Soft Bit Viterbi Equalizer Using Partially Collapsed Metrics

Abstract

A system and method of decoding data are provided. A particular method includes determining a first partially collapsed metric value for each bit in a symbol transition, where each first partially collapsed metric value is a cumulative metric of starting in a first state and ending in a second state in L transitions. The first partially collapsed metric value is equal to the cumulative metric value for each of the states plus first minimum sum of branch metric values.

1 . A method of decoding data transmitted via a communication channel, the method comprising:
 determining a first partially collapsed metric value for each bit in a symbol transition, each first partially collapsed metric value being a cumulative metric of starting in a first state and ending in a second state in L transitions, wherein the first partially collapsed metric value is equal to the cumulative metric value for each of the states plus first minimum sum of branch metric values.   
   
   
       2 . The method of  claim 1 , further comprising determining a second partially collapsed metric value for each bit in the symbol transition. 
   
   
       3 . The method of  claim 2 , further comprising for each bit in the symbol transition:
 determining a relative likelihood value based on its respective first partially collapsed metric value and its respective second partially collapsed metric value; and   decoding at least one symbol based on a hard decision performed using the relative likelihood value for each bit in the symbol transition.   
   
   
       4 . The method of  claim 2 , wherein each second partially collapsed metric value is a cumulative metric of starting in the first state and ending in the second state in L transitions with the respective bit in the symbol transition being a second value, the second value differing from a first value. 
   
   
       5 . The method of  claim 4 , wherein with the respective bit in the symbol transition being the first value, the first state and the second state are internal states of a Viterbi algorithm. 
   
   
       6 . The method of  claim 1 , wherein each relative likelihood value is a log likelihood ratio based on its respective first partially collapsed metric value and its respective second partially collapsed metric value. 
   
   
       7 . The method of  claim 1 , wherein the first minimum sum of branch metric values are over all state transitions and parallel transitions in L symbol transitions in which the respective bit in the symbol transition is a first value. 
   
   
       8 . The method of  claim 7 , wherein determining a second partially collapsed metric value is based on the cumulative metric value and second minimum sum of branch metric values over all state transitions and parallel transitions in L symbol transitions in which the respective bit in the symbol transition is a second value. 
   
   
       9 . The method of  claim 1 , further comprising determining another cumulative metric value based on the first and second partially collapsed metric values. 
   
   
       10 . A receiver to decode data transmitted via a communication channel, the receiver comprising:
 a soft bit equalizer configured to:
 determine a first partially collapsed metric value for each bit in a symbol transition, each first partially collapsed metric value being a cumulative metric of starting in a first state and ending in a second state in L transitions, wherein the first partially collapsed metric value is equal to the cumulative metric value for each of the states plus first minimum sum of branch metric values; and 
 determine a second partially collapsed metric value for each bit in the symbol transition. 
   
   
   
       11 . The receiver of  claim 10 , wherein the first minimum sum of branch metric values are over all state transitions and parallel transitions in L symbol transitions in which the respective bit in the symbol transition is the first value. 
   
   
       12 . The receiver of  claim 10 , wherein the soft bit equalizer is further configured to determine another cumulative metric value based on the first and second partially collapsed metric values. 
   
   
       13 . The receiver of  claim 10 , wherein each second partially collapsed metric value is a cumulative metric of starting in the first state and ending in the second state in L transitions with the respective bit in the symbol transition being a second value, the second value differing from the first value. 
   
   
       14 . The receiver of  claim 13 , wherein with the respective bit in the symbol transition being a first value, the first state and the second state are internal states of a Viterbi algorithm. 
   
   
       15 . The receiver of  claim 10 , wherein the soft bit equalizer is further configured to, for each bit in the symbol transition:
 determine a relative likelihood value based on its respective first partially collapsed metric value and its respective second partially collapsed metric value; and   decode at least one symbol based on a hard decision performed using the relative likelihood value for each bit in the symbol transition.   
   
   
       16 . The receiver of  claim 15 , wherein each relative likelihood value is a log likelihood ratio based on its respective first partially collapsed metric value and its respective second partially collapsed metric value. 
   
   
       17 . The receiver of  claim 15 , wherein the soft bit equalizer is further configured to determine the second partially collapsed metric value based on the cumulative metric value and a second minimum sum of branch metric values over all state transitions and parallel transitions in L symbol transitions in which the respective bit in the symbol transition is the second value. 
   
   
       18 . A computer-readable medium having computer readable program code to cause a computer processor to perform acts comprising:
 determining a first partially collapsed metric value for each bit in a symbol transition, each first partially collapsed metric value being a cumulative metric of starting in a first state and ending in a second state in L transitions, wherein the first partially collapsed metric value is equal to the cumulative metric value for each of the states plus first minimum sum of branch metric values; and   determining a second partially collapsed metric value for each bit in the symbol transition, wherein each second partially collapsed metric value is a cumulative metric of starting in the first state and ending in the second state in L transitions with the respective bit in the symbol transition being a second value, the second value differing from a first value.   
   
   
       19 . The computer readable medium of  claim 18 , wherein the acts further comprise, for each bit in the symbol transition:
 determining a relative likelihood value based on its respective first partially collapsed metric value and its respective second partially collapsed metric value; and   decoding at least one symbol based on a hard decision performed using the relative likelihood value for each bit in the symbol transition.   
   
   
       20 . The computer readable medium of  claim 18 , wherein with the respective bit in the symbol transition being the first value, the first state and the second state are internal states of a Viterbi algorithm.