Characteristics of electromagnetic wave propagation in time-varying magnetized

plasma in magnetic window region of reentry blackout mitigation

, , , , , and

Citation: AIP Advances 7, 025114 (2017); doi: 10.1063/1.4977544

View online: http://dx.doi.org/10.1063/1.4977544
View Table of Contents: http://aip.scitation.org/toc/adv/7/2
Published by the American Institute of Physics
 AIP ADVANCES 7, 025114 (2017)

Characteristics of electromagnetic wave propagation

in time-varying magnetized plasma in magnetic
window region of reentry blackout mitigation
Hui Zhou (),1,a Xiaoping Li (),1 Kai Xie (),1,b
Yanming Liu (),1 Bo Yao (),1 and Wei Ai ()2
1 School of Aerospace Science and Technology, Xidian University, Xian 710126, China
2 Science and Technology on Space Physics Technology Laboratory, Beijing 100076, China
(Received 16 January 2017; accepted 12 February 2017; published online 23 February 2017)

The magnetic window is considered a promising means to eliminate reentry com-

munication blackout. However, the turbulence of plasma sheath results in phase
jitter and amplitude turbulence of electromagnetic (EM) wave and may influence
the eliminating effect. Therefore, the effect of fluctuating property of reentry plasma
sheath on EM wave propagation when a magnetic field is used for eliminating black-
out is investigated. For this purpose, a time-varying electron density model, which
includes both temporal variation and spatial turbulence, is proposed. Hybrid matrix
method is also employed to investigate the interaction between time-varying magne-
tized plasma and EM wave. The EM wave transmission coefficients in time-varying
magnetized and unmagnetized plasmas are likewise compared. Simulation results
show that amplitude variation and phase jitter also exist on transmitted EM wave,
and the turbulent deviation increases as the degree of plasma fluctuates. Mean-
while, the fluctuation of transmitted EM wave attenuates at low-frequency pass-
band and increases at high-frequency passband with the increasing magnetic field.
That is, comparing with unmagnetized time-varying plasma, the fluctuation effect
can be mitigated by using a magnetic field when the EM wave frequency is at
low-frequency passband. However, the mitigating effect can be influenced by the
nonuniformity of magnetic field. 2017 Author(s). All article content, except where
otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license
(http://creativecommons.org/licenses/by/4.0/). [http://dx.doi.org/10.1063/1.4977544]

I. INTRODUCTION
When spacecrafts reenter the Earths atmosphere or fly at hypersonic velocity, the surrounding
molecules are dissociated and ionized due to the shock heating of the air or ablation of the heat shield
material. The ionized gases, called reentry plasma sheath, result in the loss, or severely degrade,
communication, navigation, and telemetry signal.1 Considerable effort has been made to address the
problems associated with the reentry plasma sheath, and a number of techniques have been proposed
to mitigate or eliminate the reentry communication blackout.24 These methods achieved some good
results but introduced other problems. Magnetic window, a method that was advocated by Hodara5
in the 1960s, is a promising method that is resurfacing. This method does not affect the aerodynamic or
heat shielding configuration. Several theoretical studies on the magnetic window method have been
performed.69 All these studies have revealed that when a right-handed (RH) circularly polarized
wave propagates in the direction of the static magnetic field B0 , a window appear below the gyro
frequency
b = (eB0 )/m, (1)
where e and m are charge and mass of the electron.

a
[email protected].
b
[email protected]

2158-3226/2017/7(2)/025114/14 7, 025114-1 Author(s) 2017

025114-2 Zhou et al. AIP Advances 7, 025114 (2017)

In the above studies, only the effect of magnetic window method on the improvement of
electromagnetic (EM) wave propagation in a steady-state plasma has been considered. However,
the parameters, such as electron density, of plasma sheath vary with the angle-of-attack variations,
flying attitude adjustment of hypersonic vehicle, turbulent boundary layer, and nonuniform abla-
tion.10,11 Thereby, these dynamic physical factors inevitably affect the EM waves.1217 Yao et al.
have established a model of time-varying reentry plasma sheath and discussed the effect of turbu-
lence perturbation on wave-phase jitter and amplitude fluctuation.12 In Ref. 17, the influences of
reentry time-varying plasma on the EM wave propagation were also analyzed. The result indicated
that communication interruption persists, even though the carrier frequency is larger than the plasma
resonant frequency. This finding can be explained by the parasitic modulation effect caused by
amplitude fluctuation and phase jitter.18,19 The parasitic modulation effects can be observed clearly
when an EM wave transmits through a time-varying plasma produced by a ground-plasma-glow
discharge device.18 Yang et al. analyzed the effect of time-varying plasma on the propagation of
quadrature phase-shift keying signals.19 They reported that the wave amplitude fluctuation and phase
jitter lead to the rotation and distortion of constellation of the receiving signals, resulting in the
increasing error rate of signals. Previous studies confirmed that the amplitude fluctuation and phase
jitter of transmitted EM wave caused by plasma fluctuation are unneglected factors for the reentry
communication interruption.
The dynamic property of plasma sheath cannot be ignored when a magnetic field is applied to
mitigate blackout problem. Several previous literatures focused on the EM wave propagation in tran-
sient or rapid created magnetoplasma.20,21 Meanwhile, the reentry plasma sheath varies continuously,
and the interaction between the EM waves and the continuous time-varying plasma is quite differ-
ent.19 Therefore, the propagation properties of EM waves in a continuous time-varying magnetized
plasma should be given attention. However, less attention has been provided on relevant study, par-
ticularly the effect of plasma fluctuation on wave phase. A laboratory experiment on communication
signal propagation in continuous time-varying magnetized plasma was carried out in Ref. 22. The
mean plasma density of plasma source is homogeneous, and the plasma fluctuation frequency is a
single frequency. However, the experiment plasma is too simplified compared with the actual plasma
sheath. The fluctuation properties, such as combined nonuniform plasma electron density and mul-
tiple perturbation frequency components, of reentry plasma sheath are complicated. Therefore, the
characteristics of EM wave propagation in continuous time-varying magnetized plasma are presented
in this study.
The rest of this study is organized as follows. Section II describes an accurate mathematical
model of the time-varying electron density of dynamic reentry plasma sheath. Section III elaborates
the physical model of time-varying magnetized plasma. A method used for studying the properties of
EM wave propagation in continuous time-varying magnetized plasma is also introduced in this section.
Simulation results and discussion are presented in Section IV. Section V provides the conclusion.

