See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/341680196

Advanced Electromagnetic Theory - Lecture Notes - Dr. Serkan AKSOY

Research · May 2020

CITATIONS READS

0 4,360

1 author:

Serkan Aksoy
Gebze Technical University
103 PUBLICATIONS 302 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of pulse induction (multi-frequency) metal detectors View project

Mode expansions in time domain View project

All content following this page was uploaded by Serkan Aksoy on 27 May 2020.

The user has requested enhancement of the downloaded file.

Advanced Electromagnetic Theory Dr. Serkan Aksoy

Advanced
Electromagnetic
Theory
Lecture Notes

Dr. Serkan Aksoy

These lecture notes are heavily based on the book of Advanced Engineering Electromagnetics (C. A. Balanis), 2012. For
future version or any proposals, please contact with Dr. Serkan Aksoy ([email protected]).
Advanced Electromagnetic Theory Dr. Serkan Aksoy

Content
1. THEOREM & PRINCIPLES ------------------------------------------4
1.1. Duality Theorem------------------------------------------------------------------------------------------ 4
1.2. Uniqueness Theorem ------------------------------------------------------------------------------------ 4
1.3. Image Theory ---------------------------------------------------------------------------------------------- 4
1.4. Reciprocity Theorem ------------------------------------------------------------------------------------ 4
1.5. Reaction Theorem ---------------------------------------------------------------------------------------- 5
1.6. Volume Equivalence Theorem ------------------------------------------------------------------------ 5
1.7. Surface Equivalence Theorem ------------------------------------------------------------------------ 6
1.8. Induction Equivalent ------------------------------------------------------------------------------------ 6
1.9. Physical Optics Equivalent ---------------------------------------------------------------------------- 6
1.10. Equivalency Evaluation --------------------------------------------------------------------------------- 6
2. SCATTERING ------------------------------------------------------------7
2.1. LINE SOURCE - CYLINDRICAL WAVE ---------------------------------------------------------- 7
2.1.1. Electrical Line Source ( Mode) ---------------------------------------------------------------------- 7
2.1.2. Magnetic Line Source -------------------------------------------------------------------------------------- 8
2.1.3. Electrical Line Source above Infinite PEC ------------------------------------------------------------- 8
2.2. PLANE WAVE SCATTERING (PWS) -------------------------------------------------------------- 8
2.2.1. PWS by Planar Structures --------------------------------------------------------------------------------- 8
2.2.2. PWS from a Strip ------------------------------------------------------------------------------------- 8
2.2.3. PWS from a Flat Plate -------------------------------------------------------------------------------- 9
2.3. CYLINDRIC WAVE TRANSFORM ----------------------------------------------------------------- 9
2.3.1. Plane Waves by Cylindrical Wave ---------------------------------------------------------------------- 9
2.3.2. Addition Theorem of Bessel Function ------------------------------------------------------------------ 9
2.3.3. Addition Theorem of Hankel Function ---------------------------------------------------------------- 9
2.4. CIRCULAR CYLINDER SCATTERING ----------------------------------------------------------10
2.4.1. Normal Incidence PWS: Polarization ---------------------------------------------------------- 10
2.4.2. Normal Incidence PWS: Polarization ---------------------------------------------------------- 10
2.4.3. Oblique Incidence PWS: Polarization --------------------------------------------------------- 10
2.4.4. Oblique Incidence PWS: Polarization ---------------------------------------------------------- 10
2.4.5. Electric Line Scattering: Polarization ---------------------------------------------------------- 10
2.4.6. Magnetic Line Scattering: Polarization -------------------------------------------------------- 11
2.5. CONDUCTING WEDGE SCATTERING ---------------------------------------------------------11
2.5.1. Electric Line Scattering: Polarization ---------------------------------------------------------- 11
2.5.2. Magnetic Line Scattering: Polarization -------------------------------------------------------- 11
2.5.3. Electric & Magnetic Line Scattering ------------------------------------------------------------------ 11
2.6. SPHERICAL WAVE ORTOGONOLITY ----------------------------------------------------------11
2.6.1. Vertical Dipole Spherical Wave Radiation ---------------------------------------------------------- 11
2.6.2. Orthogonality Relations --------------------------------------------------------------------------------- 12
2.6.3. Wave Transformations & Theorems ------------------------------------------------------------------ 12
2.7. CONDUCTING SPHERE SCATTERING --------------------------------------------------------12
3. DIFFRACTION --------------------------------------------------------- 13
3.1. GEOMETRICAL OPTICS -----------------------------------------------------------------------------13
3.1.1. Amplitude Relation --------------------------------------------------------------------------------------- 13
3.1.2. Phase & Polarization Relation -------------------------------------------------------------------------- 14
3.1.3. Reflection from Surfaces --------------------------------------------------------------------------------- 14
3.2. GEOMETRIC THEORY of DIFFRACTION------------------------------------------------------14
3.2.1. Amplitude, Phase & Polarization --------------------------------------------------------------------- 14
3.2.2. Straight Edge Diffraction & Normal ------------------------------------------------------------------ 14
3.2.3. Straight Edge Diffraction & Oblique ----------------------------------------------------------------- 15
3.2.4. Curves Edge Diffraction & Oblique ------------------------------------------------------------------ 15
3.2.5. Slope Diffraction ------------------------------------------------------------------------------------------ 15

