The purpose of this work is to establish how a moving emitter can be localized by a passive receiver through the use of out-of-plane multipath signals reflected by the terrain. This is a novel localization technique that assumes no a priori knowledge of the location of the multipath sources. The emitter parameters of range, heading, velocity, and altitude are estimated by exploiting the correlation between the direct-path signal and the delayed, attenuated, and Doppler modulated signals reflected by the terrain. Two basic assumptions about the terrain scattering properties lead to different maximum likelihood (ML) estimators of emitter parameters. The first assumption is that the terrain scattering is fundamentally homogeneous; in this case the ML estimator is found to have the structure of a timevarying FIR filter. The second assumption is that the terrain scattering is fundamentally inhomogeneous and dominated by a number of discrete scatterers. This assumption leads to a two-part estimator which first estimates the scattering parameters of azimuth, differential delay, and Doppler shift, and then estimates the emitter parameters using the scatterer parameter estimates. The Cramer-Rao lower bounds of the variances for each estimator are derived and used to study estimator performance for several scenarios. The proposed estimators are successfully demonstrated using field data collected at White Sands Missile Range during the DARPA/Navy Mountaintop program. Several extensions to the basic results are considered such as localizing pulsed and self-correlated emitters, multiple emitters, and the effects of receiver motion. in PREFACE The material presented in this report is identical to that in a dissertation submitted to Northeastern University, Department of Electrical and Computer Engineering, August 1997, in partial fulfillment of the requirements for the degree of Doctor of Philosophy. ACKNOWLEDGMENTS I would like to thank many individuals, at both Northeastern and Lincoln Laboratory, for their contributions toward the successful completion of this work. I would like to thank Mr. Gerard Titi first for suggesting this fruitful research topic and also, along with Dr. Steven Krich, for providing encouragement, and support including the computer resources that permitted me to work effectively at home while maintaining a full-time position at Lincoln Laboratory. I also wish to thank Dr. Kenneth Senne for many helpful suggestions during the preparation of my ASAP presentations on this material. I am deeply indebted to Prof. Ram Raghavan, my principal advisor, for many stimulating technical discussions and invaluable guidance on completing and writing up this work. Thanks also to committee members Dr.