Methods for decoding data in digital communication systems
AssigneeMotorola, Inc.
InventorsDana J. Taipale, Paula F. Such
Life of this patent
Estimating the term…
Loading patent…
AssigneeMotorola, Inc.
InventorsDana J. Taipale, Paula F. Such
Estimating the term…
The present invention includes methods relating generally to communication systems. In decoding a sequence of symbols output by an encoder having a first constituent encoder and a second constituent encoder, where the output of these encoders correspond to unterminated trellises, a forward calculation is performed along each trellis to compute a forward state metric value for a first node of each trellis. Prior to performing a backward calculation along each trellis, a backward state metric value of a second node of each trellis is initialized to the forward state metric values of the corresponding first nodes of each trellis. Using these forward and backward state metric values, the sequence of symbols is decoded. Other embodiments include decoding terminated trellises in a similar manner, or dividing each trellis into smaller windows and processing the series of windows.
A method for decoding a sequence of symbols output by an encoder having a first constituent encoder and a second constituent encoder, wherein the output of the first constituent encoder corresponds to a first unterminated trellis and the output of the second constituent encoder corresponds to a second unterminated trellis, comprising:
A method for decoding a sequence of symbols output by an encoder and transmitted over a channel, wherein the output of the channel corresponds to a trellis having a trellis length, a first window, and second window, each window having a plurality of nodes and a window length that is less than the trellis length, comprising:
A method for decoding a sequence of symbols output by an encoder and transmitted over a channel, wherein the output of the channel is represented by a trellis having a plurality of nodes, including an ending node, comprising:
Dependencies are inferred from the claim text.