Channel decoding method and tail biting convolutional decoder
The disclosure provides a channel decoding method and a tail biting convolutional decoder. The method comprises: the tail biting convolutional decoder performs add-compare-select processing on metric values of 2 N states according to input data to be decoded and N zeros input after the data to obtain final cumulative metric values of the 2 N states, and stores selection result of add-compare-select processing, wherein N is the constraint length of tail biting convolutional coding; the tail biting convolutional decoder starts to backtrack from one state of the 2 N states according to the selection result to obtain intermediate decoding result after the add-compare-select processing is finished; the tail biting convolutional decoder outputs the bits except the last N bits of the intermediate decoding result as decoded result of the data to be decoded. The tail biting convolutional decoder comprises an add-compare-select module, a backtracking module and an outputting module.
1. A channel decoding method, comprising:
decoding, by a tail biting convolutional decoder, input data to be decoded;
after the decoding of the input data to be decoded is completed, continuing to perform additional M steps of add-compare-select processing by using 0 as a path metric value on metric values of 2 N states to obtain final cumulative metric values of the 2 N states by the tail biting convolutional decoder, wherein M is more than or equal to N, and N is a constraint length of tail biting convolutional coding;
backtracking, by the tail biting convolutional decoder, from any one state of the 2 N states to obtain an intermediate decoding result, after the add-compare-select processing is finished; and
outputting, by the tail biting convolutional decoder, bits except last M bits of the intermediate decoding result as decoded result of the data to be decoded.
2. The method according to claim 1 , wherein obtaining, by the tail biting convolutional decoder, the final cumulative metric values of the 2 N states comprises:
performing, by the tail biting convolutional decoder, the add-compare-select processing on the metric values of the 2 N states according to the input data to be decoded to obtain intermediate cumulative metric values of the 2 N states; and
by taking the intermediate cumulative metric values of the 2 N states as initial values, continuing to perform additional M steps of add-compare-select processing by using 0 as the path metric value on the metric values of the 2 N states to obtain the final cumulative metric values of the 2 N states.
3. The method according to claim 2 , wherein performing the add-compare-select processing comprises:
in the process that the 2 N states transit from a previous moment to a current moment, for each state, respectively accumulating a path metric value to state metric values of two optional states corresponding to the state before the transition, wherein the path metric value is an Euclidean distance calculated by using input data; and
selecting a larger value of accumulating as a new cumulative metric value of a corresponding state after the transition.
4. The method according to claim 1 , wherein performing the add-compare-select processing comprises:
in the process that the 2 N states transit from a previous moment to a current moment, for each state, respectively accumulating a path metric value to state metric values of two optional states corresponding to the state before the transition, wherein the path metric value is an Euclidean distance calculated by using input data; and
selecting a larger value of accumulating as a new cumulative metric value of a corresponding state after the transition.
5. The method according to claim 1 , wherein backtracking, by the tail biting convolutional decoder, from one state of the 2 N states according to the selection result to obtain the intermediate decoding result comprises:
by taking one state of the 2 N states as the starting point of backtracking, searching, by the tail biting convolutional decoder, for a parent state according to the selection result until all history states are found; and
composing, by the tail biting convolutional decoder, lowest bits output by all the history states into the intermediate decoding result, according to a negative direction of backtracking.
6. A tail biting convolutional decoder, comprising:
an add-compare-select module, configured to decode input data to be decoded; and after the decoding of the input data to be decoded is completed, continue to perform additional M steps of add-compare-select processing by using 0 as a path metric value on metric values of 2 N states to obtain final cumulative metric values of the 2 N states, wherein M is more than or equal to N, and N is a constraint length of tail biting convolutional coding;
a backtracking module, configured to backtrack from any one state of the 2 N states to obtain an intermediate decoding result after the add-compare-select module finishes the add-compare-select processing; and
an outputting module, configured to output bits except last M bits of the intermediate decoding result as decoded result of the data to be decoded.
7. The tail biting convolutional decoder according to claim 6 , wherein the add-compare-select module comprises:
a first processing module, configured to perform the add-compare-select processing to the metric values of the 2 N states according to the data to be decoded which is input to the tail biting convolutional decoder to obtain intermediate cumulative metric values of the 2 N states; and
a second processing module, configured to, by taking the intermediate cumulative metric values of the 2 N states as initial values, continue to perform additional M steps of add-compare-select processing by using 0 as the path metric value on the metric values of the 2 N states to obtain the final cumulative metric values of the 2 N states.
8. The tail biting convolutional decoder according to claim 7 , wherein performing the add-compare-select processing by the add-compare-select module comprises:
in the process that the 2 N states transit from a previous moment to a current moment, for each state, respectively accumulating a path metric value to state metric values of two optional states corresponding to the state before the transition, wherein the path metric value is an Euclidean distance calculated by using the input data; and
selecting a larger value of accumulating as a new cumulative metric value of a corresponding state after the transition.
9. The tail biting convolutional decoder according to claim 6 , wherein performing the add-compare-select processing by the add-compare-select module comprises:
in the process that the 2 N states transit from a previous moment to a current moment, for each state, respectively accumulating a path metric value to state metric values of two optional states corresponding to the state before the transition, wherein the path metric value is an Euclidean distance calculated by using the input data; and
selecting a larger value of accumulating as a new cumulative metric value of a corresponding state after the transition.
10. The tail biting convolutional decoder according to claim 6 , wherein the backtracking module comprises:
a searching module, configured to, by taking one state of the 2 N states as the starting point of backtracking, search for a parent state according to the selection result until all history states are found; and
a composing module, configured to compose lowest bits output by all the history states into the intermediate decoding result, according to a negative direction of backtracking.