On the Improvement of the Phased Array Weather Radar Data - - - PowerPoint PPT Presentation

On the Improvement of the Phased Array Weather Radar Data

• Advantage and Disadvantage -

Tomoo Ushio, H. Kikuchi, T. Mega (Tokyo Metropolitan University, Osaka University)

• S. Satoh, T. Iguchi (NICT), F. Mizutani, M. Wada (Toshiba)

Scan system Elevation: Electronic scan Azimuth: Mechanical scan Coverage 3D scan (100 elevations) / 10 sec (- 1 min) Parameters Zh, vh, σvh (single-polarization)

• Achieves high speed

volume scan of 30 sec to detect severe storms instantaneously in urban area

Phased Array RADAR in Osaka University, 2012

Antenna （Patch Antenna with RF-CMOS） Control/Signal Processing Units （MMSE algorithm） Terminal data Internet Toshiba NIED etc Warning Users Real time Transfer MP-PAWR Dome

Dual Polarimetric Phased Array Radar System in 2017

Parabolic Radar Phased Array Radar

• Pencil beam

→mechanically scanning both in elevation and azimuth

• Parabolic antenna goes around

in a certain elevation angle PARABOLIC TYPE ・Fan beam → electrically scan in elevation PHASED ARRAY TYPE

Fast Scanning Strategy

Dual Polarimetric obs. by MP-PAWR

気象学会2018年度秋季大会 C460 5

Zh V Zdr φdp Kdp ρhv

Zh Vh Zdr ρhv

PPI(Pr)

7

Clear Sky (2018/08/02 12:00) Rainy (2018/07/28 16:00)

PPI(Φdp)

8

Clear Sky (2018/08/02 12:00) Rainy (2018/07/28 16:00)

Advantages and Disadvantages

• Advantage

– Fast scanning and dense observation – Early and accurate warning

• Disadvantage

– Ground clutter contamination – ghost precipitation in data assimilation – less accurate warning

• How can we reduce this disadvantage?

10

Problem

2-way beam pattern is poor

Fan beam pattern for transmitting Receiving beam pattern by uniform phase shift

• 13.6 dB

2-way beam pattern

High side lobe level

The received signal from precipitation is seriously contaminated by the relatively high received power from ground and strong precipitation echoes near by through the side lobes of the 2 way beam pattern.

Strong Ground Clutter

Transmitting Broad Beam

Masking Problem

High side lobe level from the two way beam pattern

A beam forming technique using the MMSE (Minimum Mean Square Error) formulation has been proposed and tested. This approach can adaptively mitigate the masking interference that results from the standard digital beam forming method in the vicinity of ground clutter and strong precipitation area

12

Signal Processing in PAWR

[ ]

1 , 1 , 1 , ´

• Î

=

N T l M l l l

y y y C y !

[ ]

1 , 1 , 1 , ´

• Î

=

M T l M l l l

x x x C x !

T d M j d j

e e ú û ù ê ë é =

• q

l p q l p

q

cos ) 1 ( 2 cos 2

1 ) ( ! s λ: wave length, l: pulse number Range = r0 M associated complex amplitudes N received complex amplitude Steering vector

unknown known

Range = r0

13

Signal Processing in PAWR

y S x = +

N M

[ ]

M N N ´

Î = C s s s S ) ( ) ( ) (

2 1

q q q !

v

1 ´

Î

N

C v : Additive Gaussian noise vector

unknown known

Range = r0

14

Signal Processing in PAWR

Beam forming

l H m l m

x y w =

,

ˆ

[ ]

1 , 1 , 1 ,

ˆ ˆ ˆ ˆ

´

• Î

=

M l M l l l

x x x C x !

[ ]

1 1 1 ´

• Î

=

N N m

w w w C w ! : Complex weighting vector Phased array beam forming is … – to calculate a weighting average of received signals in each element – to calculate an optimal weights

15

Fourier Method

Fourier Beam Forming (FR)

– Equivalent to a Fourier transform of y vector

( )

N

m m

q s w =

m

q

n

d

m n

d q sin

S

Origin Amplitude Phase Shifter

) (t En

) (t Esum

ú û ù ê ë é

• =

2

y w x

H MMSE m m

m

E J

Minimum Mean Square Error Formulation

16

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

Lower sidelobes

m

x

: Reference signal m q

y : Received signal

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

(λ: Lagrange Multiplier)

