-reliability information including a scaled difference of soft-bit metrics, the soft-bit metric indicating a computed node metric.
18. The system of claim 17 , wherein computation of a soft estimate is based on:
one or more leaf metrics; and
one or more metrics associated with nodes pruned at levels with an index smaller than a level threshold (N′), where N′ is smaller or equal to the total number N of levels in the tree configuration.
19. The system of claim 17 , further comprising means for saturating the at least one soft-bit metric to a saturation threshold R, wherein saturating the at least one soft-bit metric includes determining a value R of the saturation threshold.
20. The system of claim 19 , further comprising means for optimizing a value of the saturation threshold (R), wherein optimizing includes at least one of the following:
determining R=5.0 for a 2×2 MIMO system utilizing Quadrature phase shift keying (QPSK) modulation;
determining R=0.75 for a 2×2 MIMO system utilizing 16-signal quadrature amplitude modulation (QAM);
determining R=0.15 for a 2×2 MIMO system utilizing 64-signal quadrature amplitude modulation;
determining R=4.0 for a 4×4 MIMO system utilizing Quadrature phase shift keying modulation;
determining R=0.65 for a 4×4 MIMO system utilizing 16-signal quadrature amplitude modulation; and
determining R=0.1 for a 4×4 MIMO system utilizing 64-signal quadrature amplitude modulation.