| 摘要 |
A target device's location is estimated by a location estimation module (LEM) that comprises a probabilistic model (PM) for a plurality of sample points, each of which comprises a sample location and an expected distribution of signal values at that sample point. The location estimation module (LEM) makes a sequence (OS) of observations o<SUB>n</SUB>, n=1, 2, 3 . . . , of signal values. Each observation corresponds to a respective location q<SUB>n </SUB>along the target device's path. The sequence of observations and the respective location constitute a hidden Markov model. The module estimates the target device's location q<SUB>n </SUB>based on the probabilistic model (PM) and the sequence of observations, wherein the sequence of observations comprises one or more future observations o<SUB>n+m </SUB>for which m is a positive integer. In other words, the target device's location is estimated, at least partially, based on knowledge of its future.
|