II. MATHEMATICAL MODEL OF TIME-VARYING ELECTRON DENSITY OF DYNAMIC

PLASMA SHEATH
The most important plasma parameter that dominates EM wave propagation is electron density.
Electron is often subject to a nonuniform distribution and mainly determined by altitude and Mach
number for a steady-state reentry plasma. The main factors affecting plasma dynamics are the flight
status and the internal flow disturbances changing the physical parameters at different rates. Then, the
electron density varies randomly with spatial distance changing and time increasing. Spatialtemporal
electron density Ne (z, t) can be described as follows:
Ne (z, t) = Ne (z, t) + Ne (z, t) = Ne (z, t) [1 + Ne (z, t)/Ne (z, t)], (2)

where z is the radial distance from the aircraft surface, t is the time axis, Ne (z, t) is the mean value
of Ne (z, t) (this expression is also applicable to other physical quantity), and is the disturbance
variables. Ne (z, t) can be considered as a steady-state plasma electron density. A sketch of plasma
sheath and antenna location is plotted from the result of Refs. 23 and 24 in Fig. 1. The plasma electron
density decreases both from the stagnation point region to rear region of spacecraft body and from
025114-3 Zhou et al. AIP Advances 7, 025114 (2017)

FIG. 1. Sketch of plasma sheath and antenna location.

the spacecraft surface to free space.24 For improving the antenna transmission efficiency, the antenna
on the spacecraft was located on the rear region of spacecraft body.23 This position was deliberately
chosen as having the least plasma and still being in a position to radiate to selected ground sites. As
the order of magnitude the antenna window size scale is much less than the spacecraft size scale, the
variation of electron density along the spacecraft at antenna window can be neglected. Meanwhile,
the electron density along the direction vertical to the spacecraft surface at antenna location can be
represented as two Gaussian distributions, and the plasma layer width at antenna location is regarded
as constant approximately.23,25 Thus, the mathematical expression of Ne (z, t) at antenna location is
given as
2
Nep exp[1 (z zp ) ] 0 z zp

Ne (z, t) =
N exp[ (z z )2 ] z z z , (3)
ep 2 p p max

where Nep is the peak electron density, 1 and 2 represent the shape of each curve, z is the radial
distance from the aircraft surface, zp determines the location of maximum electron density, and zmax
is the total thickness of the plasma layer. Ne (z, t) is independent with variable t and can be rewritten
by Ne (z). Meanwhile, Ne (z, t) is the two-dimensional function of z and t.
The relations among the variables of electron density N e , gas temperature T, and pressure P in
fluid field are described in previous literatures Refs. 14 and 15. The equations for fixed position za
are given as
!
Ne (za , t) 1 Qi T (za , t)
= + , (4)
Ne (za ) 4 2KT (za ) T (za )
and
1 P(za , t)
! !
Ne (za , t) 1 Qi
= + , (5)
Ne (za ) 4 2KT (za ) P(za )
where K is the Boltzmann constant, Qi is the ionization energy, and is a specific heat ratio. Here,
Ne (za , t), T (za , t), and T (za , t) are only dependent on variable t, and Ne (za ), T (za ), and P(za ) are
constants. Therefore, Ne (za , t) is approximately proportional to temperature variation and pressure
variation, that is,
T (za , t) P(za , t)
Ne (za , t) . (6)
T (za ) P(za )
As found in a hypersonic wind tunnel experiment, the probability density function of
T (za , t)/T (za ) follows Gaussian distribution,26 as well as Ne (za , t) that can be obtained from
Equation (6). The test conditions of this experiment were: temperature = 695 K, total pressure = 34
atm, free-stream unit Reynolds number = 10.9 106 /m, and free-stream Mach number = 7.0, The
test core diameter was approximately 0.7 m with an axial Mach number gradient of less than 0.12/m.
025114-4 Zhou et al. AIP Advances 7, 025114 (2017)

Then, the probability density function of Ne (za , t) can be written as

2
1 (Ne (za , t)- a)
PDF[Ne (za , t)] = G[Ne (za , t), a , a ] = e 2a2
, (7)
2a
where PDF() is the probability density function, Gauss() is the function of Gaussian distribution,
and a and a are the mean value and standard deviation of Ne(za , t). In hypersonic boundary
layer, the turbulences of fluid field are principally caused by second mode instability wave; thus, a
peak appears at both zero frequency and high frequency between 200 kHz and 300 kHz in the power
spectrum.27 The second mode instability wave also influences the pressure in a hypersonic wind
tunnel, the test conditions of which were: temperature = 418 K, total pressure = 8.5 atm, free-stream
unit Reynolds number = 10.45 106 /m, free-stream Mach number = 6.0. As shown in Equation (6),
Ne(za , t) is proportional to pressure variation magnitude. Thus, electron density is also affected by
second mode instability wave. By analyzing the above fluctuation properties of electron density, the
power spectrum of Ne(za , t) can be defined as Ref. 12
PSNe (za , t) ( f ) = G[ f , 0, ga ] + G[ f , fh , ha ], (8)
where f is the fluctuating frequency, fh is the center frequency of high frequency, and ga and ha are
the shape parameters of the power spectrum. The power spectrum of electron density can be regarded
as the summation of two Gaussian functions of which peaks are at zero and high frequency. The
wind tunnel experiments gave attention on hypersonic flow which is similar to the reentry fluid fields.
Thus, the fluctuation laws of wind tunnel flow can be used on reentry plasma approximately.
The fluctuation characteristics of electron density at single-point are described above. In general,
the electron density of plasma sheath is nonuniform, and fluid velocities vary widely at different
locations. Therefore, the amplitudes of destabilization at different locations are unequal. A linear
eddy model (LEM) is used in Ref. 16 to achieve the spatial turbulence and has been very effective
with the triplet mapping process (eddy event) as follows:

