Progress In Electromagnetics Research B, Vol.

37, 2142, 2012

AN INTERACTION STUDY BETWEEN PIFAS HANDSET ANTENNA AND A HUMAN HAND-HEAD IN PERSONAL COMMUNICATIONS
A. M. Montaser1 , K. R. Mahmoud2, * , and H. A. Elmikati3
1 Sohag

University, Sohag, Egypt

2 Department

of Electronics and Communications Engineering, Faculty

of Engineering, Helwan University, Cairo, Egypt
3 Mansoura

University, Mansoura 35516, Egypt

AbstractIn this paper, the interaction of a planar inverted-F

antennas array, mounted on a mobile handset, with a human hand-head
phantom is investigated in the 1.9 GHz band. The hybrid approach
involving the particle swarm optimization (PSO) and Nelder-Mead
(NM) algorithm is considered to optimize the complex excitations of
the adaptive array elements in a mutual coupling environment for
different beamforming synthesis. Firstly, the effect of the human handhead on the handset radiation characteristics is studied. Then, the
spatial-peak specific absorption rate (SAR) values of 2- and 4-element
PIFA arrays for mobile handset in the vicinity of a human handhead are evaluated numerically for different scenarios. The antenna
is analyzed completely using finite difference time domain (FDTD)
method while the interaction is performed using the CST Microwave
Studio software.
1. INTRODUCTION
Multiple input and multiple output (MIMO) antenna systems are
seen as one of the potential technologies for enhancing performance
and capacity of future wireless communication systems [15]. They
are based on utilizing multiple antennas for both transmission
and reception of signals and can allow performance and capacity
enhancement without the need for additional power or spectrum [1
3]. Wireless communication devices have also become associated
Received 15 September 2011, Accepted 25 November 2011, Scheduled 1 December 2011
* Corresponding author: Korany Ragab Mahmoud
(KURANY HAMEDA@h-eng.helwan.edu.eg).

22

Montaser, Mahmoud, and Elmikati

with possible health effects surrounding the exposure of humans

to electromagnetic radiation [6, 7]. One of the most widely used
parameters for the evaluation of exposure is the specific absorption rate
(SAR). Therefore, some regulations and standards have been issued
to limit the radiation exposure from the mobile handsets not only to
decrease the SAR but also to increase the antenna systems efficiency.
The SAR quantifies the power absorbed per unit mass of tissue. This
quantity is defined as:

SAR =
|Ei |2
(1)
2
where Ei is the max value of the electric field strength in the tissue in
V/m, is the conductivity of body tissue in S/m, and is the density
of body tissue in kg/m3 . The SAR limit specified in IEEE C95.1: 2005
has been updated to 2 W/kg over any 10-g of tissue [8]. This new
SAR limit specified in IEEE C95.1: 2005 is comparable to the limit
specified in the International Commission on Non-Ionizing Radiation
Protection (ICNIRP) guidelines [9]. In designing antennas for mobile
communications, it is important to investigate the SAR value produced
by the radiation from the mobile handsets.
Since the use of adaptive array antennas is one of the most
effective ways of increasing the spectral efficiency and improving the
communication quality in the mobile communication system, a great
deal of attention has turned to how to design antenna arrays on mobile
handsets to meet the needs of significantly higher bit-rates for the nextgeneration wireless communications [1012]. Many studies on the SAR
value of a single antenna for mobile handsets have been carried out in
different frequency bands [1326]. In [27], an investigation of the SAR
caused by two-element antenna array at 1.9 GHz has been reported
and it was shown that the SAR value reaches a maximum when the
phase difference is near 180 degrees and a minimum when the phase
difference is approximately 0 degree. The effect of the human head on
the accuracy and depth of adaptive nulling for a linear dipole array at
5.0 GHz band is investigated in [28]. This effect is dependent on the
separation between the array and the head. A metallic plate is inserted
between the array and the head to suppress this effect. In [29], the peak
SAR value at 2 GHz and 5 GHz is evaluated numerically as a function
of the distance between the array antenna and a spherical head model
when the two elements of an antenna array are voltage-fed co-phase or
reverse-phase.
A study of how the SAR is affected by different shapes and
electrical properties of the human head exposed to a cellular phone
have been reported [30]. The results showed that the shape of the
human head plays a minor role in calculating the SAR induced in

