Development and evaluation of Gradient-Based Numerical Algorithms for object localization in multistatic passive radar system

Abstract

Passive radar does not transmit its own signals but
relies on external sources of illumination. Estimating a target
position requires distance measurements from at least three
spatially separated sensors. This paper presents an adaptation of
a maximum likelihood estimator (MLE) for target localization
using multiple (three or more) bistatic range or DToA (Difference
Time of Arrival) measurements. The Gauss–Newton, Levenberg,
and Levenberg–Marquardt methods are compared with respect
to numerical stability and localization accuracy under different
noise levels and initialization conditions.