Existing multistatic positioning techniques can attain optimal performance under weak noise conditions. However, the localization models of these algorithms are designed solely to estimate the target position and velocity, and they are not able to determine the target acceleration in some general localization situations (such as the detection and localization of maneuvering targets). To address the multistatic radar localization problem for unknown object position, velocity, and acceleration estimates with sensor parameter uncertainties, we develop two efficient estimators using bistatic range (BR), BR rate (BRR) and bistatic acceleration (BA) measurements. One method estimates the target parameters in closed form by introducing extra variables and using the weighted least squares (WLS) technique. The other method employs the semidefinite relaxation (SDR) approach to recast the localization problem as a semidefinite programming (SDP) problem, which is solved using off-the-shelf toolboxes. The former method offers appealing computational efficiency, whereas the latter has high noise tolerance. Theoretical investigations indicate that the estimation accuracy of the target position and velocity is improved by introducing the BA measurements and the unknown acceleration variable. In addition, these two methods are proven in both theory and simulations to attain the Cramer–Rao lower bound (CRLB) under small noise and error conditions. Furthermore, simulations demonstrate that the proposed methods outperform the current algorithms in terms of position and velocity estimation.
Efficient Multistatic Radar Localization Algorithms for a Uniformly Accelerated Moving Object With Sensor Parameter Errors
IEEE Transactions on Aerospace and Electronic Systems ; 59 , 6 ; 7559-7574
2023-12-01
2947937 byte
Article (Journal)
Electronic Resource
English
Multistatic Radar Imaging of Moving Targets
IEEE | 2012
|British Library Conference Proceedings | 2010