z0 + 1 (z z0 ) z [z0 , z0 + 1 l]

+

z l ( )(z z0 )
0 2 2 1



0 LEM
z z [z0 + 1 l, z0 + 2 l] , (9)
z0 + 2 l + (1 2 )(z z0 )






z [z0 + 2 l, z0 + l]



where z and z 0 are the original and fluctuating coordinates, z0 and l are the initial point and the length
of fluctuation, and 1 and 2 are the two parameters determining the relative length of the three

FIG. 2. Electron density model for spatial turbulence and temporal variation (N ep = 11017 /m3 ).
025114-5 Zhou et al. AIP Advances 7, 025114 (2017)

parts of the turbulence. Then, the original profile at fluctuant interval [z0 , z0 +l] can be replaced by
using Equation (9). Meanwhile, parameters outside the fluctuating interval are unchanged. Typically,
the initial point follows uniform distribution: p(z0 ) U[0, zp ] and the probability density function of
the turbulence length l is proportional to the inverse of its square, that is PDF(l) 1/l 2 . Thus, the
time-varying electron density (Equation (2)) can be rewritten as
Ne (z, t) = NLEM (z, t) [1 + Ne (z, t)/NLEM (z, t)], (10)

where NLEM (z, t) is mapped by Ne (z) for arbitrary time t. Some parameters mentioned above can be
chosen as = 0 in Equation (7), fh =250 kHz and ga = la =2104 in Equation (8),12 and 1 =1/3 and
2 =2/3 in Equation (9)28 in the following simulation. One example of the simulated electron density
profile is shown in Fig. 2.

III. INTERACTION BETWEEN EM WAVE AND TIME-VARYING PLASMA

A. Physical model of time-varying plasma
The physical model for EM wave propagation in homogeneous time-varying magnetized plasma
is presented in this section. This model is considered the problem of propagation through time-varying
plasma of finite thickness. Ion motion is neglected and electron-neutral collisions are considered as
a constant . A constant magnetic field is assumed as B0 = B0 z parallel to the vector propagation
constant k. The special case of wave propagation along the magnetic field is investigated. Thus, the
wave electric field is expected to be in the xy plane. Assuming that the time-varying plasma is
regarded as a lossy medium, the effect of magnetic field on the plasma turbulence is neglected. Then,
the equivalent complex dielectric constants for a RH circularly polarized wave are readily obtained
by Ref. 29
p2 (t)/ p2 (t)( b ) p2 (t)
= 1 =

j
1 1 , (11)
( b ) j [( b ) 2 + 2 ] [( b )2 + 2 ]

where p (t) = [Ne (t)e2 ]/( 0 m) and Ne (t) are the fluctuant plasma resonant frequency and elec-
p

tron density. 0 is the permittivity of vacuum and is the angular frequency of EM wave. Using
Equation (11), the refractive and attenuation indices of these cyclotron waves can be prescribed
respectively as

1 p2 (t)( b )
(t) = Re( /2 ) =
1
1 
( b ) +
2 2

2


1/2 1/2
2 2
p2 (t)( b ) p2 (t)

1 *
+ .1   + 

+ , (12)
2 + 2 2 + 2
 /
2 ( b ) ( b )
, -

1 2 (t)( b )
(t) = Im( /2 ) = 1  p
_1



2
( b ) 2 + 2

1/2 1/2
2 2
p2 (t)( b ) p2 (t)

1
+ *. +

1  +/ , (13)
( b )2 + 2 ( b )2 + 2
  
2
, -
where Re[] indicates taking a real part, and Im[] indicates taking an imaginary part. For visually
illustrating the influences of fluctuation on the time-varying plasma, we assume that 2 are much less
than ( b )2 and ( b ) p2 . In this condition, the collision frequency can be neglected. As
known from Ref. 25, the collision frequency distributions vertical to the vehicle surface are regarded
as uniform. And the collision frequencies of plasma sheath under different altitudes are obtained from
Ref. 30. The altitude and velocity of RAM C-II reentry trajectory are introduced in Ref. 31. Thus,
the collision frequencies under typical altitudes or velocity and temperatures are list in Table I.
025114-6 Zhou et al. AIP Advances 7, 025114 (2017)

TABLE I. Collision frequencies (GHz) under typical altitudes or velocity and temperatures.

