Аннотация:We comment on partial parallel interference cancellation as discussed in the paper by Divsalar et al. (see ibid. vol.46, p.258-68, 1998). The aforementioned work showed that by multiplying the symbol estimates by a factor less than unity in the early stages of cancellation, the performance of parallel cancellation can be improved relative to full ("brute force") cancellation. In this paper we analyze the improvement of parallel cancellation when using partial cancellation, and provide additional insight into the gains. Specifically, we show that the decision statistic is biased when linear (soft) estimates of the symbol or channel are used for cancellation. Partial cancellation improves the performance in this case by reducing the decision statistic bias.