Progress In Electromagnetics Research B, Vol. 37, 2012

23

human head models. In [31], the effect of the human head heterogeneity
and shape on the radiation characteristics of dipole antennas has been
discussed at 5.0 GHz band. It was found that the input impedance
and the radiation pattern of the dipole antenna are not sensitive to
the heterogeneity and the shape of the head model.
In [32], the capability of a 5-element monopole array into a
handheld device for 4G communication systems in the 5.0 GHz band
for beamforming synthesis in a mutual coupling environment is
investigated. The interaction of the antenna array, mounted on a
mobile handset, with a human head phantom is studied. The spatialpeak specific absorption rate (SAR) values of the smart handset in
the vicinity of a spherical phantom of a human head are evaluated
numerically as a function of the distance between the handset and the
head phantom. It is found that, the resulting SAR values for the smart
handset in different scenarios were under the limits set by IEEE C95.1:
2005 or ICNIRP standards.
In this paper, the handset is analyzed completely using CST
Microwave Studio, then the CST is linked with the PSO-NM algorithm,
Matlab-coded, to optimize the elements weight.
Interchanging
information between CST Microwave studio and Matlab allows
the implementation of mathematical or optimization algorithms
and graphical possibilities not included in the Microwave studio
environment itself. The approach developed here is very flexible and
utilizes the inherent capability of Microwave studio to execute Visual
Basic for Applications (VBA) programs. The general idea of this
method is that Matlab and CST Microwave studio can interchange
information using external text files, which can be read and written
by both programs [33]. The PSO-NM algorithm is used to adjust
the relative phase shifts and the amplitudes of the excitations of the
array elements for beam synthesis in a mutual coupling environment.
In particular, we evaluate the potential of a 2- and 4- element PIFA
arrays incorporated into a handheld device for beamforming in the
1.9 GHz band. Also, the interaction of the antenna array, mounted on
a mobile handset, with a human hand-head phantom is investigated.
In addition, the spatial-peak SAR values of 2- and 4-element antenna
arrays for mobile handset in the vicinity of a human hand-head are
evaluated numerically for different scenarios. The total power delivered
to all antenna arrays for all cases is held constant (100 mW).
The organization of the present paper is as follows: In Section 2, a
brief introduction to the hybrid particle swarm optimization and nelder
mead (PSO-NM) algorithm is presented. In Section 3, handset design
and simulation results for two and four elements PIFA array mounted
on the mobile handset are discussed. The interaction between the

24

Montaser, Mahmoud, and Elmikati

handset and the human hand-head is investigated in Section 4. Finally,

Section 5 presents the conclusions.
2. PARTICLE SWARM OPTIMIZATION NELDER
MEAD (PSO-NM) ALGORITHM
PSO is a population based optimization tool, where the system is
initialized with a population of random particles and the algorithm
searches for optima by updating generations [34]. Suppose that the
search space is D-dimensional. The current position of the i-th particle
can be represented by a D-dimensional vector Xi = (xi1 , xi2 , . . . , xiD )
and the velocity of this particle is Vi = (vi1 , vi2 , . . . , viD ). The
best previously visited position of the i-th particle is represented by
Pi = (pi1 , pi2 , . . . , piD ) and the global best position of the swarm found
so far is denoted by Pg = (pg1 , pg2 , . . . , pgD ). The fitness of each
particle can be evaluated through putting its position into a designated
objective function. The particles velocity and its new position are
updated as follows:

k+1
k
(2)
vid
= vid
+ c1 r1k pkid xkid + c2 r2k pkid xkid
k+1
xkid = xkid + vid
t

(3)

where d {1, 2, . . . , D}, i {1, 2, . . . , N }, N is the population size,

the superscript k denotes the iteration number, is the inertia weight,
r1 and r2 are two random values in the range [0, 1], c1 and c2 are the
cognitive and social scaling parameters which are positive constants
(c1 = 2.8 and c2 = 1.3). In this paper, the PSO algorithm is employed
with a population size of 20 and 50 iterations. For beamforming
synthesis, the amplitude was allowed to vary between 1.0 and 3.0 and
the phase was allowed to vary between and .
A very simple objective function for the antenna array is used
for maximizing the output field toward the desired signal at i and
minimizing the total output field in the direction of the interfering
signals at j .
XN
XM
Objective-function =
ai |E(i )|
bj |E(j )|
(4)
i=1

j=1

where E() is the total electric field and the constants ai and bj are
the weights that control the contribution from each term to the overall
objective function. The constants N and M represent the number of
desired signals and interferers respectively. In our analysis, we take
the weights ai = 2 and bj = 1 to give some priority to maximizing
the output field toward the desired signal rather than minimizing the
output field in the direction of the interfering signals.

Progress In Electromagnetics Research B, Vol. 37, 2012

25

Once the global optima using PSO algorithm is completed; the

Nelder-Mead (NM) local optimization technique is followed to fine
optimize the results. The NM method thus falls in the general class of
direct search methods. It is based on the comparison of the function
values at the (D + 1) vertices for D-dimensional decision variables.
Selection of these points can be prescribed, but random selection allows
the potential to fully investigate the merit space [35].
3. HANDSET DESIGN AND SIMULATION RESULTS
The geometry of the PIFAs and printed circuit board (PCB) layout
for our prototype are shown in Fig. 1. Fig. 1(a) shows the geometry of
the 2-PIFA antenna elements and Fig. 1(b) for the four elements PIFA
array. The dimensions of the PCB are (l = 110 mm, w = 50 mm)
while the size of the PIFAs are PL = 34 mm, Pw = 10 mm with heights
of h = 6 mm and PIFA separation, d = 20 mm. The FDTD method
written with MATLAB software [36] is used to design a two and four
elements PIFA array and the results are compared with that obtained
using CST Microwave Studio. The parameters for FDTD computation
were set as follows for 2-PIFA array: the domain was 147 100 20
cells with a cell size of x = 0.75 mm, y = 0.5 mm, z = 0.5 mm.
While for 4-PIFA elements, the domain was 147 100 40 cells with
the same cell sizes used for 2-PIFA elements. The FDTD lattice
needs to be terminated by perfectly matched layer (PML) on all sides;
a spatially varying conductivity should be used in order to avoid
numerical reflections at the interface of FDTD/PML regions. For two

(a) 2-PIFA elements

(b) 4-PIFA elements

Figure 1. Geometry of the two and four element PIFA array.