Temperature

Altitude (velocity) 2000K 3000K 4000K 5000K

21km (3.93km/s) 23.27 35.11 46.65 54.57

30km (6.55km/s) 6.21 8.79 12.45 14.12
40km (7.38km/s) 1.46 2.21 2.93 3.89
53km (7.62km/s) 0.1984 0.3088 0.4241 0.5121
61km (7.65km/s) 0.0602 0.0938 0.1247 0.1587
71km (7.65km/s) 0.0161 0.0242 0.0322 0.0389
76km (7.65km/s) 0.0074 0.0109 0.0149 0.0186

When the flight altitudes are in the range of 50km and 80km, the collision frequency satis-
fies the condition under which the collision frequencies can be ignored. Then, the expressions of
Equations (12) and (13) can be simplified as follows:
p2 (t) 1/2
(t) 1 , (14)
[( b )]

p2 (t)
(t) f . g 1/2 . (15)
2( b )3/2 b p2 (t)

The refractive index (4) has a zero (cutoff condition) for both RH waves
q
R (t) = b /2 + p2 (t) + b2 /4 , (16)

and a pole (resonance condition) at the cyclotron frequency

= b (17)

for RH wave. In general, an important feature, which can be derived from Ref. 29, is that, two
passpands exist for < b and > R in RH waves. However, R (t) varies with the fluctuant plasma
from Equation (6), that is, the passband > R has a time-varying property, which may impair
the effect of the magnetic window mitigation. Meanwhile, the phase is also affected by plasma
instability from Equation (14).

B. Simulation model of EM wave propagation in time-varying plasma

The simplified uniform time-varying magnetized plasma physical model is introduced in Section
to illustrate the fluctuation effects. This part presents a model for calculating the characteristics of EM
wave propagation in time-varying plasma. The propagation velocity of EM wave is approximately
equal to the speed of light and is much larger than the speed of electron density fluctuation. Parameters
can be considered as constants at a certain time when EM wave passes through the plasma sheath.
Thereby, the time-varying plasma can be divided into large numbers of static cell. For each cell,
the nonuniform plasma can be treated as several layers of homogenous thin plasma slabs. Then,
the transmission coefficients with time variation can be obtained by calculating the transmission
coefficient in each cell. The schematic of simulation model is shown in Fig. 3.
A method to solve the problem of the EM wave in nonuniform magnetized plasma is presented
to explore the transmission coefficient of EM wave for arbitrary time t n (n = 1, 2. . . N) in celli . The
geometry of a general multilayered magnetized plasma slab in the z direction is shown in Fig. 4
and the EM wave incident in xz plane. In the Fourier spectral domain, the tangential electric (Etn )
and magnetic (Htn ) fields for time t n satisfy a first-order differential equation within a source-free
025114-7 Zhou et al. AIP Advances 7, 025114 (2017)

FIG. 3. Schematic of stratified time-varying plasma simulation model.

FIG. 4. Geometry of EM wave incident on stratified magnetic plasma for each time t i . The left and right bounding interfaces
of each layer m are represented by Zm< and Zm> , and L m is the thickness of each layer m.

homogeneous medium32
E E
d tn = At tn .
n (18)
dz H tn H tn
f gT
Here, E tn , H tn = [Ex,tn , Ey,tn , Hx,tn , Hy,tn ]T ; Atn is a 44 matrix, whose elements are functions of
frequency, spectral variables, and time-varying magnetized plasma constitutive parameters; and sub-
script t n represents time term. The complex elements of Atn are given in Appendix A. Equation (18)
admits solutions in terms of the superposition of eigenwaves for each layer m:33
E
tn =
m, tn Pm, tn (z) cm, tn = m, tn wm, tn (z). (19)
H tn
Here, m, tn is a 44 eigenwave matrix comprising the eigenvectors; Pm, tn (z) is a diagonal matrix
whose diagonal elements are p(i) (i) (i)
m, tn (z) = exp(jkm, tn z) (i = 1, 2, 3, 4, and km, tn being the ith wavenumber
for m layer); and cm, tn is a 4 1 vector of unknown coefficients to be determined from the boundary
and/or radiation condition. The matrices and vectors in Equation (19) can be partitioned as
P > (z) 0
m, tn
Pm, tn =
<
, (20)
0 Pm, tn (z)

e> <
em,
m, t tn
m, tn = > n <
, (21)
hm, tn hm, tn

where the superscripts > and < stand for right-bounded and left-bounded decompositions.
Then, the relations among the field vector elements for each layer m can be defined from
Equation (18)34
E (z < ) H (z < ) Hm Hm
tn m = Hm tn m , Hm = 11, tn 12, tn
. (22)
t t
H tn (zm> ) n
E tn (zm> ) n
Hm 21, tn Hm
22, tn
025114-8 Zhou et al. AIP Advances 7, 025114 (2017)

The matrix Hm tn is called the layer hybrid matrix because it is a mixture of admittance, impedance,
and transfer elements. Using the eigenwaves in each layer, the layer hybrid can be determined as
> < P < (L ) > < < 1
em, em, hm, hm, tn Pm, tn (Lm )

tn tn m, tn m tn
Hm = > > < > > . (23)
ev<

tn
hm, tn m, tn (Lm )
P hm, tn em, tn Pm, tn (Lm )


In solving the multilayered problem, the stack hybrid matrix is defined as
E (z < ) H (z < )
tn l = H(l,m) tn l , (24)
tn
H tn (zm> ) E tn (zm> )

where H(l,m)
tn is the total stack hybrid matrix from layer l to layer m. Then, the overall hybrid matrix
(m,M)
Htn of the stack from m layer to M layer is deduced in Ref. 34.

