Abstract
This paper focuses on the problem of multistatic sonar localization with a transmitter where the known signal transmission speed, the transmitter and receiver positions are all contaminated by Gaussian noise. The Cramer-Rao lower bound (CRLB) are derived for the object position and the localization parameters. The analysis in CRLB shows that the localization parameters have a chance to be updated to more accurate ones. Two solutions are then proposed to estimate the object position using the time measurements and angle measurements. One is a two-step closed-form solution based on weighting least squares and the other is a generalized trust region subproblem (GTRS) solution using Newton’s method. A recursive MLE is also proposed to update the localization parameters and a more accurate propagation speed can be obtained from the proposed MLE especially when the propagation speed noise is large. Simulations show that the two localization solutions and the propagation speed updated from the proposed MLE can reach their CRLBs.
Highlights
The problem of localization of an object, whether active or passive, has been an interest basis research due to its various applications in many areas, including wireless sensor networks (WSN), Internet of Thing (IoT), radar and sonar, etc [1]–[8]
Most of the research focused on passive localization which are usually based on time difference of arrival (TDOA), angle of arrival (AOA), or frequency difference of arrival (FDOA), or their combinations [9]–[13]
We focus on multistatic sonar localization with a transmitter
Summary
The problem of localization of an object, whether active or passive, has been an interest basis research due to its various applications in many areas, including wireless sensor networks (WSN), Internet of Thing (IoT), radar and sonar, etc [1]–[8]. In [22], a three-step method based on WLS was developed to jointly estimate the position and the unknown signal propagation speed. These methods are simple and computationally efficient and they can reach the Cramér-Rao lower bound (CRLB) in the small error region. The authors in [27] proposed efficient closed-form solutions for multistatic sonar localization where the sound speed is supposed to be a Gaussian random variable. Considering the complexity of the underwater environment where the propagation speed together with sensor positions are all time-variant, in this paper, they are supposed to be unknown constants with respect to time during a short measurement period.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
