Figure 2.
Soft decoding algorithms evaluated in this work. (A) The trellis architecture and connections of states used by the Beam Trellis algorithm (Chandak et al.). (B) Edge scoring mechanism of the Beam Trellis algorithm to account for CTC-encodings of candidate messages. (C) State merging required by the Beam Trellis algorithm to account for duplicate state messages. (D) Architecture and scoring methodology used by our novel alignment matrix algorithm. Arrows in the alignment matrix represent data used to calculate newly added rows for both the new blank and base symbols.

Soft decoding algorithms evaluated in this work. (A) The trellis architecture and connections of states used by the Beam Trellis algorithm (Chandak et al.). (B) Edge scoring mechanism of the Beam Trellis algorithm to account for CTC-encodings of candidate messages. (C) State merging required by the Beam Trellis algorithm to account for duplicate state messages. (D) Architecture and scoring methodology used by our novel alignment matrix algorithm. Arrows in the alignment matrix represent data used to calculate newly added rows for both the new blank and base symbols.

Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close