2
Advanced Electromagnetic Theory Dr. Serkan Aksoy

3.2.6. Multiple Diffractions ------------------------------------------------------------------------------------- 15

3.2.7. Equivalent Diffraction Current ------------------------------------------------------------------------ 15
4. INTEGRAL EQUATIONS ------------------------------------------ 16
4.1. POINT MATCHING METHOD --------------------------------------------------------------------16
4.1.1. Basis Functions -------------------------------------------------------------------------------------------- 16
4.1.2. Subdomain Functions ------------------------------------------------------------------------------------ 16
4.2. METHOD of MOMENTS -----------------------------------------------------------------------------17
4.3. EFIE and MFIE -------------------------------------------------------------------------------------------17
4.3.1. Electric Field Integral Equation ------------------------------------------------------------------------ 17
4.3.2. Magnetic Field Integral Equation --------------------------------------------------------------------- 17
4.4. FAST MULTIPOLE METHOD ----------------------------------------------------------------------17
4.5. FINITE DIAMETER WIRES --------------------------------------------------------------------------17
4.5.1. Pocklington’s Integral Equation ----------------------------------------------------------------------- 18
4.5.2. Hallen’s Integral Equation ------------------------------------------------------------------------------ 18
4.5.3. Source Modeling ------------------------------------------------------------------------------------------ 18

3
Advanced Electromagnetic Theory Dr. Serkan Aksoy
1. THEOREM & PRINCIPLES ( ) ( )( )

1.1. Duality Theorem where defining and , it is clear

that the condition for a unique solution is .
A sample circuit with its reciprocal circuit is given for better
understanding the duality theorem. Using energy conservation, it is possible to prove
for lossy medium. For lossless medium, the same can be
proved as a limit case. In lossy medium, the following special
conditions are appeared in prove of uniqueness:

- If is defined on a surface (such as zero), fields

are unique valued. The reason for this is the tangential
components, but the normal components are ineffective.
- If is defined on a surface (such as zero), fields
are unique valued. The reason for this is the tangential
and are written from the circuit as
components, but the normal components are ineffective.
- If is defined on some certain section of surface, and
is defined on the other part of the same surface,
fields are unique valued.
where , , show the reciprocal values that 1.3. Image Theory
solution of both circuits are equal to each other. Similar to
It can be applied to PEC, flat and infinite extended surfaces.
this, and fields are also dual of each other. It means that
Since the boundary conditions on the tangential electric field
solution of fields can be found by using solution of fields.
components are satisfied over a closed surface ( to ), the
In other words, with proper arrangements, a solution for
solution is unique. Below the conductor, the equivalent system
( ) , ( ) can be used for a solution for
does not give the correct fields but are not concerned.
( ) , ( ) . These arrangements are given
below.

( ) , ( ) ( ) , ( )
( ) ( )
( ) ( )
( ) ( )
( ) ( )

⁄ ⁄

According to the duality theorem, the problems can be solved

by interchanging the quantities. For example, a duality of a
problem of is equal to
. The duality theorem gives chance to calculate fields
excited by equivalent magnetic sources using fields excited by
electric field sources (such as that fields of magnetic dipole
can be calculated by using electric dipole fields).
1.2. Uniqueness Theorem
It is necessary to know that the uniqueness of the solution of
electromagnetic problems are provided under which condition
(mean they have no other solutions). In a lossy medium with The direction of current can be evaluated from the polarization
and parameters, let and fields are excited by of the reflected fields or boundary condition on the surface.
sources of and . In both cases, Maxwell’s equations
1.4. Reciprocity Theorem
, Reciprocity theorem means that ideal voltage sources (zero
internal impedance) can be changed with ideal current source
, s (infinite internal impedance) without changing value read on
ampermeter in a linear electric circuit. For electromagnetic
where subtracting the second equation from the first theory, the reciprocity theorem means that location of
transmitter and receiver can be changed. Thus, difficulties of
( ) ( ) cylindrical or spherical wave propagation from a source can
be mitigated by transforming the problem plane wave
( ) ( ) ( ) propagation. To explain the reciprocity theorem, let us
consider that independent , and , sources having
where and , then, the second equation same frequencies excite , and , fields in a same
linear and isotropic medium. In this case, Maxwell’s equations
Advanced Electromagnetic Theory Dr. Serkan Aksoy
Reciprocity theorem must be satisfied for two different wave
, guide modes and so on.