26

Montaser, Mahmoud, and Elmikati

antenna structures, the computational domain was terminated with

perfectly matched layer (PML) of 8 cells in all directions with a time
step of 1.452 ps. For the CST Microwave Studio simulator, which based
on the Finite Element Method (FEM), the following settings were
used for time domain simulations: the minimum mesh step = 0.535,
maximum mesh step = 12.637 and the mesh cells = 22.185 (N x = 52,
N y = 30, N z = 16). The mesh line ratio limit was set to 50 with an
equilibrate mesh ratio of 1.19. Open add space boundary condition
is applied in all directions with thermal boundaries isothermal (T =
constant).
The primary difference between the FDTD and the FEM method
is that, the FDTD discretize the problem into small rectangular
cells, these cells possible allocation of the six electric and magnetic
components are located at the edges and on the surfaces of the
cell [36]. While in the FEM, the problem space is discretized into
small shapes which results in a mesh with grid nodes, the solution can
be approximated with a simple function, the so-called shape function,
in each element, which acts as a contribution to the approximation of
the global solution.
Figures 2(a), (b), (c), (d) show the results comparison between
the FDTD method, the EM Simulator of CST Microwave Studio
software, and measurement results [27] for S parameters (S11 , S12 ,
S21 , S22 ) for two elements PIFA array. In general, good agreement
is seen between the computed values and measurements up to the
readability of the figures in [27]. There are some differences between
the calculated results due to the different applied numerical techniques
(FDTD and FEM). Compared to the measured results, the FDTD
results are slightly more accurate than the CST MWS results. From
Fig. 2, it can be observed that both PIFA antennas achieve S11 and
S22 of less than 10 dB around 1.9 GHz with a mutual coupling less
than 9 dB. Although mutual coupling of 9 dB is not insignificant
it is generally accepted that envelope signal correlations between
the antennas in realistic mobile environments only needs to be less
than 0.5 [2, 4, 5]. Using the bound e < 25|S12 |2 (which is valid if the
magnitude of the S-parameters satisfy the condition |S11 | + |S22 | +
|S11 ||S22 | + |S12 ||S21 | < 0.25 and is also satisfied in our results) it
can be shown that e < 0.39 indicating that sufficient diversity would
be achieved [2, 4, 5]. Figs. 3(a), (b), (c), (d) show the comparison
between the simulated results using the FDTD method and the CST
Microwave Studio software for S parameter (S11 , S21 , S31 , S41 ) for
4-elements PIFA array. It is found that, the 4-PIFA antenna array has
a resonance frequency of 1.9 GHz with a return loss less than 11.5 dB
with acceptable coupling between antenna elements.

Progress In Electromagnetics Research B, Vol. 37, 2012

(a) S 11

(b) S 12

(c) S 21

(d) S 22

27

Figure 2. The comparison between simulated results using the FDTD

method, CST microwave studio software and measurement results for
S parameter (S11 , S12 , S21 , S22 ) for two elements PIFA array.
The PSO-NM algorithm has been programmed in Matlab in
connection with the CST simulator for the calculation of the fitness
value of each particle. The process for communicating CST and Matlab
is as follows: First, using PSO-NM as it is described in Section 2,
the values for the parameters to be optimized are generated using
Matlab. These values are saved in a text file. A Visual Basic for
Applications macro (VBA) is the most straightforward way to run the
CST simulation from Matlab. Second, Matlab calls CST and a VBA
macro loads the data from the text file and stores the values in CST
variables. Then, a new macro is called for creating the structure of
the proposed device according to the values of the variables previously
loaded and starts the simulation. When the simulation finishes, the
results for radiation pattern will be saved in a new text file and CST
is closed. Finally, Matlab reads the radiation pattern results from the
file and the fitness function is calculated. This process is repeated for
each particle. Fig. 4 shows a flowchart diagram of the main steps to
link the Matlab with the CST simulator.

