Communications: directive radio wave systems and devices (e.g. – Determining direction – Low angle processing
Reexamination Certificate
2001-12-27
2003-02-25
Tarcza, Thomas H. (Department: 3662)
Communications: directive radio wave systems and devices (e.g.,
Determining direction
Low angle processing
C342S156000, C342S159000
Reexamination Certificate
active
06525685
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to multipath radar interferometers, and more particularly, to a method and apparatus for detecting and eliminating signal angle-of-arrival (AOA) errors caused by multipath. More specifically, the present invention substantially improves AOA estimation accuracy and reliability when using super resolution algorithms.
BACKGROUND OF THE INVENTION
Ship borne search and track radars typically are directed to an emitter by azimuth and elevation coordinates generated from passive measurements in an electronic support (ES) subsystem. The ES system is said to cue the radar. Such cueing avoids time consuming radar target searches over a large volume of space, and hence minimizes the radar's time-to-range. The ES angle outputs may also be used to drive an emitter tracker.
To be effective in cueing, the ES system must direct the radar to numerous closely spaced targets. This requires azimuth and elevation angle measurement precision comparable to that of the radar. The preferred approach for obtaining this precision as depicted in
FIG. 1
, is an array of antennas
103
utilized as an interferometer. If the spacing between the outermost interferometer antennas is tens of signal wavelengths, the array can measure signal angle-of-arrival (AOA) to a fraction of a degree. The interferometer can do this over a wide field of view, and over a wide range of frequencies. For both cueing and tracking, the interferometer AOA outputs must be available at an output rate of several Hertz over a period of many seconds.
The interferometer measures the phase difference between pairs of antennas in the array. The signal angle-of-arrival resolution is related to this phase measurement by an accuracy proportional to the spacing d
1
101
and d
2
102
between the antenna pairs. Hence the requirement for spacings many wavelengths long. But when the antenna pair are more than half a wavelength apart, the phase measured is ambiguous modulo 360°. A special algorithm, typically tailored to the specific antenna spacings used in the array, must resolve these ambiguities before emitter AOA is found. Conventional ambiguity resolution algorithms (ARA) assume only one signal is present.
Unfortunately, many times in shipboard use this single-signal requirement is not met because of multipath. The presence of two or more simultaneous signals, such as strong specular or diffuse multipath, can cause spurious outputs. Diffuse multipath may produce large phase measurement noise, larger than the ambiguity resolution algorithm was designed to handle robustly. This noise can be so large it causes incorrect ambiguity resolution or “gross errors”. When gross errors occur the resulting spatial errors can be on the order of tens of degrees.
Diffuse multipath affects mainly arrays oriented in azimuth, or with axis
104
parallel to the water's surface. Specular multipath does not affect a strictly azimuth array. But specular multipath has a profound impact on an interferometer used for elevation angle measurements, i.e. with axis
104
mounted normal to the reflecting surface. And, if the antenna platform is not stabilized, ship pitch and roll assures azimuth and elevation arrays are both affected by specular and diffuse multipath.
Specular multipath induces unacceptable phase errors in an elevation interferometer array by creating a second signal interfering with the direct path one. At low elevation angles these simultaneous interfering signals can be of nearly equal strength Antenna element beam shaping cannot mitigate this phenomenon sufficiently at low elevation angles. Since the ARA and ultimate interferometer AOA generating algorithm are based on measuring phase created by a single plane wave, specular multipath renders them useless for generating elevation measurements to cue the radar or provide angle estimates to a passive emitter tracker.
A straight forward multipath mitigation approach is phase measurement data editing. Specular multipath interference creates a standing wave detected at the array by the varying signal amplitude induced across the different antennas. So one method of data editing is to reject the phase measurements when the amplitude variation exceeds a predetermined threshold. Another method involves histogramming to determine outliers. But such data editing is not viable when cueing a radar or establishing emitter tracks. So many measurements may be rejected that large gaps in elevation angle output, averaging many seconds in length, can result. This is disastrous for cueing and tracking support. Reliable angle estimates at regular, predetermined intervals are required in these applications.
Therefore, the conventional interferometer algorithm must be augmented, especially for elevation processing, in a manner that recovers the direct path phase from multipath corrupted measurements. Currently this augmentation is typically done using a super resolution approach that separates the true signal from it's specular reflection. In particular, the MUiltiple SIgnal Classification (MUSIC) method, described by Schmidt in “Multiple Emitter Location and Signal Parameter Estimation,”
Proc. RADC Spectrum Estimation Workshop,
October 1979, has been extensively studied for this application. When used for multipath processing the original MUSIC subspace approach had a drawback: the interfering signals must be uncorrelated. This is not the case for multipath. Multipath is the original signal simply shifted in phase and amplitude. But work-a-rounds have been introduced involving decorrelation by spatial averaging, and these work-a-rounds are referred to generically as modified MUSIC.
The spatial averaging is accomplished by implementing interferometer arrays with clusters of elements having certain symmetries. Roy, Paulraj and Kailaith detail the use of such spatial averaging in U.S. Pat. No. 4,965,732, “Methods and Arrangements for Signal Reception and Parameter Estimation.” Their approach does not utilize MUSIC, but a super resolution algorithm based on the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT). ESPRIT has also been extensively studied for shipboard interferometer use.
MUSIC and ESPRIT are called subspace techniques because they utilize the fact that all signal vectors are orthogonal to all vectors in the measurement noise subspace. Dividing by the inner product of the candidate signal vector and noise space vectors, these algorithms generate peaks at the true signals. The peak would be at divide-by-zeros if there were no noise. But the presence of measurement error limits their magnitude. The magnitude of the peak is thus a function of the signal-to-noise ratio (SNR). When applying super resolution methods to the interferometer array processing application, where the array antennas are spaced several wavelengths apart, many peaks are generated besides the direct path signal and the multipath signal. Some peaks are caused solely by noise; some peaks appear at the interferometer's gross error points, or ambiguous AOAs; and some peaks, at higher frequencies, are caused by aliasing. Thus to work robustly, the super resolution approach, whether modified MUSIC, the implementation of ESPRIT disclosed by Roy, or some other subspace technique, must consistently determine which of the many possible outputs is the direct path. The success the method has doing this depends strongly on the direct and reflected signal SNR. When the two signals have comparable SNR, performance can degenerate dramatically for all current subspace approaches.
In the ship board radar cueing and tracking application, comparable SNRs occur often when the emitter is below 10° elevation. The reflected signal strength can also be particularly high when the emitter's signal is horizontally polarized. To assess the performance of subspace algorithms in the critical low elevation region, Litton Advanced Systems (LAS) conducted a field test with the array shown in
FIG. 1
during the Fall of 1999. This array has a folded symmetry abo
Andrea Brian
Lowe Hauptman & Gilman & Berner LLP
Northrop Grumman Corporation
Tarcza Thomas H.
LandOfFree
Method and apparatus for detecting and eliminating signal... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for detecting and eliminating signal..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for detecting and eliminating signal... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3137996