, 1.5. Reaction Theorem

The reciprocity relation can not be evaluated in the sense of
where multiplying the first equation with and the last power because of no conjugate, but it may be preted a reaction
equation with (coupling) between sources and fields as

∭( )

is found. Subtracting the first equation from the second ∭( )

The first equation relates the coupling of the fields , to

the sources , , . Due to
where using a relation of ( ) ( ) reciprocity theorem
( ), it becomes
.
( )
The reaction theorem is used to calculate the mutual
In a similar manner, multiplying the second equation with impedance and admittance between aperture antennas and can
and the third equation with also be expressed by voltages and currents.
1.6. Volume Equivalence Theorem
The sources of and can create the fields and in
free space or and in a dielectric medium. In that case,
Maxwell's equations
where subtracting the second equation from the first

where again using a relation of ( ) ( ) After subtraction second one from the first one, one can obtain
( ), it becomes
( ) ( )
( )
( ) ( )
where subtracting this equation from ( ) equation
If and are defined as scattered
( ) (disturbance) fields, then rearranging

is found. This equation is known as a differential Lorentz , ( )- ( )

reciprocity principle. Applying volumetric integral and using
divergence thermo to this equation , ( )- ( )

∬( ) using the definition ( ) and

( ) , the equations

∭( )

is found. This equation is known as a integral Lorentz

reciprocity principle. Specially, if there are no sources in a
region as , then, the following where and are only exist in the
equations are satisfied. material and give chance to formulate Integral Equations for
scattering by dielectric objects. Moreover the surface
( )=0 equivalence theorem is more useful for scattering by PEC
surfaces and for analysis of antenna aperture radiations.

∬( )

5
Advanced Electromagnetic Theory Dr. Serkan Aksoy
1.7. Surface Equivalence Theorem and are due to the obstacle . In the step V,
means the shorted by conductor. As an example, the induction
Inside Outside On surface of
equivalent for scattering by conducting surface of infinite
I , by , , by , -
( )
extend is given at below:
II , by ,
( )
III , by ,

IV , PEC , by ,

V , PMC , by ,

This theorem is also called Huygen's Principle and the step III
is known as Love's Equivalence Principle. One of the key The induction equivalent for scattering by conducting surface
point in the theorem is that the medium is the same ( and of infinite extend.
) and homogeneous inside and outside the up to step IV.
The different applications of the theorem are also given at 1.9. Physical Optics Equivalent
below for magnetic source radiation near a PEC material and If one can consider the Induction Theorem for a PEC
for a waveguide aperture mounted on an infinite flat electric
ground plane. Inside Outside On surface of
I -

II

( ) ( )

Considering boundary condition,

( ) ( )
Magnetic source radiation near a PEC material.
where, the currents can be formed for equivalence problem as

( ) ( ( ) )
( ) ( ( ) )

This equivalency is known as Physical Equivalent and is

based on the EFIE and MFIE for unknown current densities. If
the conducting obstacle is an infinite, flat and PEC conductor,
the Physical Equivalent problem can be stated as

| ( )| |
with following figure.

A waveguide aperture on an infinite flat electric ground plane.

1.8. Induction Equivalent
The Induction (or Induction Equivalent) Theore is related to
Surface Equivalence Theorem and used for scattering problem
from aperture.

Inside Outside On surface of

I , by , , by , -
II -
( ) Physical equivalent problem for PEC materials.
III
( )
1.10. Equivalency Evaluation
IV
At a first glance, the solution of a scattering problem
V , PEC by the Induction Equivalent or Physical Optics Equivalent will
be the same in any sense. Whenever the Induction Equivalent
where are transmitted fields in is different medium gives a known current placed on the surface of
having and after step I. Whereas in outside and . the obstacle can not be used for scattering calculation because
the medium within and outside the obstacle is not the same.
6
 Advanced Electromagnetic Theory Dr. Serkan Aksoy
The new boundary value problem (as difficult as the original 2. SCATTERING
one) must be solved to know the currents on the surface of the
obstacles. Nevertheless the Physical Equivalent currents
can be used for calculation of the scattered field
because the same medium is present within and outside of the  Modal Solution: Needs orthogonal systems ( ), poor
surface. The fundamental difficulty in this way is not to know convergent series.
the current density on the surface of the obstacles and  IE, MoM: Arbitrary shapes, not to many wavelengths.
generally as difficult as to find the original solution of the  GO (GTD), PO (PTD): Many wavelengths.
problem with requiring the knowledge of the total fields in
the original problem. 2.1. LINE SOURCE - CYLINDRICAL WAVE
If one assumes that the PEC obstacle is enough large
electrically (locally flat), the image theory can be used for 2.1.1. Electrical Line Source ( Mode)
Induction Equivalent and the known current can
be used to calculate scattered fields. In many cases, a closed
form solution can not be obtained easily due to the inability
for integration along closed surface of the obstacle. In this
case, the integration can be performed only part of the surface
(visible for transmitter) that the current density is intense for
major contribution of the scattered field. With the electrically
large assumption, the Physical Equivalent currents can be used
with the form of known as Physical Optics
Approximation. For backscattering calculations, the Induction
Equivalent and the Physical Equivalent give the same results ⃗⃗⃗ : No variation (infinite length of wire)
and these approaches must be extended for conductors (not
PEC) and dielectrics.