M)
H(m,
11, tn
m (m+1,N)
11, tn + H12, tn H11, tn
= Hm [I Hm (m+1,N) 1
22, tn H11, tn ] Hm
21, tn
M)
H(m,
12, tn
= Hm (m+1,N)
12, tn [I H11, tn Hm
22, tn ]
1
H(m+1,N)
12, tn
M) . (25)
H(m,
21, tn
= H(m+1,N)
21, tn
[I Hm (m+1,N) 1
22, tn H11, tn ] Hm
21, tn
M) M)
H(m,
22, tn
= H(m+1,N)
22, tn
+ H(m+1,N)
21, tn
Hm (m+1,N)
22, tn [I H11, tn Hm
22, tn ]
1
H(m+1,
12, tn

For a special case, H(M,M)

tn is equal to HM (1,M)
tn , and using self-recursive algorithm, Htn can be obtained.
The reflection and transmission coefficients of EM waves in magnetized plasmas can be computed
simply via the hybrid matrix. Let the EM wave be incident on layer 0 and the external layer M+1 is
semi-infinite. The matrices of the coefficients of transmission and reflection can be defined as
g 1 f
Rtn = Hs, tn h0< e0< e0> Hs, tn h0> ,
f g
(26)

> (1,M) > 1 (1,M) > <

f g
Ttn = [hM+1, tn H22, tn eN+1, tn ] H21, tn h0, tn + h0, tn R tn , (27)

> >
Hs, tn = H(1,M)
11, tn
+ H(1,M)
12, tn
[hM+1 (eM+1 (1,M)
)1 H22, tn
] H(1,M)
21, tn
. (28)
Here, the media of layers 0 and M+1 are free space for which the eigenmatrices (Equation (21)) are
expressed in Appendix B.
From the results of Ref. 5, the coefficient of transmission for the RH waves is given considerable
attention. The electric fields of both RH and left-hand (LH) waves can be written in terms of a TE
plane wave and a TM plane wave.29 That is,

Er = (ETE + jETM ) / 2
. (29)
El = (ETE jETM ) / 2
For an incident RH wave, the electric field (ET , tn ) of transmission wave has the following form:
ET , tn = Tr, tn Er + Tl, tn El , (30)
where Tr, tn and Tl, tn are the coefficients of transmission for the RH and LH waves at time t n . Then,
substituting Equation (29) into Equation (30), we obtain

ET , tn = [(Tr, tn + Tl, tn )ETE + j(Tr, tn Tl, tn )ETM ]/ 2. (31)
 T
Moreover, with the unit RH wave vector defined as Er = 1, j , and using Equation (27), another
expression for ET is obtained,
" #
E
ET , tn = Ttn Er = T 1, tn . (32)
ET 2, tn
025114-9 Zhou et al. AIP Advances 7, 025114 (2017)

From Equations (22) and (23), the coefficients of transmission for the RH and LH waves can be
written as

Tr, tn = (ET 1, tn jET 2, tn )/ 2
. (33)
Tl, tn = (ET 1, tn + jET 2, tn )/ 2
Then, the time-varying transmission coefficient of RH wave in fluctuant magnetized plasma can be
expressed as
Tr (t) = [Tr, t1 , Tr, t2 , ..., Tr, tn ,..., Tr, tN ]. (34)

IV. NUMERICAL RESULT AND DISCUSSION

According to the typical data collected in RAM C-II by NASA Langley Research Center,35
the peak of electron density of the reentry plasma sheath is chosen as 1 1017 /m3 (71km) in
this part. The plasma sheath temperature is used 3000K according to the results of temperature
at antenna location in Ref. 36. Thus, the plasma collision frequency is considered to be 0.0242
GHz.30 The other parameters of plasma sheath are 1 =1000, 2 =500, zp =0.03 m, and zmax =0.12 m.35
To discuss the effect of fluctuation on multi-frequency transmission coefficients, the communica-
tion quality during reentry is considered affected by the attenuation of EM wave and the fluctuating
degree of EM wave amplitude and phase, except the fixed phase. Thereby, the mean value of ampli-
tude attenuation and the standard deviations of amplitude and phase are shown in the following
simulations.

A. Effect of fluctuation of plasma on EM wave propagation in magnetized plasma

Comparison of transmission coefficients between steady-state and fluctuating plasmas at different
standard deviations is discussed in this part to illustrate the effect of plasma fluctuation on EM wave
propagation in magnetized plasma. For the electron density variable quantities less than the peak
value,13 the standard deviation in Equation (7) can be chosen as 0.05, 0.15, and 0.25, and the
steady plasma can be treated as = 0. The electron gyro frequency corresponding to the magnetic
field induction is chosen as b = 5p .
As shown in Fig. 5, the transmission coefficients of EM wave propagation in both steady-state
and turbulent magnetized plasmas are plotted. As can be seen from Fig. 5(a), the mean values of
attenuation for fluctuating plasma at different are equal to the transmission coefficient EM wave
propagation in a steady-state plasma. Two passbands occur at low-frequency passband < b and
high-frequency passband > R . That is, low-frequency EM wave can pass through the plasma when
a certain degree of magnetic field is applied. The order of magnitude of attenuation of transmitted
EM wave cannot be influenced by the fluctuation. However, the fluctuations can give rise to parasitic
modulation, which may cause the communication failure both on the amplitude and phase of EM wave.
Moreover, the larger the turbulent ranges of amplitude and phase are, the worse the communication
quantity is.19 The standard deviations, which can reflect the fluctuation degree, of attenuation and
phase are presented in Figs. 5(b) and (c). In both Figs. 5(b) and (c), the standard deviation of EM
wave for steady-state plasma can be regarded as zeros. As can be seen from the plots, the standard

