Professional Documents
Culture Documents
SR
E (0) , H (0) VR \V2
Ed, H d
E (1) , H (1)
V−
E (2) , H (2) S2
V+ r V2
+
...
...
| {z } | {z }
E+, H + E−, H −
Fig. 4. The surfaces SR , S2 , enclose the volumes VR and V2 , respectively.
Fig. 3. Decomposition of fields E , H generated from an antenna based on The volume VR \V2 is between the two surfaces SR and S2 .
source origin into the fields E + , H + and E − , H − . Scattering components,
E (n) , E (n) , are denoted with their scattering order n, i.e. the number of
reflection they have been under. The direct components, E d = E (0) , H d = First, we will focus on the second of these four integrals, i.e.
H (0) , is included in E + , H + .
(12) in the above expression. The integral (12) depends on the
fields E2+ , H2+ from Antenna 2, and the fields E1− , H1− from
II. T HE R EACTION F OR S CATTERED F IELDS Antenna 1 that are scattered on objects enclosed by S2 .
A. Field Decomposition
B. The Radiation Condition
We consider an antenna generating the fields E, H that are
scattered on some objects, as in Fig. 3. We use a fictitious We let a sphere SR enclose a volume VR that include both
surface to separate the transmitting antenna (if necessary, also antennas and the surface S2 , see Fig. 4, and calculate the
including parts of the platform) and scatterers in two disjoint reaction (1) over the surface SR . When the radius r of the
regions, V + and V − , as in Fig. 3. sphere goes to infinity r → ∞, the radiated fields decay as
The fields from the antenna will induce a current J on r−1 and the cross products as r−2 . Since the area of integration
the platform and all scattering objects. We note that, from goes as r2 , the integral will not obviously converge. We have
the method of moment approach, given the induced current J to convince ourselves that the reaction over the large sphere
and material parameters ε(r), µ(r), the fields E, H can be SR with the decomposed fields is zero.
uniquely determined, see e.g. [9]. Clearly, E, H are linearly We assume that the medium far away from the antennas is
dependent of J . The current J is decomposed into two parts, homogeneous and isotropic, and that the sources are localized.
With these assumptions, the EM fields will propagate outwards
J = J + + J −, (4) from the antennas and, asymptotically (i.e. infinitely far away),
based on the origin of the currents, so that form a locally plane wave.
For large but finite distances from the source, the wave
J+ = J , in V + , (5) will not be perfectly plane. We consider the Silver-Müller
− + radiation condition, which gives information about the asymp-
J = 0, in V , (6)
totic behavior of the fields far from the sources. In [10]
J+ = 0, in V − , (7)
− −
the Silver-Müller radiation condition is given as (with a
J = J , in V . (8) omitted/normalized impedance Z = 1)
Based on the superposition principle, we consider the fields lim r (H × n̂ − E) = 0. (15)
E + , H + to be generated by the sources J + and the fields r→∞
E − , H − by the sources J − , so that Using the little-o notation1 , the radiation condition (15) for
a large, but finite, r and a general impedance Z can be
E = E+ + E−, (9)
formulated as a plane wave including a correction term as
H = H + + H −. (10)
E = Z(H × n̂) + o(r−1 ), (16)
Consequently, the fields E1− , H1−
are source free in V and +
the fields E2+ , H2+ are source free in V − . in which it is understood that each field component goes as
We apply the decomposition (9)–(10) to E1 , H1 and o(r)−1 . Using (16), we can write the integrand in the reaction
E2 , H2 in (1). We let the surface that separates V + and integral (1) as,
V − coincide with the integration surface S2 in Fig. 1–2. The E2+ × H1− − E1− ×H2+ = (17)
reaction (1) separates into four terms,
I = Z(H2+ × n̂) + o(r −1
)) × H1−
E2+ × H1+ − E1+ × H2+ · n̂2 dS (11) −(Z(H1− × n̂) + o(r −1
)) ×H2+ =
h2, 1i =
I S2 H2+ (H1− · n̂) − H1− (H2+ · n̂) + (H1− − H2+ )o(r−1 )
= Z
E2+ × H1− − E1− × H2+ · n̂2 dS (12)
+
1 For
non-zero g(r), the little-o notation f (r) = o g(r) implies that
IS2
E2− × H1+ − E1+ × H2− · n̂2 dS (13) f (r)
+ lim = 0.
r→∞ g(r)
IS2
Additionally, the little-o notation is related to the Big-O notation. If a function
E2− × H1− − E1− × H2− · n̂2 dS. (14)
+ f (r) = o(r), it implies that f (r) = O(r) [11].
S2
MALMSTRÖM: ON THE REACTION THEOREM FOR SCATTERED FIELDS 3
Far from the source, the dominant parts of H1− and H2+ outside the integration surface S2 . With this choice, there
are orthogonal to the direction of propagation n̂. We know that will be no scattering from outside S2 , and, hence, E2+ =
H1− , H2+ , and (H1− − H2+ ) all decay as r−1 for large r. The E2d , H2+ = H2d , E2− = 0, H2− = 0 (using the notation
radial components, H1− · n̂ and H2+ · n̂, decay faster, so that introduced in Fig. 3). The fields E2d , H2d , E1−, H1−, have
H1− · n̂ = o(r−1 ) and H2+ · n̂ = o(r−1 ). Hence, all terms in their sources within S2 and are source free outside S2 . With
(17) are o(r−2 ), and so is the whole integrand, i.e., this restriction, the same derivations can be repeated for the
generalized reaction, resulting in the following expression,
E2+ × H1− − E1− ×H2+ = o(r−2 ). (18) I
E2d × H1+ − E1+ × H2d · n̂ dS,
This relation implies that the reaction (1) over the large sphere h2, 1i = (25)
S2
SR goes to zero when r → ∞,
I which is valid for an empty alternative environment. Note
E2+ × H1− − E1− ×H2+ · n̂ dS = 0. that the components E1− , H1− generated by Antenna 1 and
(19)
SR scattered inside the surface S2 do not contribute to the reaction.
E2− × H1− − E1− × H2− · n̂2 dS. (24) and equivalently for the magnetic fields H1+ , H1− .
S2
The same procedure is repeated with Dipole 2, to decom-
Note that there is no interaction between the fields (·)+ and pose the fields into E2+ , H2+ and E2− , H2− .
(·)− from the different antennas. Based on the decomposed fields, E1+ , H1+ , E1− , H1− , E2+ ,
The generalized reaction (2) can also be separated into H2+ , E2− , H2− , we evaluate two formulations of the reaction;
scattering components. It becomes especially simple if the the original reaction (1), and the reactions for scattered fields
alternative environment where Antenna 2 transmits, depicted (24) derived in this letter. We use a dipole separation distance
in Fig. 2(b), is chosen to be homogeneous (e.g. vacuum) D = 1.8 m, with each dipole placed a = 1.1 m from
MALMSTRÖM: ON THE REACTION THEOREM FOR SCATTERED FIELDS 4
a
- D - a-
Fig. 8. The mutual impedance magnitude |Z21 | calculated with the reaction
h2, 1i as in (1), and (24). For illustration, the integrals (22) and (23) are also
plotted. We see that the agreement of (1) and (24) is excellent and that (22)
and (23) are both very close to zero.
(/m)
(/m)
1 1 1 1 11 1 1 1 1 1 1 11 1 1 1 1 1 1 11 1 1 1 1 1 1
z (/m)
z (/m)
z (/m)
zz (/m)
z (/m)
z (/m)
zz (/m)
z (/m)
zz (/m)
z (/m)
z (/m)
z (/m)