The field solution is the form of

( ) ( )
[ ( ) ( )]
, ( ) ( )-
, -

 Outward direction,
 Lowest order mode,
 No variation,

In that case, the solution

( )
( ) ( )

Then
,
( )
( )
( )
( )

The calculation of in the field equation from Amper law

( )
∫ ⏟ ∫ ( )

( )
where ( ) , then the fields

( )
( ) √
√

7
Advanced Electromagnetic Theory Dr. Serkan Aksoy
2.2. PLANE WAVE SCATTERING (PWS)
( )
( ) √ 2.2.1. PWS by Planar Structures
√ Radar Cross Section (RCS): Area intercepting the amount of
power that, when scattered isotropically, produced at the
where ⁄ is impedance and receiver a density that is equal to the density scattered by the
actual target. ( or ) is known as scattering width
( ) ( )
( ) √ , ( ) √ . (RCS per unit length), ( or ) is RCS. RCS pattern
is a function of space coordinates.
2.1.2. Magnetic Line Source
Physically not realizable, but used for modeling of radiating ⃗ |⃗ | |⃗ |
apertures. Using Duality Principle * + [ ] [ ]
⃗ |⃗ | |⃗ |
( )
( ) √ ⃗
√ * +

( )
|⃗ | |⃗ |
( ) √ [ ] [ ]
√ |⃗ | |⃗ |

Relation between and ( is the target length)

2.1.3. Electrical Line Source above Infinite PEC

⃗ ⃗ |
√

and definition of RCS can be approximated when the

target is placed in the far-field of the source (Specular
Original problem Near-field equivalent Far-field equivalent reflections satisfies Snell’s law).

- Monostatic (Backscattered): Tx and Rx are in same place.

( ) ( ) - Bistatic: Tx and Rx are in the different place.
{ [ ( ) ( )] }

2.2.2. PWS from a Strip

In the asymptotic case:

{ ( )√ }
√ √

⃗ ( )
At large distance: for amplitude and
, for phase, the far-field approximation
⃗ ( ) ( )

( )
⃗ ( )

√ ( )
{ }
√ ⃗ ( ) ( )

for finite strip, for infinite strip,

and for PEC material. Thus reflected fields are

⃗ ( )

8
Advanced Electromagnetic Theory Dr. Serkan Aksoy
⃗ ( ) ( )

Using the far-field transformation ( ), the field

Using Physical Optics components , , , , , are calculated.

⃗ ⃗ | ⃗ (⃗ ⃗ )| ⃗ ⃗ |
2.3. CYLINDRIC WAVE TRANSFORM
2.3.1. Plane Waves by Cylindrical Wave
( )
∑ ( )

where the infinite sum shows the cylindrical wave function

To calculate far-zone scattered field and it can be proved that .
2.3.2. Addition Theorem of Bessel Function
⃗ , ⃗ ( ) ( )
 ( ) [ ( ) ( )]

∫ ( )
( )
∑ ( ) ( )
Then ( | |)
( )
∑ ( ) ( )
√| | | |
{ }
∫[ ∫ ]
√| | | | ( ) ( )
 ( ) [ ( ) ( )]

It is known that
( )
∑ ( ) ( )
√
∫
( )
( ) ( | |)
√ ∑ ( ) ( ) ( )

{ }
Then
2.3.3. Addition Theorem of Hankel Function
( ) ( )
( )
It is based on the equivalence of and in the far field.
∫ ( | |)
( )
( | |)
For far-zone observation | | ( ) ∑ ( ) ( )
( ) ( )

( ) ( )
( | |) √ ∑ ( ) ( )
( ) ( )

{ }

√ ( )
( | |)
√
( ) ( )
∑ ( ) ( )
Then and by using it, ⃗ , ⃗ and can be calculated.
2.2.3. PWS from a Flat Plate ∑ ( ) ( )
( ) ( )

⃗ ( ) ( ) { }

⃗ ( )

Induced current on the surface

⃗ ⃗ |

9
Advanced Electromagnetic Theory Dr. Serkan Aksoy
2.4. CIRCULAR CYLINDER SCATTERING Using the transformation

2.4.1. Normal Incidence PWS: Polarization

∑ ( )

The outward direction

( )
∑ ( )

Applying the boundary condition

| | |
⃗ ∑ ( )
,
( )
⃗ ⃗ ⃗ ( )⁄ ( )

In outward direction Far zone scattered field:

( ) ( )
⃗ ∑ ( ) ( ) √

Using the boundary condition 2.4.4. Oblique Incidence PWS: Polarization

The solution is similar to the section 11.5.3 for ⃗ .
( )
⃗ | 2.4.5. Electric Line Scattering: Polarization
( )
( )

The induced current with small radius approximation ( )

meaning the first term is dominant.

⃗ ⃗ | ( )

Far zone scattered field:

( )
√ ∑ ( )
√ ( ) ( )
( | |)

2.4.2. Normal Incidence PWS: Polarization

Using the addition theorem
The solution is similar to the section 11.5.1 for ⃗ .
2.4.3. Oblique Incidence PWS: Polarization ∑ ( ) ( )
( ) ( )

( ) ( )
∑ ( ) ( )
{ }

where ( ) is chosen for because the field should be

( )
finite everywhere including and ( ) is chosen for
because the travelling nature of the wave. Then

( )
∑ ( )

⃗ ( )

10
Advanced Electromagnetic Theory Dr. Serkan Aksoy
Applying the boundary condition as 2.5.2. Magnetic Line Scattering: Polarization
( ) , ( )-
| | | ∑ ( ) ( )
, ( )-
( ) , ( )-
may be found. Then the field components , , ∑ ( ) ( )
{ , ( )- }
and , , are found. The current
density ⃗ ⃗ | can be found.
Using Maxwell’s equations ⃗ is calculated and boundary
2.4.6. Magnetic Line Scattering: Polarization condition is applied at and with .
Then with allowable ⁄ ( ) , since
The solution is similar to the electric line scattering.
source is magnetic field ( ⁄ ( )) where
or . Far zone field ( )
2.5. CONDUCTING WEDGE SCATTERING or PWS ( ) can be found by using the asymptotic
Hankel function.
2.5.1. Electric Line Scattering: Polarization
2.5.3. Electric & Magnetic Line Scattering
Using a new coordinate system, it is possible to write ⃗
( , soft) and ⃗ ( , hard) as

⃗ ( )

⃗ ( )

where ( ) is Green function.