FIG. 5. Comparisons of transmission coefficients between propagation in steady-state and time-varying magnetized plasma.
(a) Mean value of amplitude attenuation, (b) standard deviation of amplitude attenuation, and (c) standard deviation of phase.
025114-10 Zhou et al. AIP Advances 7, 025114 (2017)

deviations of both amplitude attenuation and phase for fluctuating plasma are larger than 0 and
increase with the fluctuating deviation of plasma. The standard deviation of attenuation is less than
that of phase. That is, the fluctuation has more effects on the phase of EM wave than the amplitude.
In addition, the phase jitter is one of the most important factors in communication blackout in time-
varying plasma. Meanwhile, the attenuation and phase deviation increase with the EM frequency at
low-frequency passband and decrease with the increasing EM wave at high-frequency passband. In
particular, valley points appear at = b for each fluctuating plasma, as shown in Figs. 5(b) and (c).
This phenomenon can be explained by Equations (14) and (15) when = b , (t) and (t) , and
the effect of plasma fluctuation on (t) and (t) is decreasing. The error at = b for different plasmas
is produced by electron collision.9 In conclusion, the fluctuation of plasma also leads to the amplitude
turbulence and phase jitter of transmitted EM wave when a magnetic field is applied for blackout
mitigation.
B. Comparison of transmission coefficient between unmagnetized and magnetized
time-varying plasmas
The differences of EM wave propagation in both unmagnetized and magnetized plasmas are
discussed in this part to investigate the mitigation efficiency of magnetic window with time-varying
plasma. The peak of electron density is set to 1 1017 /m3 , and the time domain of the time-varying
transmission coefficients for the incident EM wave from 0.1p to 15p is obtained through the
combination of the quasi-static Monte Carlo method and hybrid matrix method in Section III B.
Fig. 6 illustrates the characteristics of EM wave propagation in unmagnetized and magnetized plasmas
with uniform magnetic fields b = p , b = 5p , and b = 10p .
Fig. 6(a) shows the mean value of amplitude attenuation of the transmitted EM wave. Again,
the mean attenuations of EM wave at < b in time-varying magnetized plasma are less than that
in unmagnetized plasma. These results are much similar to that in steady-state plasma. Thereby,
the deviation of amplitude attenuation and phase should be given considerable attention. The dis-
crepancies of standard deviation between unmagnetized and magnetized plasmas are investigated in
Figs. 6(b) and (c). When b = p for which the cut-off frequencies of both magnetized and unmagne-
tized plasmas are equal, the deviations of phase and attenuation in magnetized plasma is close to that
in unmagnetized plasma at low-frequency passband. Meanwhile, when b = 5p and b = 10p , the
standard deviations in magnetized plasma are less at < b /2 and larger at near b than that in
unmagnetized plasma. This result can be explained by rewriting Equations (14) and (15) as
p2 1/2
1
, (35)
[( b )]

p2 (t)
f . g 1/2 , (36)
2( b )3/2 b p2 (t)
where represent the turbulent quantity. When is near b , small p may cause the refractive and
attenuation indices to vary significantly. Moreover, when << b , the effect of p on refractive

FIG. 6. Comparisons of transmission coefficients between propagation in time-varying unmagnetized and magnetized plas-
mas. (a) Mean value of amplitude attenuation, (b) standard deviation of amplitude attenuation, and (c) standard deviation of
phase.
025114-11 Zhou et al. AIP Advances 7, 025114 (2017)

and attenuation indices can be neglected. That is, the effect of turbulence of plasma on EM waves
can be mitigated by using a certain magnetic field at low-frequency band. In addition, the larger
the strength of magnetic field, the better the elimination effect is. On the contrary, when > R in
high passband, the magnetic field intensifies the effect of turbulence on EM waves. A large magnetic
field causes drastic fluctuations. This result can be deduced from Equations (35) and (36). When
> R > b , | b | decreases with the increasing magnetic field, and the effect of p on
and is extremely enhanced. Therefore, the magnetic flied can deduce the fluctuation of plasma
on EM wave when < b /2, and the mitigation efficiency increases with increasing magnetic field
induction.
C. Effect of nonuniform magnetic field on EM wave propagation in time-varying
magnetized plasma
According to BiotSavarts Law (Equation (37)), which can be expressed as
0 Idl r
dB = , (37)
4 r 3
the magnetic induction intensity varies with distance from its source. A stop-band effect caused by
nonuniform magnetic field in steady-state plasma is found in Ref. 9. Thereby, the effect of nonuniform
magnetic field on EM wave propagation in time-varying plasma is investigated in this part. The
nonuniform magnetic field can be generated by several magnets. The forms of magnet are chosen as
rectangle, and the sizes are given as Magnet A (30 mm40 mm5 mm), Magnet B (150 mm150
mm100 mm), and Magnet C (300 mm200 mm100 mm). The axial direction outer magnetic field
distributions of permanent magnets are calculated by using the algorithm in Ref. 37 and shown in
Fig. 7.
The transmission coefficients versus frequency for different nonuniform magnetic fields in time-
varying magnetized plasma are shown in Fig. 8. Here, the mean values of amplitude attenuation are
plotted in Fig. 8 (a), (b) and (c) show the standard deviation of the amplitude and phase. As can be seen
from Fig. 8(a), wide range stopbands are also formed versus different nonuniform magnetic fields in
time-varying magnetized plasma. The locations and widths of stopbands on frequency domain has
been determined in previous study. That is, b,min (the minimum b ) and R,max (the maximum R )
determine the left and right edges of the stopband.9 The stop-band effect also exists in standard
deviation of amplitude attenuation and phase, as shown in Figs. 8 (b) and (c). Comparing Figs. 8 (a)
and (b), the cut-off frequencies of both mean value and standard deviation of amplitude attenuation are
consistent. The standard deviation of phase shows two significant peaks for each different nonuniform
magnetic fields. Moreover, the frequencies corresponding to the peaks are also the cut-off frequency
in Fig. 8(a). The standard deviations in Figs. 8 (b) and (c) both decrease at < b,min and increase at
> R,min with the increasing magnetic field. These properties are similar to that when the magnetic
field is uniform. Thus, combining the two stopbands of amplitude attenuation and phase standard
deviation, we can find that the nonuniform magnetic field causes the frequency range of large deviation

FIG. 7. Variation of axial magnetic induction intensity with distance from vehicle surface.
025114-12 Zhou et al. AIP Advances 7, 025114 (2017)

FIG. 8. Characteristics of EM waves in time-varying magnetized plasma for different nonuniform magnetic fields. (a) Mean
value of amplitude attenuation, (b) standard deviation of amplitude attenuation, and (c) standard deviation of phase.

to widen. The effect of magnetic field on eliminating the fluctuation of the transmission coefficient
may become worse.

