Semester of Graduation

Spring 2018

Degree

Master of Science in Electrical Engineering (MSEE)

Department

Electrical Engineering Department

Document Type

Thesis

Abstract

Bistatic passive radar (BPR) system does not transmit any electromagnetic signal unlike the active radar, but employs an existing Illuminator of opportunity (IO) in the environment, for instance, a broadcast station, to detect and track the targets of interest. Therefore, a BPR system is comprised of two channels. One is the reference channel that collects only the IO signal, and the other is the surveillance channel which is used to capture the targets' reflected signals. When the IO signal reflected from multiple targets is captured in the surveillance channel (SC) then estimating the delays and Doppler shifts of all the observed targets is a challenging problem. For BPR system, the signal processing algorithms developed so far models the IO waveform as a deterministic process and discretizes the delays and Doppler shifts parameters.

In this thesis, we deal with the problem of jointly estimating the delays and Doppler shifts of multiple targets in a BPR system (i.e., a two channel system) when the unknown IO signal is modeled as a correlated stochastic process. Unlike the previous work, we take all the delays and Doppler shifts as continuous-valued parameters to avoid straddle loss due to discretization and propose a computationally efficient Expectation-Maximization (EM) based algorithm that breaks up the complex multidimensional maximum likelihood optimization problem into multiple separate optimization problems. The EM algorithm jointly provides the estimates of all the delays and Doppler shifts of the targets along with the estimate of each target's component signal in the SC and the estimate of the unknown IO signal. We also derive the Cramer-Rao lower bound for the considered multitarget estimation problem with stochastic IO signal. Numerical simulations are presented where we compare our proposed EM-based multi-target estimator with the widely used conventional cross correlation estimator under different multitarget environments.

Date

4-3-2018

Committee Chair

Naraghi-Pour, Morteza

DOI

10.31390/gradschool_theses.4678

Download

Share

COinS