2.6. SPHERICAL WAVE ORTOGONOLITY

( ) ( )
∑ ( ) ( )
2.6.1. Vertical Dipole Spherical Wave Radiation
( ) ( )
∑ ( ) ( )
{ }

∑ ( ) ( ) , ( )- , ( )-

( )
∑ ( ) ( ) , ( )- , ( )-
{ }

when , two must be identical, then and

( )
( ) ( ), ( ) ( ). When and
( )
, ⃗ will vanish. It means that Using spherical Hankel function ( ) ⁄
⁄ ( ) . Then ⃗ can be calculated
by Maxwell’s equation. Induced current density ( )
( )

⃗ ⃗ | ∑ , ( )- , ( )- For a linear magnetic current element using duality

( )
Since the Fourier series for a current located at and ( )
, then
If the source is removed from origin as
∑ , ( )- , ( )-
( )

where ⁄ ( ) and depends on source type.

Far zone field ( ) or PWS ( ) can
be found by using the asymptotic form of Hankel function.

11
Advanced Electromagnetic Theory Dr. Serkan Aksoy
( )
( )
( )
( ) ( ) ( ) ( )
∑
( , ( )-)
( )
( ) ( ) ( ) ( )
∑
{ ( , ( )-) }

2.7. CONDUCTING SPHERE SCATTERING

( )
( | |)

( )
( | |)

( ) | |
where ( | |) ⁄ | |.

2.6.2. Orthogonality Relations

In spherical coordinates, Legendre ( ) functions and ⃗
Associated Legendre (Zonal Harmonics) functions ( )
form a complete orthogonal set for . Therefore, Using the transformation for ⃗ ⃗ ⃗ ⃗
their series form Legendre polynoms

( ) ( ) ( ) ( )

are used to represent arbitrary functions. The polynoms

(Tesseral Harmonics) form a complete set on the sphere
surface. These form also Fourier-Legendre series

( ) ∑ ( ) where is also can be written as

where ( ⁄ )∫ ( ) ( ) and the

following condition holds
Using the spherical transformation

∫ ( ) ( ) { }
⁄( ) ∑ ( ) ( ) , ( )-
2.6.3. Wave Transformations & Theorems

⃗ ∑ ( ) ( ) ∑ ( ) ( ) ( )

Using the orthogonality relation ( ).

∑ ( ) ( ) ( )
 Addition theorem of spherical wave functions
( )
( )
( ) ( ) ( )( ) The solution can be constructed as
∑
( , ( )-)
( ) :
( ) ( ) ( )
∑
{ ( , ( )-) } :

Using the , it can be proven that

12
Advanced Electromagnetic Theory Dr. Serkan Aksoy

∑ ̂ ( ) ( )
3. DIFFRACTION
GTD: Keller, extended by Pathak (diffraction)
PTD: Extended by Ufimtsev (nonuniform fringe edge current)

∑ ̂ ( ) ( ) Diffraction is a local phenomenon depends on

- the geometry of the object (edge, vertex, curve)

( )⁄ ( ) and ( ) ̂ ( )⁄ . - the amplitude, phase and polarization of the field
where
Due to the field component are uniform plane waves, and
can be constructed for scattered field as 3.1. GEOMETRICAL OPTICS
Phase of rays is assumed that equals to product of optical ray
length (with ) from reference. Amplitude of rays is assumed
∑ ̂( ) ( ) ( ) that vary in a narrow tube with the principle of energy
conservation. The phases at the caustics should be rearranged.
Specular Reflection is only allowed (Snell law). The Fermat
principle equations is used for GO as
∑ ̂( ) ( ) ( )

∫ ( )
To determine and , the boundary conditions are applied

( ) where is variational differential, ( ) ( )⁄ is

refraction index. If ( ) is constant, paths are straight lines.
( ) 3.1.1. Amplitude Relation

Then, it can be proved that

̂ ( ) ̂ ( )
( ) ( )
( ) ̂ ( )

Then , , and can be calculated. Using the spherical

Hankel function and Hankel function relations

Due to the energy conservation .

( )
̂ ( )( ) √ ( )
( ) ( )
( ) ( )
In the far field region ( )
where because ( ⁄ )| ⃗ | , then
|⃗ |
[ ] ,| | | | -
|⃗ | |⃗ |
√
|⃗⃗⃗⃗ |
The RCS of a sphere is given below.
The areas can be written by the radii of curvature as

|⃗|
 Spherical √
|⃗⃗⃗⃗ |
|⃗|
 Cylindiral √ √ √
|⃗⃗⃗⃗ |

|⃗|
 Plane √
|⃗⃗⃗⃗ |

Phase and polarization information are absent yet and the

caustic problem is present.

