On the parallel MAP algorithm

摘要

The MAP algorithm used in turbo decoders is very parallelizable.This property can be used to speed them up. The speedup is attained atthe expense of increased complexity and processing hardware. Theincrease in processing hardware is linear with the speedup but theincrease in complexity of the parallel implementation compared to thetraditional implementation is extremely small and this is the mainobservation of this paper. We discuss the parallel implementation of theMAP algorithm, the increase in complexity and memory and contrast thebehaviour with that of the Viterbi algorithm. The results given herealso suggest a change in design methodology of the sliding windowimplementation of turbo decoders