28

Montaser, Mahmoud, and Elmikati

(a) S 11

(b) S 21

(c) S 31

(d) S 41

Figure 3. The comparison between simulated with FDTD method,

CST Microwave Studio software for S parameter (S11 , S21 , S31 , S41 )
for four elements PIFA array.
Table 1. Descriptions of the environmental scenarios.
Scenario #1
Desired
Interference

0
90
180 270
Scenario #3
Desired
Interference

120
0
90 270

Scenario #2
Desired
Interference

30
90
180 270
Scenario #4
Desired
Interference

210
0
90 270

As an example of adaptive beamforming, four scenarios were

considered as shown in Table 1. Fig. 5 shows the capability of the
2-element array geometry to direct the maximum field towards the
direction of Signal Of Interest (SOI) while placing deeper nulls towards
the angles of Signal Not Of Interest (SNOI) for different scenarios. The
required amplitude and phase excitations of each element to obtain

Progress In Electromagnetics Research B, Vol. 37, 2012

29

MATLAB
Implement the PSO-NM algorithm
to create the feeding variables and
write them to text file then
Run VBA program

VBA Program
Reads the variables from text file
and Write it to MWS file
Interfaces to MWS

MWS
Simulate the antenna and performs
the fitness computations then Write
the data to file

VBA Program
Saves everything

MATLAB
Reads the file which
contains the fitness

Iteration itermax

Stop

Figure 4. Flowchart showing the main steps to link the Matlab with
the CST simulator.
the beam patterns in Fig. 5 are shown in Table 2. It is found that
an acceptable gain of 4.12, 3.85, 4.35, and 4.55 dBi are obtained in
the direction of SOI for different scenarios 1, 2, 3, and 4 respectively.
Similarly, the same scenarios are applied to the 4-elements PIFA array
as shown in Fig. 6. The required amplitude and phase excitations of
each element to obtain the beam patterns in this figure are shown in
Table 3. The calculated gain for different scenarios (1, 2, 3, 4) are

30

Montaser, Mahmoud, and Elmikati

90

90
4

120

60

120

60

2
150

30

150

30
1

180

180

(V/m)

(V/m)

Scenario# 2
Scenario# 1

210

240

330

Scenario# 3
Scenario# 4

210

300

300

240

270

330

270

SNOI directions

SOI directions)

(a)

(b)

Figure 5. The radiation pattern |E| for the two element PIFA array.
( SNOI directions SOI directions).
90

90

60

120

120

60

30

150

150

30

180

180

(V/m)

(V/m)

Scenario# 1
Scenario# 2

210

330

240

300

240

330

300
270

270

Scenario #3
Scenario #4

210

SNOI directions

SOI directions)

(a)
(b)
Figure 6. The radiation pattern |E| for the four element PIFA array.
( SNOI directions SOI directions).

found to be 4.7, 4.81, 4.67, and 4.46 dBi respectively. It is found that
for a uniform feeding PIFA array (scenario # 0) where all elements
are excited with the same amplitude and phase (2, 0 ), the gain of the
4-PIFA elements is slightly higher than that of the 2-PIFA elements by

Progress In Electromagnetics Research B, Vol. 37, 2012

31

Table 2. The required amplitude and phase excitations of each

element for 2-element PIFA array for different scenarios with the
corresponding gain.
Element No.
1
2
Element No.
1
2

Scenario #1
amplitude
phase
2.97
75.728
2.55
47.33
Scenario #3
amplitude
phase
1.73
73.25
2.05
11.13

Scenario #2
amplitude
phase
2.61
81.19
2.49
116.71
Scenario #4
amplitude
phase
2.65
97.44
1.93
129.94

Table 3. The required amplitude and phase excitations of each

element for 4-element PIFA array for different scenarios with the
corresponding gain.
Element No.
1
2
3
4
Element No.
1
2
3
4

Scenario #1
amplitude
phase
2.41
68.025
1.72
128.97
1.64
135.77
2.72
8.974
Scenario #3
amplitude
phase
2.76
24.21
2.44
23.20
1.64
5.272
1.23
156.84

Scenario #2
amplitude
phase
2.98
94.096
1.97
145.456
1.76
139.96
2.98
10.981
Scenario #4
amplitude
phase
1.72
40.671
1.46
1.996
2.85
95.573
1.04
4.568

0.32 dB. Generally, the obtained gain for different scenarios (1, 2, 3, 4)
is found to be higher than that of the uniform feeding case (scenario
# 0); due to directing the power to only one direction (Base Station).
4. INTERACTION BETWEEN HANDSET AND HUMAN
HAND-HEAD
In this section, the interaction between the mobile handset and the
human hand-head has been studied. The described antennas in

32

Montaser, Mahmoud, and Elmikati

Section 3 are now covered with a dielectric material (r = 2.1), having

external dimensions of 113.6 53.6 9.6 mm for 2-PIFA antenna
elements and 113.6 53.6 15.6 mm for 4-PIFA case. The SAM
phantom head and hand that provided by CST Microwave Studio is
used [37]; the tissue that it contained had a relative permittivity and
conductivity as shown in Table 4. These tissue-equivalent dielectric
parameters were chosen according to [38] for simulating brain tissue at
1.9 GHz. The head phantom was finely meshed to give voxels of 8 mm3 ,
corresponding to a sampling mesh of 2 mm along each Cartesian axis.
The final number of mesh-cells sums up to be around 11 Million cells.
The relative position of the handset relative to human head model is
illustrated in Fig. 7. The interaction between the mobile handset and
the human hand-head is studied from two viewpoints, the effect of the
human hand-head on the handset radiation characteristics in different
scenarios and the impact of the handset on SAR.
Firstly, we will study the effect of the human hand-head on the
handset radiation characteristics. As shown in Fig. 7, the handset has
Table 4. The conductivity and relative dielectric constant at 1.9 GHz
for the head model.
Tissue
Avg. Brain
Avg. Skull
Avg. Muscle
Palm (Skin) Dry

(a) Front view

Permittivity
43.374870
15.465057
54.304787
38.714

Conductivity
1.203850
0.456124
1.447971
1.2245

(b) Side view

Figure 7. Human hand-head model and handset configuration.