13
Advanced Electromagnetic Theory Dr. Serkan Aksoy
3.1.2. Phase & Polarization Relation 3.2.1. Amplitude, Phase & Polarization
Luneberg Kline series is given as
( )
√
( )
( )
( ) ∑
( ) where is eikonal surface, ( ) is diffracted field factor.
Substituting to wave equations
Substituting this equation to wave equation, Eikonal,
Transport and Conditional Equation can be found as
( ) ( )√
( ) ( )
Eikonal equation | |
( )
Transport equation , - where ( ) [ ⁄√ ] ( ) is diffracted field at
reference. The diffracted field has to satisfy the following
Conditional equation relation

( )√ ( )
where ⁄ and is wave front surfaces (may be
plane, cylindrical, spherical), then using the transport equation
and reference field values Thus, the diffracted field

( ) ⏟( ) ⏟ ⏟( ) ⏟
( )
( ) ⏟( ) √ ⏟
⏟
where the spreading factor for a curved surface ( )
The solution from Luneburg-Kline series predicts phase and √ ⁄ ( ). And is a function of wavefront curvature
polarization information. More accurate results may be angles of incident and diffraction radius of edge curvature.
obtained by higher order terms. But diffraction mechanism
can not be treated. In the caustics, the field is singular and  If the edge is straight
another approach is used.
( ) Incident Wave
3.1.3. Reflection from Surfaces For plane and conical
Snell law is applied. Near the reflection point √ wave incidences
For cylindrical wave
- Reflecting surface is approximated to plane, √ √ incidences
- Incident wave front is assumed to be planar.
For spherical wave
√
Then, the reflected field is given by √ incidences
( )

( ) ⏟( ) ⏟ √ ⏟ The important key is to find diffraction coefficient.

( ) ( )
⏟
3.2.2. Straight Edge Diffraction & Normal
The radiation mechanism nearby the edges needs to separate
where and can be approximated by and with focal space surrounding wedge into three different regions with
point and . If incident ray is spherical, spherical wave is reflection shadow boundary and incident shadow boundary.
treated. The principal radii of curvature (Principal Plane) are To remove discontinuity on boundaries (modify the fields for
defined. It is assumed that the reflecting surface is well- physically realizable field), diffraction has to be included.
behaved (smooth and continuous). Diffraction coefficient can be extracted by steps as
3.2. GEOMETRIC THEORY of DIFFRACTION
- Find Green’s function in series form of far field region
Edge diffraction can be evaluated by a diffraction coefficient - Convert series form Green’s function to integral form
needs canonical problem. Applying Fermat principle, Law of - Evaluation of the integral by Steepest Descent method
Diffraction is obtained. GO fails
In the case of electrical (or magnetic) line source
- Incident rays are tangent to curved surface
- Present an edge, vertex, corner
- Present caustics

New semi-heuristic approach must be proposed. where the asymptotic solution of ( ) ( )

14
Advanced Electromagnetic Theory Dr. Serkan Aksoy
3.2.6. Multiple Diffractions
( ) ∑ * ( ( )) If the structures have multiple edges, coupling between edges
( ( ))+
should be considered. Especially this is important for close
edges. The multiple interactions are known as Multiple
where depends on Dirichlet or Nuemann boundary Diffraction (higher order diffraction). To account especially
condition. The series converges rapidly when small , but third and even higher order diffraction, a procedure adopted
slowly converges when large . To overcome this problem, for accounting all diffraction is used and known as Self
the series are transformed to integral form Consisting Method. It is based on the multiple reflection and
transmission coefficients and series representations of
Multiple Diffractions.
( ) ∫ ( )* ( ) ( )+

( )
3.2.7. Equivalent Diffraction Current
⏞ In contrast to straight edges diffraction, curved edges
( ) ∫ ( ) ( ) diffraction creates caustics. To correct field near to caustics,
Equivalent Current (an equivalent two dimensional electric
∫ ( ) ( ) (soft polarization) or magnetic (hard polarization) line current
⏟
( ) technique can be used. In that case, by equated the currents
and diffracted fields, the currents
( ) ∫ ∫ ∫
⏟ ⏟ √
( )

( ) ∫ ∫ ∫ √
⏟ ⏟ ( )

If the wedge has finite length , the currents have also finite.
After contour integration, Steepest Descent and Modified The field created by the currents can be calculated using
Steepest Descent methods, the solution can be arranged as the standard way. Moreover in the case oblique angle, the
( ) ( ) ( ). After evaluating the Equivalent Currents should be modified. In the case of curved
solution, diffraction coefficient can be extracted for incident edge diffraction, Equivalent Currents are modeled by quarter-
and reflected waves in the sense of polarizations. Thus the wavelength monopole mounted on a circular ground plane in
incident and reflection shadow boundaries can be clarified. which rim of it is modeled as a ring radiator.