V. CONCLUSION
Magnetic window method is used to eliminate the communication blackout problem caused by
reentry plasma sheath. Few studies focus on the influence of the fluctuation of plasma. In this study,
the characteristics of EM wave propagation in time-varying magnetized plasma are investigated.
An electron density fluctuation model is introduced, and the transmission coefficients are calculated
under this condition. LEM is used for simulation of plasma electron density spatial turbulence, and the
time domain variation is generated as Gaussian random process. Then, the effects of magnetic field
on EM wave propagation in fluctuant plasma are numerically investigated by using the time-varying
hybrid matrix method. The simulation results show that the fluctuation of plasma can also result in
amplitude variation and phase jitter of EM wave when a magnetic field is applied, and the fluctuating
standard deviation increases with the turbulent degree of plasma. Comparing the characteristics of EM
propagation with a time-varying unmagnetized plasma, the standard deviations of transmitted EM
wave amplitude and phase attenuate at low-frequency passband < b /2. Conversely, in the other
passband > R , the fluctuation of transmission coefficients is enlarged by the magnetic field. That
is, the fluctuations of amplitude and phase caused by the instability of plasma sheath can be mitigated
by the applied magnetic field when the EM wave frequency is less than b /2. Fluctuation mitigating
effect also improves with the increasing magnetic field. Furthermore, the nonuniformity of magnetic
field can also affect the mitigating effect of turbulence. The stopband effect is found not only in the
steady-stated magnetized plasma, but also in the time-varying one. That is, wide turbulent ranges of
transmission coefficients appeared on both standard deviation of phase and amplitude attenuation.
Moreover, the fluctuating elimination effect in pass band < b is attenuated. Therefore, applying
a uniform magnetic field may eliminate the influence of fluctuation of plasma sheath on EM wave
propagation.
Although progress has been made and the conclusion has important significance to blackout
problem, further work should be done in the future. In this study, the time-varying plasma sheath
is considered as a lossy anisotropic medium. The fluctuation of plasma sheath is caused by vehicle
attitude adjustment, ablation, and turbulence disturbance. The fluctuation caused by magnetic field
is neglected. Given that the magnetic field can change the track of electron motion, investigating the
effect of magnetic field on the fluctuation of plasma in succeeding work is necessary. A study on
the magnet with uniform magnetic field for improving the magnetic window in blackout mitigation
should be also done in future works.

ACKNOWLEDGMENTS
This work was supported by the National Basic Research Program of China under Grant
2014CB340204, by the National Natural Science Foundation of China under Grants 61473228 and
61601353, and the Natural Science Foundation of Shaanxi Provincial Department of Education under
Grant 2016JM6016.
025114-13 Zhou et al. AIP Advances 7, 025114 (2017)

APPENDIX A: COEFFICIENT MATRIX Atn

The complex matrix Atn can be given as
Atn = [Mtt, tn + (Mtz, tn Ktz, tn ) M1
zz, tn (Kzt, tn Mzt, tn )], (A1)
where
0 0 0 1
0 0 1 0
= ,
0 1 0 0
1 0 0 0
xx xy 0 1
yy 1 0
"
0
#
Mtt = yx , Mzz = zz ,
0 1 0 0
1 0 0
(A2)
xz, tn 0
0
"
zy, tn 0 0
#
Mtz = yz, tn , Mzt = zx, tn ,
0 0 0 0 0 0
0 0
0 0 " #
1 0 kx, tn 1 0 0 0 kx, tn
Ktz = , Kzt = .
0 0 0 kx, tn 0 0
kx, t
n 0
In Equation (A2), kx, tn is the x-component of wave vector propagation k; is the permeability of
plasma, which can be assumed as 0 ; and tn is the permittivity tensor of time-varying magnetized
plasma. tn can be expressed as
xx, t j xy, t 0
n n
tn = j yx, tn yy, tn 0 , (A3)
0 0 zz, tn
where
(p2 (t)/2 )(1 j/)
= =


xx, t yy, t 1 ,
(1 j/)2 b2 /2


n n




(p2 (t)/2 )(b /)





= = j , (A4)
xy, t yx, t
(1 j/)2 b2 /2
n n





p2 (t)/2



zz, tn = 1

,


1 j/

APPENDIX B: EIGENMATRIX OF FREE SPACE

The eigenmatrices (Equation (21)) of layers 0 and M+1 can be written as
e> e< e> <
eM+1
M+1
0 = 0> 0
=
h0 h0< > <

hM+1 hM+1
kz 0 kk0z 0
k0 . (B1)
0 1 0 1
= 0 kz
00 kk0z
q q
q0 0

0 k0
0 0
q
0 0 0 0

025114-14 Zhou et al. AIP Advances 7, 025114 (2017)

where 0 , 0 , and k0 are the permittivity, permeability, and wave vector of free space, respec-
tively.

