Complexity Reduction of Turbo Decoding Based on Decision Thresholds

In this paper we propose a method for reducing turbo decoding complexity. Based on decision thresholds, some calculations are previously considered negligible, and it is pos- sible to reduce the number of calculations required to perform turbo decoding. Simulations show that almost 40% of the state metrics (α and β) and 25% of the LLRs (Log-Likelihood Ratios) calculations can be eliminated. With the same idea, decision thresholds can be used to take decisions on the information bits based on the LLRs, achieving good decoding complexity reduction, especially for high SNR (Signal to Noise Ratio).