Académique Documents
Professionnel Documents
Culture Documents
Outline
J ( y j , t)
Applications in sensing
Raman spectroscopy for bio and environmental sensing
Photonic Crystal
spectrometer
(in nano-scale)
Human
tissue
Spatially incoherent
light
Output B
Output C
Output D
Output E
Output F
Incident
Spatially Incoherent
Field
Heterogeneous
Photonic Structure
Spectrally
Diverse
Field
Detectors
Input Source
(incoherent)
at A
Output at B
(electric field
intensity)
J ( y, t )
E 2 ( y, t )
Maxwell Equation
Maxwell equations:
H (r, t )
E(r, t ) =
t
E(r, t )
H (r, t ) = (r )
+ J (r, t )
t
E = 0
H = 0
E(r, t ) = (r )
2
t
t
J p ( y j , t ) = 0, ( j i ).
Compute output electric field at B Eip ( y, t )
2
Total electric intensity at B:E ( xB , y, t ) =
p
2
(
E
(
x
,
y
,
t
))
.
i B
i
i, j
A Stochastic Model
( E ( x, y, t )) xx + ( E ( x, y, t )) yy ( x, y )( E ( x, y, t ))tt = ( J ( x, y, t ))t
a
t t0 2
V (t ) = sin ( (t t 0 ) ) exp
T
Stochastic model
X ( y ) = dW ( y ),
W ( y ) Brownian Motion.
More general: J t ( x, y, t ) = f ( x, y, t ) dW ( x, y, t ),
Monte Carlo
Not many computational methods available.
Traditional methods, Monte Carlo (MC)
simulations,
Solve the equations realization by realization.
Each realization, the equations become deterministic and
solved by classical methods.
The solutions are treated as samples to extract statistical
properties.
Monte Carlo
MC can be very expensive:
1
Has slow convergence, governed by law of large numbers O( ),
n
and the convergence is not monotone.( n is the number of
MC realizations)
Need to resolve the fine scales in each realization to obtain
the small scale effects on large scales, while only large scale
statistics are of interests, such as long time and large scale
behaviors.
Hard to estimate errors
Must involve random number generators, which need to be
carefully chosen.
T ( ) = H i ( i ).
i =1
u (t , x, dW ) = u (t , x)Ta ( ),
where
u (t , x) = E (u (t , x, )T ( )).
Eu (t , x) = u0 (t , x),
variance
E (u )(t , x) = u
2
E (T0 ( )) = 1,
E (T ( )) = 0, 0
0
E (T T ) =
1 =
Ghanem, et al (91,99, ).
E ( x, y, t , dW ) = Ei ( x, y , t )Ti ( )
Take advantage of
dW ( y ) = i mi ( y )
i =1
J ( y , t , dW ) = V (t )mi ( y ) i
Percentage error
20ax10a
Number of coefficients
For 15 coefficients the gain in simulation time is 32, i.e., 32 times faster simulation
Less than 1% error and more than one order of magnitude faster simulation,
Over 2 order of magnitude faster simulations for practical photonic crystals.
Conclusion
Proposed a stochastic model for incoherent source
Design a fast numerical method based on WCE to
simulate the incoherent source for photonic crystals.
The method can be coupled with other fast Maxwell
equations solvers.
More than 2 order of magnitude faster simulations can
be achieved for practical structures.
The model and method are general and can be applied
to other types of stochastic problems involving incoherent
sources.