17

Lower sidelobes Preserve power in main lobe

m

x

: Reference signal

m

q

y : Received signal

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

minimize

) ( ) ( ) (

1 1 m m H m MMSEm

q q q s R s s R w

• =

v H x

R S SR R + =

x

R ：Covariance of x

v

R ： Covariance of noise

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

18

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

minimize

) ( ) ( ) (

1 1 m m H m MMSEm

q q q s R s s R w

• =

v H x

R S SR R + =

x

R ： Covariance of x

v

R ： Covariance of noise

unknown

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

19

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

minimize

) ( ) ( ) (

1 1 m m H m MMSEm

q q q s R s s R w

• =

v H x

R S SR R + =

x

R ： Covariance of x

1) Initial value

l H BF BF

m m l

y w x =

,

ˆ

v

R ： Covariance of noise

Beam former method

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

20

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

minimize

) ( ) ( ) (

1 1 m m H m MMSEm

q q q s R s s R w

• =

v H x

R S SR R + =

x

R ： Covariance of x

1) Initial value 2) Calculation of weighting vector

) ( ) ( ) (

1 ) ( 1 ) ( ) ( m i m H m i MMSE i

m

q q q s R s s R w

• =

v

R ： Covariance of noise

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

21

l H BF BF

m m l

y w x =

,

ˆ

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

3) Weighting average

l MMSE i i MMSE

m H l

y w x

) ( ) (

ˆ =

22

minimize

) ( ) ( ) (

1 1 m m H m MMSEm

q q q s R s s R w

• =

v H x

R S SR R + =

x

R ： Covariance of x

1) Initial value 2) Calculation of weighting vector

) ( ) ( ) (

1 ) ( 1 ) ( ) ( m i m H m i MMSE i

m

q q q s R s s R w

• =

v

R ： Covariance of noise

( ) { }

1 ) ( Re

2

• +

ú û ù ê ë é

• =

m H MMSE H MMSE m m

m m

E J q l s w y w x

22

l H BF BF

m m l

y w x =

,

ˆ

Minimum Mean Square Error Formulation

Weighting factor is calculated by minimizing the residuals, defined as the difference between the reference and the weighting average of the received signal

23

MMSE Beam Forming

Re-iterative super resolution (RISR) algorithm

1) Initialization (i=0) As a prior information, FR solution is calculated.

[ ]

, 1 , 1 ,

ˆ ˆ ˆ ˆ

l M l l l

x x x

• =

! x

( )

m m FR m

q s w w = º , ˆ

, l H m l m

x y w =

2) Determination of MMSE weights

) ( ) ]( ) ( [

1 2 m v H i x i m i m

x E q q s R S R S w

• +

=

4) Computation of MMSE solution and Re-iteration

[ ]

1 , 1 , 1 , 1

ˆ ˆ ˆ ˆ

+

• + =

i l M l l i l

x x x ! x , ˆ

1 , l i H m i l m

x y w =

+

3) Gain control

,

i m i m

gw w = g : gain control factor

wave source : 28dB from 25[deg]

1

x x : 60dB from 30[deg]

DBF Results

Numerical Calculations

25

Simulation results

With a strong clutter

• MMSEs properly suppresses strong clutter
• The iteration in MMSE improves the solution.

MMSE-4 MMSE-1 Capon (CP) Fourier (FR) Truth Precipitation Clutter

Application to real data

MMSE FR method Corrected MMSE

FR:elv=0° FR:elv=3° FR:elv=5° MMSE:elv=0° MMSE:elv=3° MMSE:elv=5° MMSE with correction elv=0° MMSE with correction elv=3° MMSE with correction elv=5°

Application to real data

MMSE method FR method Corrected MMSE Clear Sky Rainy Condition

Filtered image (0 deg. ) 除去なし

Clutter mitigation with MMSE technique in doppler domain

Original image

Summary

• PAWR has rapid scanning and dense
• bservation capabilities.
• It also has the advantage of observing without

any ground clutters.

• The only drawback would be a cost.

Kikuchi et al. IEEE TGRS, 2017

31

Flow –Transmission

Fan beam is transmitted by feeding power into 24 elements (max) (with about 10 deg beam width)

Scattered signals are received by all 128 antennas 128 data are stored by 128 ADCs

Flow –Reception

33

Sharp received beams (of about 1 deg) are formed by digital signal processing

Flow –Signal Processing

Outline

• Introduction to Phased Array Weather Radar in

Osaka University and Tokyo Metropolitan area

• Clutter mitigation technique in Phased Array

Weather Radar