1 J. P. Rybak and R. J. Churchill, IEEE Trans. Aero Elec Sys. 7(5), 879894 (1970).
2 W. F. Hinson, P. B. Gooderum, and D. M. Bushell, NASA Tech. Rep. TN D-5861 (NASA, Washington, DC, USA, 1970).
3 G. Rosen, Phys. Fluids 5(6), 737738 (1962).
4 M. Kim, M. Keidar, and I. D. Boyd, Journal of Spacecraft and Rockets 45(6), 12231229 (2008).
5 H. Hodara, Proc. IRE 49(12), 18251830 (1961).
6 M. Laroussi and J. R. Roth, IEEE Trans. Plasma Sci. 21(4), 366372 (1993).
7 C. Thoma, D. V. Rose, C. L. Miller, R. E. Clark, and T. P. Hughes, J. Appl. Phys. 106, 043301 (2009).
8 X. Yin, H. Zhang, S. Sun, Z. Zhao, and Y. Hu, Prog. Electromagn. Res. 137, 159186 (2013).
9 H. Zhou, X. P. Li, Y. M. Liu, B. W. Bai, and K. Xie, IEEE. Trans. Plasma. Sci. 45(1), 1523 (2017).
10 A. Demetriades and R. Grabow, AIAA J. 9(8), 15331538 (1971).
11 L. G. Zang, M. Takeuchi, N. Nishino, T. Mizuuchi et al., Plasma Sci. Tech. 15(3), 213216 (2013).
12 Y. Bao, X. P. Li, and L. Shi, Proceeding of the Chinese Society for Optical Engineering Conferences Held July 2016,

Changchun, China, 24 July 2016, edited by Y. lv, W. Bao and G. Zhang, pp. 101410J-1.
13 L. Shi, Y. M. Liu, S. X. Fang, X. P. Li, B. Yao, L. Zhao, and M. Yang, IEEE. Trans. Plasma. Sci. 44(7), 10831093 (2016).
14 T. C. Lin and L. K. Sproul, Comput. Fluids 35(7), 703711 (2006).
15 Z. W. Liu, W. M. Bao, X. P. Li, D. L. Liu, and B. W. Bai, IEEE Trans. Plasma Sci. 43( 9), 31473154 (2015).
16 G. L. He, Y. F. Zhang, J. Zhou, and N. Ge, IEEE Trans. Plasma Sci. 44(3), 232238 (2016).
17 M. Yang, X. P. Li, K. Xie, Y. M. Liu, and L. Shi, J. Astronaut. 34(6), 842847 (2013).
18 M. Yang, X. P. Li, Y. M. Liu, L. Shi, and K. Xie, Acta. Phys. Sin. 63(8), 085201 (2014).
19 M. Yang, X. P. Li, K. Xie, and Y. M. Liu, Phys. Plasmas 22(2), 022120 (2015).
20 D. K. Kalluri, V. R. Goteti, and A. M. Sessler, IEEE. Trans. Plasma. Sci. 21(1), 7076 (1993).
21 D. K. Kalluri and T. T. Huang, IEEE. Trans. Plasma. Sci. 26(3), 10221030 (1998).
22 Y. Bo, Q. Zhao, X. Luo, J. Fan, Y. Liu, and J. Liu, Acta. Phys. Sin. 65(5), 055201 (2016).
23 R. Rawhouser, in Proceeding of the Entry Plasma Sheath Effects Space Vehicle Electromagn. Syst., NASA Langley Res.

Center, Washington, DC, USA, Oct. 1970, pp. 317, NASA SP-252.
24 M. Kundrapu, J. Loverich, K. Beckwith, P. Stoltz et al., J Spacecraft Rockets 52(3), 813862 (2015).
25 B. Bai, X. P. Li, Y. M. Liu, J. Xu, L. Shi, and K. Xie, IEEE. Trans. Plasma. Sci. 42(10), 33653371 (2014).
26 F. K. Owen, C. C. Horstman, and M. I. Kussoy, J. Fluid. Mech. 70(2), 393413 (1975).
27 M. Estorf, R. Radespiel, S. Schneider, H. Johnson, S. Hein, in Proceeding of the AIAA Aerospace Science Meeting and

Exhibit, Reno, Nevada, USA, Jan. 2008, pp. 20081153.

28 A. R. Kerstein, J. Fluid Mech. 392, 277334 (1999).
29 M. A. Heald and C. B. Wharton, Plasma Diagnostics with Microwaves(Wiley, 1965), pp. 7194.
30 R. A. Hartunian, G. E. Stewart, S. D. Fergason, T. J. Chrtiss and R. W. Seibold, Causes and Mitigation of Radio Frequency

(RF) Blackout During Reentry of Reusable Launch Vehicles, Aerospace Report No. ATR-2007(5309)-1, Jan, 2007.
31 W. L. Grantham, Reentry Plasma Measurement Using a Four-frequency Reflectometer, NASA, Washington, D C, Tech.

Rep. NASA/SP-252, pp. 65107, Oct, 1970.

32 D. W. Berreman, J. Opt. Soc. Amer. 62(4), 502510 (1972).
33 P. G. Cottis and G. D. Kondylis, IEEE Trans. Antennas Propagation 43(2), 154161 (1995).
34 N. Jing and E. L. Tan, Mater. Design 30(6), 20612066 (2009).
35 See National Technical Information Service Document No. AD718428 (J. P. Rybak, Scientific Report No. 1, 1970). Copies

may be ordered from the National Technical Information Service, Springfield, VA 22151.
36 G. L. He, Y. F. Zhan, N. Ge, Y. K. Pei, and B. Wu, IEEE. Trans. Plasma. Sci. 42(12), 39753981 (2014).
37 X. Guo, Y. Yang, and X. Zheng, Appl. Math. Mech. 25(3), 271278 (2004).