Progress In Electromagnetics Research B, Vol. 37, 2012

33

placed in parallel tendency according to the situation palm of the hand

to simulate the actual status of the handset user. The distance between
the nearest points for palm hand and the human head model is 5 mm.
Fig. 8 shows the handset radiation pattern in the presence of the SAM
phantom head and hand that provided by CST Microwave Studio for
both 2- and 4-PIFA handsets. It is noted that, the radiation pattern
is affected by the presence of the human hand-head. Therefore, the
beam patterns need to be re-optimized in the presence of the human
hand-head.
90

90

2.5

2.5

60

120

60

120

1.5

1.5

30

150

30

150
1

0.5

0.5

180

180

(V/m)

(V/m)

Scenario# 1
Scenario# 2

210

240

Scenario# 3
Scenario# 4

210

330

240

300

330

300
270

270

(a) Two elements PIFA array mounted on the mobile handset

90

90

60

120

120

60

3
4

150

30

30

150
2

180

180

(V/m)

(V/m)

210

Scenario# 2
Scenario# 1

240

300
270

330

210

Senario #4
Senario #3

240

330

300
270

(b) Four elements PIFA array mounted on the mobile handset

Figure 8. The radiation pattern |E| of the handset in the presence of

the human hand-head phantom (pre-optimized beam pattern).

34

Montaser, Mahmoud, and Elmikati

Figure 9 shows the re-optimized beam patterns of the handsets in

the presence of the SAM phantom human hand-head for scenarios 1,
2, 3 and 4. The figure shows the ability of the design to direct the
maximum field towards the direction of SOI while placing deeper nulls
towards the angles of SNOI in the presence of the human hand-head
for different scenarios. The required amplitude and phase excitations
of each element to obtain the beam patterns for two and four elements
PIFA array on a handset in this figure are shown in Tables 5 and 6,
respectively.
90

90
6

120

60

120

60

3
4

150

30

150

30

80

180

(V/m)

(V/m)

Scenario# 1
Scenario# 2

210

240

330

Scenario# 4
Scenario# 3

210

300

240

270

330

300
270

(a) Two elements PIFA array mounted on the mobile handset

90

90

6
120

120

60

60

4
150

150

30

30
2

180

180

(V/m)

(V/m)

Scenario# 2
Scenario# 1

210

300

240
270

330

Scenario# 3
Scenario# 4

210

240

330

300
270

(b) Four elements PIFA array mounted on the mobile handset

Figure 9. The radiation pattern |E| of the handset in the presence of
the human head phantom (optimized beam pattern).

-5

-5

-10

-10

Return Loss S11 (dB)

Return Loss S11 (dB)

Progress In Electromagnetics Research B, Vol. 37, 2012

-15
-20
-25
-30
Scenario #1
Scenario #2

-35

35

-15
-20
-25
-30
Scenario #0 (Uniform feeding handset)
Scenario #1

-35

Scenario #2

Scenario #3
Scenario #4

-40

Scenario #3

-40

Scenario #0 (Uniform feeding handset)

-45
1.5

1.6

1.7

1.8

1.9

2.1

2.2

2.3

2.4

Scenario #4

-45
1.5

2.5

1.6

1.7

1.8

Frequency (GHz)

1.9

2.1

2.2

2.3

2.4

2.5

Frequency (GHz)

(a) 2-PIFA handset in free space

(b) 4-PIFA handset in free space

0

0
-2

-5
Return Loss S11 (dB)

Return Loss S11 (dB)

-4
-6
-8
-10
Uniform feeding handset (free space)

-12

Scenario #1 (in presence of human hand-head)

-10

-15

Uniform feeding handset (free space)

Scenario #0 (in presence of human hand-head)
Scenario #1 (in presence of human hand-head)

Scenario #2 (in presence of human hand-head)

-14

-20

Scenario #3 (in presence of human hand-head)

Scenario #2 (in presence of human hand-head)

Scenario #3 (in presence of human hand-head)

Scenario #4 (in presence of human hand-head)

-16

Scenario #4 (in presence of human hand-head)

Scenario #0 (in presence of human hand-head)

-18
1.5

1.6

1.7

1.8

1.9
2
2.1
Frequency (GHz)

2.2

2.3

(c) 2-PIFA handset in the presence

2.4

2.5

-25
1.5

1.6

1.7

1.8

1.9
2
2.1
Frequency (GHz)

2.2

2.3

2.4

2.5

(d) 4-PIFA handset in the presence

of human head-hand model

of human head-hand model

Figure 10. The return loss results comparison of both handsets in

free space and in the presence of human head-hand model for different
scenarios.