3.2.3. Straight Edge Diffraction & Oblique

In reality, not only for principal pattern, but also for all pattern
(directions), the diffraction coefficient has to be calculated.
Although in the case of normal incidence, the diffraction
coefficient becomes scalar, it becomes Dyadic form for
oblique incidence.

3.2.4. Curves Edge Diffraction & Oblique

Because the diffraction is a local phenomenon, curved edges
(Convex, Concave) can be approximated as a wedge and
application of the wedge diffraction theory meaning the scalar
diffraction coefficients.

3.2.5. Slope Diffraction

In classical evaluation, the diffraction coefficient is zero when
the field is zero at the point of diffraction. But in fact if the
normal derivative (slope) of the incident field is also causes a
higher order diffraction known as Slope Diffraction creating
the currents on the wedge surface. The slope diffraction
coefficients can be calculated separately for different
polarizations.

15
Advanced Electromagnetic Theory Dr. Serkan Aksoy
4. INTEGRAL EQUATIONS then , - , -, - in which , - is the unknown. The
system linear equation each with unknowns by applying
The solution of realistic finite sized objects can be found by the boundary condition at discrete points technique is
Integral Equation, IE (numerically Method of Moments). The known as Point Matching (or Collocation) method.
total current density can be Physical Optics (with fringe wave)
currents. Using the current with IE, the fields can be 4.1.1. Basis Functions
calculated. But because the current is not known, current is Basis functions should be chosen to have ability for accurately
approximated to series solution with a basis function and an representation of anticipated unknown function while
unknown coefficient for a boundary value of the field. The set minimizing the computational efforts. Basis functions can be
of linear equation is obtained as a matrix equation. classified as a Subdomain (non-zero over a part of the
domain) and an Entire Domain (exist entire domain) basis
4.1. POINT MATCHING METHOD
functions. Do not chose basis functions with smoother
Electrical line source above a segmented strip can produce the properties than the unknown function.
scattered electric field because of line source can be written as
4.1.2. Subdomain Functions
( ) These are most common and can be used without prior
( ) ( )
knowledge of the unknown function. The subdomain approach
based on the subdivision of the structure into no
where because of segmentation, the current is ( ) overlapping segments. Some subdomain basis functions
( ) , then
- Pulse function,
- Triangle function,
( )
∑ ( ) ( ) - Piecewise sinusoid,
- Truncated cosine.

If every segment is so small ( ) , then As an example, triangle function is overlap adjacent function
which is smoother than pulses but the cost of increased
⁄ computational complexity.
∫ ( ) ( )
( | |) If ( ) is subdomain type (exist only over one segment)
means that ( ) only for , otherwise
⁄
is zero, then the integral is as
Using the boundary condition for electric field |
( )
⁄ ∫ ( | |)
( ) ( )
( ) ∫ ( ) ( | |)
⁄ The closed form evaluation of this integral is not possible
because of the self term on surface (for
This equation is known as Electric Field Integral Equations application of boundary condition on the surface) is zero
(EFIE). If one expands the current density ( ) as causing the singularity of Hankel function. To overcome this,
the observation point is chosen away from the surface. But
will still sufficiently small that the computation of
( ) ∑ ( ) Hankel function may not be very accurate. The approximation
of Hankel function for small argument is used and
approximate closed form of the integral can be evaluated for
then the EFIE
diagonal and nondiagonal terms. Specifically the average
⁄ value of arguments is considered for also curved space.
( ) ( )
( ) ∑ ∫ ( ) ( |
4.1.2.1. Entire Domain Functions |)
⁄
A common one is sinusoidal function (similar to Fourier
series) useful for modeling sinusoidal distribution as the wire
which takes the general form as ∑ ( ) where current. The main advantages of it are assumed a priori to
is the known excitation function, is the linear integral follow a known pattern. Such entire domain functions may
operator. If one considers the problem in observation render the unknowns with a fewer terms. Because using a
points, EFIE is finite number of functions, the modeling of arbitrarily or
complicated functions have difficulty by entire functions
⁄
which can be generated from polynomials.
( )
[⏟ ( )] ∑ ⏟ [ ∫ ( ) ( )( | |) ]
⏟ ⁄

