TDOA approach for target localization based on improved genetic algorithm

Probabilistic 5G Indoor Positioning Proof of Concept with Outlier Rejection

“…As done e. g., in [8], an estimate of the UE's position x can be computed by solving the linear equation system…”

“…Moreover, the performance of maximum likelihood (ML) algorithms can be enhanced by removing outliers, created e. g., due to the already mentioned multipath propagation, and by weighting the remaining measurements based on the confidence in them. In literature, various approaches based on time difference of arrival (TDoA) exist that could be employed to determine an initial search point, e. g., [8]- [11]. However, [8]- [10] require the definition of a reference locator, which has the drawback that the reference itself could exhibit an outlier measurement.…”

“…In literature, various approaches based on time difference of arrival (TDoA) exist that could be employed to determine an initial search point, e. g., [8]- [11]. However, [8]- [10] require the definition of a reference locator, which has the drawback that the reference itself could exhibit an outlier measurement. On the other hand, [11] avoids the definition of a reference locator, but can not detect outliers and does not provide the possibility of assigning weights to the measurements.…”