Detector with soft pruning
Various embodiments of the present invention provide apparatuses, systems and methods for data detection in a detector with soft pruning. For example, a data detector is disclosed that includes a branch metric calculator operable to calculate branch metrics for transitions between states in a trellis for the data detector, and a branch metric offset circuit operable to apply branch metric offsets to the branch metrics to yield soft pruned branch metrics. The branch metric offsets comprise a range of probability values from zero percent to one hundred percent.
1. A data detector comprising:
a branch metric calculator operable to calculate branch metrics for transitions between states in a trellis for the data detector; and
a branch metric offset circuit operable to apply branch metric offsets to the branch metrics to yield soft pruned branch metrics, wherein the branch metric offsets comprise a range of probability values from zero percent to one hundred percent, wherein a first group of the states originate at least one transition prohibited by the branch metric offsets and wherein a second group of the states do not originate transitions prohibited by the branch metric offsets, and wherein the branch metric offset circuit is operable to apply the branch metric offsets to the branch metrics for the transitions from the first group of states and to the branch metrics for the transitions from the second group of states.
2. The data detector of claim 1 , wherein the branch metric offsets are derived from conditional state transition probabilities for encoded data at an input of the data detector based on at least one constraint on the encoded data.
3. The data detector of claim 2 , wherein the branch metric offset circuit is operable to apply the branch metric offsets by adding them to the branch metrics in a log domain.
4. The data detector of claim 2 , wherein the branch metric offset circuit is operable to disable soft pruning by not applying the branch metric offsets when the at least one constraint is violated.
5. The data detector of claim 1 , wherein the branch metric offset circuit is operable to periodically switch between applying the branch metric offsets and applying a second group of branch metric offsets.
6. The data detector of claim 1 , wherein the data detector comprises a maximum a posteriori detector, and wherein the branch metric offset circuit is operable to apply the branch metric offsets in a forward detection operation and in a reverse detection operation.
7. The data detector of claim 1 , wherein the branch metric calculator and the branch metric offset circuit are implemented as an integrated circuit.
8. The data detector of claim 1 , wherein the branch metric calculator and the branch metric offset circuit are incorporated in a storage device.
9. The data detector of claim 1 , wherein the branch metric calculator and the branch metric offset circuit are incorporated in a transmission device.
10. A method of detecting data in a data detector with soft pruning, comprising:
calculating a branch metric for each of a plurality of transitions between states in the data detector;
deriving branch metric offsets from probabilities of data patterns at an input to the data detector, wherein the branch metric offsets are derived by counting occurrences of each possible pattern in a sliding window applied to a data stream at an output of an encoder upstream of the data detector to yield probabilities of occurrence;
applying at least one of the branch metrics offset to each of the branch metrics to yield soft pruned branch metrics; and
calculating hard decisions based on the soft pruned branch metrics.
11. The method of claim 10 , wherein the branch metric offsets are applied to all transitions in the data detector.
12. The method of claim 10 , wherein the branch metric offsets are further derived by grouping the possible patterns and normalizing the probabilities of occurrence by group.
13. The method of claim 10 , further comprising disabling the application of branch metric offsets when data at an input of the data detector violates constraints upon which the branch metric offsets are based.
14. The method of claim 10 , wherein the branch metric offsets lie in a probability range between and including zero percent and one hundred percent.
15. A storage system comprising:
a storage medium maintaining a data set;
a read/write head assembly operable to write the data set to the storage medium and to read the data set from the storage medium;
a maximum transition run encoder operable to encode the data set to yield encoded data before it is written to the storage medium and to limit a number of successive transitions in the encoded data; and
a data detector operable to detect values of the encoded data using branch metrics for transitions between states in a trellis and to apply branch metric offsets to all the transitions, wherein the branch metric offsets are based on probabilities of occurrence in data patterns in the encoded data, and wherein the data detector is operable to stop applying the branch metric offsets when data at an input of the data detector violates the limit on the number of successive transitions.
16. The storage system of claim 15 , wherein the probabilities of occurrence lie in a probability range between and including zero percent and one hundred percent.
17. The method of claim 10 , further comprising periodically switching between applying the branch metric offsets and applying a second group of branch metric offsets to each of the branch metrics to yield the soft pruned branch metrics.
18. The method of claim 10 , wherein the branch metric offsets are derived from conditional state transition probabilities for encoded data at an input of the data detector based on at least one constraint on the encoded data.
19. The storage system of claim 15 , wherein the data detector comprises a maximum a posteriori detector operable to apply the branch metric offsets in a forward detection operation and in a reverse detection operation.
20. The storage system of claim 15 , wherein the data detector is operable to apply the branch metric offsets by adding them to the branch metrics in a log domain.