SOURCE LOCALIZATION USING AIRBORNE VECTOR SENSORS Hasan S. Mir and John D. Sahr University of Washington, Seattle {hmir, jdsahr}@ee.washington.edu Catherine M. Keller MIT Lincoln Laboratory keller@ll.mit.edu ABSTRACT The direction finding (DF) performance of an array of vec- tor sensor antennas on a small aircraft is assessed. DF performance in the presence of additive noise, array mani- fold perturbations, and uncertainty of polarization param- eters is examined with simulated data and compared to the Cramer-Rao lower bound. DF performance improvements are demonstrated with a least-squares type calibration tech- nique with and without the use of synthetic antenna ele- ments. 1. INTRODUCTION A key application of sensor arrays is the passive localization of a radiating signal source. The sensor-to-sensor delays contain information about the source location in terms of the source azimuth angle θ and source elevation angle φ for two dimensional array configurations. This information is exploited in direction finding (DF) algorithms that estimate the source angle-of-arrival (AOA). DF algorithms utilize an underlying model which presumes a coherent phase rela- tionship among the antenna array elements. Such a relation- ship almost never occurs in practice due to various antenna effects such as antenna pattern differences and antenna-to- receiver electrical cable length differences. Thus, array cal- ibration, which attempts to fit the actual array response to the theoretical response, is essential for obtaining accurate DF estimates. The angular resolution of an array is directly related to the size of its aperture. For airborne applications in which a sensor array is mounted on a small aircraft, the physical space available on the airframe is limited and the array aper- ture is restricted. Earlier work in [1] proposed the use of a single vector sensor consisting of two orthogonal triads of dipole and loop antennas with the same phase center for This work was sponsored by the Department of Defense under Air Force Contract F19628-00-C-0002. Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily en- dorsed by the United States Government. The authors acknowledge the work of Dan Bliss, Amanda Chan, and Alex Eapen for their modeling work to establish an appropriate vector sen- sor array for a small aircraft. source localization. Because a vector sensor uses multiple components of electromagnetic information, it can offer ac- curate AOA estimates with a smaller aperture. This paper examines the problem of using vector sen- sors mounted on a small aircraft to perform source local- ization. Section 2 contains background material and formu- lates the signal model. Section 3 presents and discusses the simulation results. Section 4 provides the conclusion. 2. PROBLEM FORMULATION Initial work with a vector sensor mounted on an aircraft has indicated that some elements of the vector sensor act as “feeds” for the airframe, rendering them unusable. The proposed solution is to use a “trimmed” vector sensor em- ploying only the elements with insignificant airframe inter- action. Multiple trimmed vectors sensors are sited at various locations on the airframe so that additional sensor elements are available for accurate AOA estimates. An example of a trimmed vector sensor is shown in Fig. 1(a).The two loop antennas measure the x and y compo- nents of the magnetic field, while the vertical dipole mea- sures the z component of the electric field. Fig. 1(b) shows a potential 8-channel configuration for placement on an air- craft that will be studied in this paper. The trimmed vector sensors on the aircraft are not all identical; the one mounted on the nose of the airframe lacks the vertical dipole. Fig. 1. (a) Trimmed vector sensor consisting of two loop and one dipole antenna. (b) 8-channel aircraft configura- tion. 2.1. Signal Model It assumed (as in e.g. [1]) that the vector sensor array is in the far-field of a narrowband signal. Following [2], define IV - 1033 0-7803-8874-7/05/$20.00 ©2005 IEEE ICASSP 2005 ➠ ➡