Using quadrant shifting to facilitate binary arithmetic with two's complement operands

A maximum a posteriori decoder, comprising:

• an alpha block for producing alpha state metrics in two's complement format;
• a beta block for producing beta state metrics in two's complement format;
• an extrinsic block having an input coupled to said alpha block and said beta block for receiving said alpha state metrics and said beta state metrics as operands, said extrinsic block responsive to said operands for producing extrinsics data;
• and
• said extrinsic block including logic coupled to said input for determining, for each operand, whether an original value is within a predetermined proximity of a maximum positive/maximum negative value boundary associated with the two's complement format, and an adjuster coupled to said logic and responsive to a determination by said logic that any of the original operand values is within the predetermined proximity for adjusting all of the original operand values such that cross over from the maximum positive value to the maximum negative value is avoided and to produce respectively corresponding adjusted operand values for use in producing the extrinsics data.