Soft-output Viterbi decoding: VLSI implementation issues

The soft-output Viterbi algorithm (SOVA) is a decoding algorithm which uses soft quantized inputs and delivers soft decision outputs. Implementation parameter effects in SOVA decoding that are related to considerable variations in area consumption of VLSI implementations are considered, i.e., the quantization of the reliability values inside the survivor memory unit, the depth of reliability updating, and the effect of a simplified update rule on the reliability values. Results of extensive simulations are presented. Area estimates obtained by logic synthesis from VHSIC hardware description language (VHDL) descriptions are given to show how these parameters translate into the area consumption of VLSI implementations.