作者: A. Averbuch , S. Itzikowitz , T. Kapon
DOI: 10.1109/71.89069
关键词:
摘要: The implementations of the Viterbi algorithm (VA) and interacting multiple model (IMM) on a shared-bus shared-memory multiple-input multiple-data (MIMD) multiprocessor are discussed. computational complexity as well speedup efficiency examined in detail. It is shown that parallel implementation these algorithms about same both memory space processing time categories. Efficiency with P processors 1-1/P for small expected to be relatively high large P, especially when many filters state measurement vectors considered. >