16
Advanced Electromagnetic Theory Dr. Serkan Aksoy
4.2. METHOD of MOMENTS ⃗ | [ ∬ ( ) ∬ ( ) ]
Boundary conditions are satisfied only at discrete points in
Point Matching Method and between these points, boundary
conditions are not satisfied and deviation as a Because unknown is expressed by incident electric field, it
. To minimize residual that its overall average over is referred as EFIE which is actually integro-differential
entire structures approaches zero, Method of Weighted equation. It can be used for closed or open surfaces. After
Residuals (MoM) is utilized and forces boundary conditions to calculating , the scattered field can be calculated. For open
be satisfied in an average sense. To do this, Weighting surfaces, a boundary condition should be supplemented to
(Testing) functions * + form inner product as
yield a unique solution for normal component of to vanish
on . The above given EFIE is for a general surface of 3D
problems but can be reduced to 2D case.
〈 〉 ∑ 〈 ( )〉
4.3.2. Magnetic Field Integral Equation
MFIE enforces boundary condition on tangential magnetic
This can be formulated as the matrix equation ( )
field and similar to EFIE but based on incident magnetic field
, - , -, - , - , - , -
| ⃗ (⃗ ⃗ )|
where must be linearly independent (matrix equation be
also linearly independent) and should be chosen to minimize
⃗ ( )
computational load. If weighting and basis function are the
same, technique is known as Galerkin's Method (Others are where
Point Collocation, Collocation, Least Squares). Specifically
choosing the set of Dirac weighting function (Point
Collocation) will reduce the computational requirements, but ∬ ( )
is forced the boundary condition at discrete points, hence
name is Point Matching. The positioning of points (equally
spaced one yield good results) depends on basis function If the observation is restricted on the surface
choosing in some configuration (such as match point does not
coincide with peak of triangle basis functions due to it's not ⃗ |
| {⃗ ∬ ( ) , ( )- } ⃗
differentiability) may cause errors. Point Matching is popular
testing technique due to its acceptable accuracy. For a strip
problem, a convenient inner product would be
Because unknown is expressed by incident magnetic field,
⁄ it is referred as MFIE which is valid for only closed surfaces.
〈 〉 ∫ ( ) ( ) After calculating , the scattered field can be calculated. The
above given MFIE is for a general surface of 3D problems but
⁄
can be reduced to 2D case.
Applying the inner product to previously given EFIE 4.4. FAST MULTIPOLE METHOD
, - , -, - The MoM treats each of basis function resulting in an
( ) scaling of memory requirements for storing the
If the weighting (testing) functions are Dirac functions as impedance matrix and in an ( ) CPU time to solve linear
( ) ( ), then and . The set of equations (number of process) if the solution of the
matrix is performed by Gaussian elimination method. By
MoM is introduced to minimize average deviation from actual
values, of the boundary conditions over the entire structure. iterative method, number of process is ( ) where is
number of iteration. Fast Multipole Method (FMM) is a kind
4.3. EFIE and MFIE of iterative method and uses memory and number of matrix-
vector multiplication as ( ⁄ ). EFIE solution by FMM is
4.3.1. Electric Field Integral Equation not preferable because preconditioners have to be more
EFIE enforces the boundary condition on total tangential elements and use more memory. MFIE solution by FMM is
electric field as ⃗ | ⃗ | ⃗ | also not preferable because internal resonance problems.
Therefore ( ) ( )
solution by FMM is used and shown less memory and less
⃗ ( ) progress time. Moreover Multi Level FMM results in
( ) scaling in memory and , ( )- in CPU time.
where
4.5. FINITE DIAMETER WIRES
∬ ( ) Three dimensional IE can be arranged to find the current
distribution on a conducting wire. The obtained forms are
If the observation is restricted on the surface known as Pocklington’s Integrodifferential Equation and
Hallen’s Integral Equation. For very thin wires, the current

17
 Advanced Electromagnetic Theory Dr. Serkan Aksoy
distribution can be assumed as sinusoidal but for finite This is Pocklington’s Integral Equation. It can be used to
diameter wires ( ), the sinusoidal distribution determine filamentary line source current of the wire. If the
assumption is not accurate. wire is very thin ( ), then
4.5.1. Pocklington’s Integral Equation
( ) ( )

more convenient form of the Pocklington’s Integral Equation

can be obtained. Point Matching method can be used to solve
Pocklington’s Integral Equation. Matching points are to be at
the interior of the wire, but by reciprocity, matching points
can also be chosen on the wires.

The boundary condition on the wire surface 4.5.2. Hallen’s Integral Equation
When and , neglecting end effects of the wire, the
| | | | | only current flows along wire means that and
satisfies the equation
At any observation point, the scattered field can be calculated
by the vector potential. But at the wire surface, only | is
enough to calculate as ( )

with boundary condition |

( )

where
⁄

∫ ∫ ∫ and its solution ( ) √ , ( ) ( | |)-

with the definition of vector potential for a line
⁄

⁄
If wire is very thin, is not a function of azimuthal angle as
∫ ( ) √ , ( ) ( | |)-
( ) ⁄
( )
This is Hallen's Integral Equation for a PEC wire. and
where ( ) is equivalent line source current at . Than can be found from boundary conditions.

⁄ 4.5.3. Source Modeling

∫ ∫ ( ) To fed finite diameter wire, Delta-Gap and Equivalent
Magnetic Ring Current (Magnetic Frill) can be used as
⁄

Because of symmetry of the scatterer, fields are not function  Delta-Gap: It is simplest and widely used, but least accurate
for impedances. Excitation voltage at feed terminals is
of chosen as and observation at the surface ( )
constant and zero elsewhere. Therefore electric field is also
⁄ constant over the gap, and zero elsewhere. Then Equivalent
Magnetic Current Density ( ⁄ ⁄ )
∫ ( ) ∫
⁄ ⏟
( )
⃗⃗ ⃗ ⃗

Then,  Magnetic Frill: It is based on near as well as far zone fields

⁄
of coaxial apertures. Circumferentially directed magnetic
( ) ∫ ( ) ( ) current density is replaced over an annular aperture. Then
⁄ Magnetic Source basing on transmission lines is
By arranging with boundary condition

⁄
⃗⃗ ⃗ ⃗
( ⁄ )
∫ ( ) *( ) ( )+
⁄

18

View publication stats