Distributed Adaptive Control and Metrology for Large Radar Apertures

bendembarrassΗλεκτρονική - Συσκευές

2 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

104 εμφανίσεις

Distributed Adaptive Control and Metrology

for Large Radar Apertures


PI: James Lux

Co
-
Is: Adam Freedman, John Huang, Andy Kissil, Kouji Nishimoto, Farinaz Tehrani

Poster No.


04
-

039

Optical sensor measures coarse
(1:500) mechanical position and
orientation by looking for LED
“stars”

Widely available 640x480 sensors provide this level of
performance.

Camera calibration / photogrammetry algorithms
needed are well known.

Other Components

Antenna Elements

CMOS image sensor

In
-
band RF measurement of precise
phase shift & path loss to/from beacons.

Evaluated two basic approaches:

PN ranging (
e.g.

GPS): acquire and synchronize to high rate code

Phase Comparison (
e.g.

Omega, DECCA, Loran)

CW beacon phase measurement and comparison selected

Directly gives us the metric we need: phase shifts

Usual problem with phase comparison (ambiguity) not an issue because we
have mechanical constraints on possible locations. RF measurements also
measure gain

Measurements must be made in two bands with multiple frequencies
per band

Bands: working (radar) frequency and link frequency

Number of frequencies must be > number of degrees of freedom

Measurements feed into microcontroller algorithms

executing on each element:


Mechanical & Structural Model


predict future position of element


Beamforming Computations


look direction, element position/orientation


Element Control


compensate for variations

Goal: Conceptual design and analysis for novel method to compensate
for inevitable movements in large (>100m
2
) lightweight radar apertures.

Conceptual Approach
: Electronically scanned
phased array

using phase and amplitude controls
at each element to form the beam and steer it.
Each element includes simple hardware to
measure its own position and RF properties

and an
embedded processor compensates for the
variations

from ideal. The measurements are made
by “looking” at beacons in the structure that have
“well known” position and characteristics (similar
to surveying benchmarks).

Quantitative Performance Requirements:

Apertures on order of 100
-
400 m
2
, 50m linear extent

Hardware ROM mass near 2kg/m
2

Compensate to <1/20
th

wavelength at L band (1.2
GHz, 23 cm) (
i.e.

1 cm)


Mechanical deviations on order of 1 meter at 1Hz
(
i.e.

few m/s velocities)


RF property changes due to adjacent element
interaction, aging, temperature

Radio

Computer

Incoming signals arrive
staggered in time and
phase, depending on
direction of arrival.

Embedded microcontroller in element commands radio
to adjust time delay, phase, and amplitude to
compensate for:

-

Direction of arrival

-

Physical position

-

Radio performance variations

Adjusted element signals are
transmitted by radio to a common
receiving point where they are
summed.

Metrology

Radio

Computer

Metrology

Radio

Computer

Metrology

Metrology on each element
measures position of element
and RF performance

Control

Wireless links used
for control, status,
and software loads
from central point
and to share
information with
adjacent elements.

Radar

Processor

Beacons:


Radiate RF CW signals with well controlled frequency and phase.


Have an optical source (LED) that elements can see. Beacon
positions are well surveyed and broadcast to all elements.


Number of beacons determined by two primary factors:


FOV of optical sensors: 4
-
6 beacons visible


Uncertainties in element antenna patterns: RF path to beacon
should be reasonably close to path to radar beam direction


Relative geometry of beacons is important also

FY04 tasks


Generate set of quantitative bounding
requirements from published literature and
proposed future missions

Size, performance, structural dynamics



Analytical and Simulation models of conceptual
design



Metrology and calibration approaches

Blend of optical and RF techniques

Tradeoff of optical FOV, resolution

Develop RF metrology requirements and concept

Parallelizable Modeling
Codes


Calculates contribution of each
element (or group of elements) and
sums electric field components.


Ready for insertion of “element
implementation”specific models:

Microcontroller algorithms

Optical sensor error model

RF metrology error model


Modeling software structure
designed to be similar to actual
element execution environment.

Breadboard Validation of RF Phase Measurement

SDR1000


“Digital Radio”

VIA EPIA


533MHz C3

5 port

Switch

802.11a WLAN

access point

Radar Freq

Converter

Link Freq

Converter

RF

In/Out

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

-
21.94

-
21.92

-
21.9

-
21.88

-
21.86

-
21.84

-
21.82

-
21.8

-
21.78

-
21.76

time

DDS error (ppm)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

-
20

-
15

-
10

-
5

0

5

10

15

20

time

Residual frequency error (ppb)

sigma = 5.02 ppb

Simple algorithm (straight line predictor) removes variations from 10 MHz DDS output


(
s
f
/
f

= 5 ppb after removal of linear trend) > 5ppb is approximately 1.8
º

phase error / 0.1second

Measured variation (approx 150 ppb) in f
osc

over two
seconds (linear trend shown in green)

Residual variation after using linear predictor

{previous 0.3 seconds predicts 0.1 second into future}

4
breadboard
elements

Metrology and Calibration Concept

Measured variation in commercial crystal oscillators in 4 breadboard
elements frequency using 2 reference RF tones from a “beacon”.