Table 5. The required amplitude and phase excitations of each

element for the 2-element PIFA handset in the presence of the SAM
phantom human hand-head for different scenarios.
Element No.
1
2
Element No.
1
2

Scenario #1
amplitude
phase
1.43
169.29
2.73
139.69
Scenario #3
amplitude
phase
1.56
67.32
2.49
103.93

Scenario #2
amplitude
phase
1.02
102.02
2.27
47.383
Scenario #4
amplitude
phase
1.59
155.20
2.92
75.61

36

Montaser, Mahmoud, and Elmikati

Figure 10 shows the comparison of the return loss results of both

handsets in free space and presence of human hand-head model for
different scenarios. It is found that for all scenarios the handsets
resonance frequency still around 1.9 GHz but with different values of
S11 . The different values of S11 from scenario to another are related
to the change in the feeding weights for each scenario.
Nowadays, all newly proposed handset antennas consider the SAR
as an important design specification that they must meet. Therefore,

(a) Scenario # 0

(b) Scenario # 1

(c) Scenario # 2

(d) Scenario # 3

(e) Scenario # 4

Figure 11. SAR in full human head phantom for different scenarios.

Progress In Electromagnetics Research B, Vol. 37, 2012

37

Table 6. The required amplitude and phase excitations of each

element for the 4-element PIFA handset in the presence of the SAM
phantom human hand-head for different scenarios.
Element No.
1
2
3
4
Element No.
1
2
3
4

Scenario #1
amplitude
phase
2.96
47.34
2.17
17.29
2.38
63.87
1.03
20.13
Scenario #3
amplitude
phase
1.75
2.108
1.37
134.29
2.01
75.175
2.18
10.045

Scenario #2
amplitude
phase
1.97
59.85
2.97
27.05
1.53
132.54
1.14
16.76
Scenario #4
amplitude
phase
2.03
22.103
1.39
140.754
1.48
14.408
2.94
101.61

Table 7. Comparison between spatial-peak SAR over 10-g, gain and

S11 of two and four elements PIFA handset in the presence of human
hand-head for different scenarios.

Scenario # 0
2-PIFA Scenario # 1
Handset Scenario # 2
Scenarios Scenario # 3
Scenario # 4
Scenario # 0
4-PIFA Scenario # 1
Handset Scenario # 2
Scenarios Scenario # 3
Scenario # 4

SAR
Gain (S11 )f =1.9 GHz
(W/kg) (dBi)
(dB)
1.087
3.31
11.7
1.167
3.98
11.12
0.852
3.7
12.14
0.0371
4.31
10.95
0.0778
4.17
11.87
1.21
4.53
13.82
1.63
5.21
12.89
1.067
4.93
17.53
0.338
4.86
19.77
0.127
4.75
15.11

in this section we will study the impact of the handset on SAR.

Fig. 11 shows the SAR human head phantom for 4-PIFA handset
in different scenarios including the uniform feeding case (scenario #
0). It is noted that, when the PIFA array elements are fed with
different voltage values to achieve certain scenario, the spatial-peak

38

Montaser, Mahmoud, and Elmikati

(a) Horizontal cross section

(b) Vertical cross section

Figure 12. Horizontal and vertical cross section in human head

phantom for scenario # 1 (worst SAR).
SAR is affected significantly. As shown in Fig. 11, the position
of the spatial-peak SAR over 10-g on the human head is changed
from scenario to another in addition to the effective field size. It
is found that the highest spatial-peak SAR over 10-g is for scenario
#1 (SAR = 1.63 W/kg) but for other scenarios such as scenario #
4 the SAR is decreased to 0.127 W/kg compared to 1.21 W/kg for a
uniform feeding case (scenario # 0). Fig. 12 shows the horizontal and
vertical cross section human head phantom for scenario # 1 which
corresponding to the worst case to illustrate the penetration depth
of field inside the human head. It is found that, the highest SAR
value is obtained in the region of the skin near the handset. In
addition, some other issues should be noted as well, as suggested by
a comparison of the SARs induced in the realistic human-head model
for the homogeneous and the inhomogeneous cases [30]. It is found
that the constitutive parameters of a human head significantly affect
the result of the SAR induced in homogeneous or inhomogeneous head
models where the local maximum SAR induced in the homogeneous
human-head model is larger than that induced in the inhomogeneous
human head model. That means that the simulated SAR values in
this paper conservative estimate for the homogeneous human head
model, however, these values of SAR for the smart handsets in different
scenarios are under the limits set by IEEE C95.1: 2005 or ICNIRP
standards with acceptable gain values. Table 7 shows the spatial-peak
SAR over 10-g for two and four elements PIFA array on a handset in
addition to handset antenna gain and return loss for all scenarios in
the presence of human hand-head at 1.9 GHz.

