Enhanced positioning system using hybrid filter

摘要

The disclosure generally relates to an enhanced positioning system and method using a combination or hybrid filter. In one embodiment, Time-Of-Flight (ToF) measurements are used to determine an approximate location for a mobile device in relationship to one or more Access Points. The ToF combined with known and unknown variables are then processed through a hybrid filter system to determine location of the mobile device. The hybrid filter system may include a Kalman Filter (KF) for processing linear models and generally Gaussian noise distribution. The KF assumes that the state probability of mobile device location is Gaussian. Such variables include, for example, WiFi ToF bias. The hybrid filter system may include a Bayesian Filter (BF) for processing variables having non-Gaussian noise distribution and non-linear models. Such variables include, for example, the coordinates of the mobile device. A probability determination from each of the KF and BF is then applied to estimate the mobile device location.

主权项

1. Processor circuitry for determining the location of a mobile device in an enclosed area, the processor circuitry programmed with logic to implement functions, comprising:
identifying a plurality of locations (Xn, Yn) in the area and assigning each location an initial Bayesian Filter (BF) probability value (Pr) for presence of the mobile device; assigning an initial Kalman filter (KF) probability value for presence of the mobile device; determining the probability that the mobile device is at a first location Pr(Xn/Yn−1) by summing the probabilities that the mobile device is located at a neighboring location and multiplying the summed probabilities by a transition probability; determining a weight value Pn as a function of the probability that the mobile device is at the first location Pr(Xn/Yn−1) and the transition probability; calculating a mean value and a variance value as a function of the weight value Pn for a respective location; calculating an updated KF value as a function of a bias value, the mean value and the variance value; and calculating an updated BF value as a function of the bias value; and updating the location of the mobile device as a function of the updated BF and the updated KF values.