Progress In Electromagnetics Research B, Vol. 37, 2012

39

5. CONCLUSION
In this paper, we investigated the capability of a 2- and 4-elements
PIFA array into a handheld device for communication systems in
the 1.9 GHz band for beamforming synthesis in a mutual coupling
environment. The PSO-NM algorithm is used to optimize the complex
excitations of the adaptive arrays elements for beamforming. Also, the
interaction of the antenna array, mounted on a mobile handset, with
a human hand-head phantom is investigated. It is found that, the
radiation pattern is affected by the presence of the human hand-head.
Therefore the beam patterns should be optimized in the presence of the
human hand-head. The spatial-peak SAR values of the smart handset
in the vicinity of the SAM phantom head and hand is also considered
for different scenarios. The numerical simulation results demonstrated
that the smart handset can work under SAR guidelines limitations.
ACKNOWLEDGMENT
We would like to acknowledge the Electronics Research Institute (ERI),
Microstrip Department for the support, encouragement, help and
cooperation during simulation process of this research.
REFERENCES
1. Lozano, A., F. R. Farrokhi, and R. A. Valenzuela, Lifting the
limits on high-speed wireless data access using antenna arrays,
IEEE Commun. Mag., Vol. 39, 156162, Sep. 2001.
2. Winters, J. H., Smart antennas for wireless systems, IEEE
Personal Commun. Mag., Vol. 5, 2327, Feb. 1998.
3. Murch, R. D. and K. B. Letaief, Antenna systems for broadband
wireless access, IEEE Commun. Mag., Vol. 40, 7683, Apr. 2002.
4. Ko, C. K. and R. D. Murch, Compact integrated diversity
antenna, IEEE Trans. Antenna Propagat., Vol. 49, 954960,
Jun. 2001.
5. Vaughan, R. G. and J. B. Andersen, Antenna diversity in mobile
communications, IEEE Trans. Veh. Technol., Vol. 36, No. 4, 149
172, Nov. 1987.
6. Rosen, A. and A. Vander Vorst, Special issue on medical
application and biological effects of RF/microwaves, IEEE Trans.
Microwave Theory Tech., Part II, Vol. 44, Oct. 1996.

40

Montaser, Mahmoud, and Elmikati

7. Rosen, A., A. Vander Vorst, and Y. Kotsuka, Special issue pm

medical applications and biological effects of RF/microwaves,
IEEE Trans. Microwave Theory Tech., Part II, Vol. 48, Nov. 2000.
8. IEEE C95.1:2005, Safety levels with respect to human exposure
to radio frequency electromagnetic fields, 3 kHz to 300 GHz,
IEEE standard, 2005.
9. ICNIRP Guidelines, Guidelines for limiting exposure to timevarying electric, magnetic, and electromagnetic fields (up to
300 GHz), Health Phys., Vol. 74, No. 4, 1998.
10. Dietrich, Jr., C. B., W. L. Stutzman, B. Kim, and K. Dietze,
Smart antennas in wireless communications: Base-station
diversity and handset beamforming, IEEE Antennas and
Propagation Magazine, Vol. 42, No. 5, 142151, Oct. 2000.
11. Denidni, T. A., D. McNeil, and G. Y. Delisle, Experimental
investigations of a new adaptive dual-antenna array for handset
applications, IEEE Trans. Veh. Technol., Vol. 52, No. 6, 1417
1423, Nov. 2003.
12. Yuan, Q., Q. Chen, and K. Sawaya, Performance of adaptive
array antenna with arbitrary geometry in the presence of mutual
coupling, IEEE Trans. Antennas Propag., Vol. 54, No. 7, 1991
1996, Jul. 2006.
13. Jensen, M. A. and Y. Rahmat-Samii, EM interaction of handset
antennas and a human in personal communications, Proceedings
of the IEEE, Vol. 83, No. 1, 11061113, Jan. 1995.
14. Vaughan, R. G. and L. S. Neil, Evaluation of antenna
configurations for reduced power absorption in the head, IEEE
Transactions on Vehicular Technology, Vol. 48, No. 5, 13711380,
Sep. 1999.
15. Li, L. W., M. S. Leong, P. S. Kooi, and T. S. Yeo,
Specific absorption rates in human head due to handset
antennas: A comparative study using FDTD method, Journal
of Electromagnetic Waves and Applications, Vol. 14, No. 7, 987
1001, 2000.
16. Kang, X. K., L. W. Li, M. S. Leong, and P. S. Kooi, A method
of moments study of SAR inside spheroidal human head and
current distribution along handset wire antennas, Journal of
Electromagnetic Waves and Applications, Vol. 15, No. 1, 6176,
2001.
17. Kouveliotis, N. K., P. J. Papakanellos, E. D. Nanou, N. I. Sakka,
V. S. G. Tsiafakis, and C. N. Capsalis, Correlation between SAR,
SWR and distance of a mobile terminal antenna in front of an
experimental validation, Journal of Electromagnetic Waves and

Progress In Electromagnetics Research B, Vol. 37, 2012

18.

19.

20.

21.
22.

23.

24.

25.

26.

27.

41

Applications, Vol. 17, No. 11, 15611581, 2003.

De Salles, A. A., C. R. Ferndadez, and M. Bonadiman, FDTD
simulations and measurements on planar antennas for mobile
phones, Microwave and Optoelectronics Conference. IMOC 2003.
Proceedings of the 2003 SBMO/IEEE MTT-S International,
Vol. 2, 10431048, Sep. 2003.
Stavros, K. and K. S. Nikita, Study of the coupling between
human head and cellular phone helical antennas, IEEE
Transactions on Electromagnetic Compatibility, Vol. 46, No. 1,
6270, Feb. 2004.
Kouveliotis, N. K., S. C. Panagiotou, P. K. Varlamos, and
C. N. Capsalis, Theoretical approach of the interaction between
a human model and a mobile handset helical antenna using
numerical methods, Progress In Electromagnetics Research,
Vol. 65, 309327, 2006.
Klemm, M. and G. Troester, EM energy absorption in the human
body tissues due to UWB antennas, Progress In Electromagnetics
Research, Vol. 62, 261280, 2006.
Ebrahimi-Ganjeh, M. A. and A. R. Attari, Interaction of dual
band helical and PIFA handset antennas with human head and
hand, Progress In Electromagnetics Research, Vol. 77, 225242,
2007.
Kuo, L.-C., Y.-C. Kan, and H.-R. Chuang, Analysis of a
900/1800-MHz dual-band gap loop antenna on a handset with
proximate head and hand model, Journal of Electromagnetic
Waves and Applications, Vol. 21, No. 1, 107122, 2007.
Khodabakhshi, H. and A. Cheldavi, Numerical analysis of
human head interaction with PIFA antennas in cellular mobile
communications, Progress In Electromagnetics Research B,
Vol. 22, 359377, 2010.
Priyadarshini, S. J., T. A. J. Mary, D. Sugumar, and
C. S. Ravichandran, Impact of SAR on human head modeling
in Elevators using IFA, Applied Electromagnetics (APACE),
2010 IEEE Asia-Pacific Conference, 15, Port Dickson, Nov. 9
11, 2010.
Kusuma, A. H., A.-F. Sheta, I. Elshafiey, Z. Siddiqui,
M. A. S. Alkanhal, S. Aldosari, S. A. Alshebeili, and
S. F. Mahmoud, A new low SAR antenna structure for wireless
handset applications, Progress In Electromagnetics Research,
Vol. 112, 2340, 2011.
Chim, K.-C., K. C. L. Chan, and R. D. Murch, Investigating
the impact of smart antennas on SAR, IEEE Trans. Antennas

42

28.

29.
30.

31.

32.

33.

34.
35.
36.
37.
38.

Montaser, Mahmoud, and Elmikati

Propag., Vol. 52, No. 5, 13701374, May 2004.

Akimasa, H., S. Mitsuzono, and T. Shiozawa, Feasibility study
of adaptive nulling on handset for 4G mobile communications,
IEEE Antennas and Wireless Propagation Letters, Vol. 3, 120
122, 2004.
Chen, Q., Y. Komukai, and K. Sawaya, SAR investigation of
array antennas for mobile handsets, IEICE Trans. Commun.,
Vol. E90-B, No. 6, 13541356, Jun. 2007.
Chen, H.-Y. and H. Yang, SAR affected by shapes and electrical
properities of the human head exposed to the cellular phone,
Microwave and Optical Technology Letters, Vol. 42, No. 1, 14,
Jul. 2004.
Yoshida, K., A. Hirata, Z. Kawasaki, and T. Shiozawa, Human
head modeling for handset antenna design at 5 GHz band,
Journal of Electromagnetic Waves and Applications, Vol. 19,
No. 3, 401411, 2005.
Mahmoud, K. R., M. El-Adawy, S. M. M. Ibrahem, R. Bansal,
and S. H. Zainud-Deen, Investigating the interaction between a
human head and a smart handset for 4G mobile communication
systems, Progress In Electromagnetics Research C, Vol. 2, 169
188, 2008.
Sanchez-Montero, R., S. Salcedo-Sanz, J. A. Portilla-Figueras,
and R. Langley, Hybrid pifa-patch antenna optimized by evolutionary programming, Progress In Electromagnetics Research,
Vol. 108, 221234, 2010.
Kennedy, J. and R. Eberhart, Particle swarm optimization,
IEEE International Conference on Neural Networks, Vol. 4, 1942
1948, Perth, Australia, 1995.
Nelder, J. A. and R. Mead, A simplex method for function
minimization, Computer Journal, Vol. 7, 308313, 1965.
Atef, E. and V. Demir, The Finite Difference Time Domain
Method for Electromagnetics: With MATLAB Simulations,
Har/Cdr Edition, SciTech Publishing, 2009.
CST Microwave Studio Suite 2011 Users Manual, www.cst.com.
Gabriel, S., R. W. Lau, and C. Gabriel, The dielectric properties
of biological tissues: II. Measurements in the frequency range
10 Hz to 20 GHz, Phys. Med. Biol., Vol. 41, 22512269, 1996.