JPRS ID: 9152 USSR REPORT ELECTRONICS AND ELECTRICAL ENGINEERING

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000200060038-2
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 95
• Document Creation Date: November 1, 2016
• Sequence Number: 38
• Content Type: REPORTS

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2
ELtu I _ I u
25 MARCH 1980
CFOUO 5r80a
1 OF 1
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
JPRS L/899?
25 March 1980
USSR Report
EIECTRONICS AND ELECTRICAL ENGINEERING
(FOUO 5/80)
A
FBIS FOREIGN BROADCAST INFORMATION SERVICE
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
NOTE
JPRS publications contain information pricnarily from foreign
newspapers, periodicals and books, but also from news agency
- transmissions and broadcasts. Materials from foreign-language
sources are translated; those from English-language sources
are transcribed or reprinted, with the original phrasing and
other charactEristics retained.
Headlines, editorial reports, and material enclosed in brackets _
are supplied by 3PRS. Processing indicators such as [Text]
or [Excerpt] in the first line of each item, or following the last line of a brief, indicate how the original inforroation was
processed. Where no processing indicator is given, the infor-
mation was summarized or extracted.
_ Unfamil4ar names rendered phonetically or transliterated are
enclosed in parentheses. Words or names preceded by a ques-
tion mark and enclosed in parentheses were not clear in the
original but have been supplied as appropriate in context.
Other unattributed parenthetical notes within the body of an
item originate with the source. Times within items are as
given by source.
The contents of this publication in no way represent the poli- -
" cies, views or at.titudes of the U.S. Government.
For fsrther information on report content
call [phone] (economic); 3468
(political, sociological, military); 2726
(life sciences); 2725 (physical sciences).
COPYRIGHT LAWS AND REGULATIONS GOVERNING OWNERSHIP OF
MATERIALS REPRODUCED HEREIN REQUIRE THAT DISSEMINATION
OF THIS PUBLICATION BE RESTRICTED FOR OFFICIAL USE ONLY.
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
JPRS L/8997
25 March 1980
- USSR REPORT
_ ELECTRONICS AHD ELECTRICAL ENGINEERING
(FOUO 5/80)
This serial publication contains articles, abstracts of articles and news
items from USSR scientific and technical journals on the specific subjects -
reflected in the table ot contents.
Photoduplications of foreign-language sources may be obtained from the
Photoduplication Service, Library of Congress, Washington, D. C. 20540.
_ Requests should provide adequate identification both as to the source and
the individual article(s) desired.
CONTENTS
PaGE
Piarine R1dio Communications Ship Antennas
(OI3'!,ORNAYA INFOI2MATSIYA, SERIYA PRUMYSLOVAYA RADIOELCKTRONNAYA
' API'ARATi1RA I PODVODNAYA TEKHNIKA, SUDOVYYE ANTENNY MORSKOY
IZADIOSVYZAI, No 2, 1979)
1
~ Investigation of the Focusing Properties of Certain Second-Order
Surfaces With the Incidence of a Plane Wave at an Angle to the
Focal Axis
(M.G. Gorlov, L.B. Tartakovskiy; RADIOTEKHNIKA, No 10, 1979)...
3
_ Spectrum Analysis of Ultrashort Wave Signal Level in Long-Range.Tropo-
spheric Propagation Over an Ocean Route
(B.S. Rybakov; ELEKTROSVYAZ', No 10, 1979)
9

Depolarization of Radio Waves in Scattering on Vegetative Soil
(A.S. Skryabin; RADIOTEKHNIKA,,No 10, 1979)
17

Approximate Uetermination of Statistical Distributions of Intensity
of Rainfall
(Ye.A. Larin; ELEKTROSVYAZ', No 10, 1979)
29

- Results of Tests of the `Luch' Data Transmission Equipment on Vessels
(A.A. Borisovskiy; k;KSPRESS-INFORMATSIYA, SERIYA PROMYSLOVAYA
RADIOELEKTRONNAYA APPARATURA I PODVODNAYA TEKHNIKA, No 11, 1979)
36
- a- [III - U5SR - 21E S&T
FOUO]
FOR OFFICIAL U5E ONLY
c
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
CONTENTS (Continued)
Page
Microwave Attenuation in a Line of Metal-Dielectric-Semiconductor-
Metal Composition .
(A.K. Balyko, A.S. Tager; RADIOTEKHNIKA I ELEKTRONIKA,
No 8, 1979) 38
Considerzition of Design and Technological Features in the Realization
of an El.ectronically Tuned Matched Filter.for the Selection of a
Complex Signal
(N.I. Smirnov, Yu.A. Karavayev; IZVESTIYA WZOV RADIOELEKTRONIKA,
Vol 22 No 8, 1979) 45
~
Fifth International Symposium on Electromagnetic Compatibility
Announced
- (ELEKTROSVYAZ', No 10, 1979) 51
_ Discretely Controlled Ferromagnetic Elements for the Conversion of
Electric Power Parameters
(DISKRETNO-UPRAVLYAYEMYYE FEItROMAGNITNYYE ELEMENTY DLYA
PREOBRAZOVANIYA PARAMETROV ELEKTROENERGII, 1979) 52
Oscillation Sources Based on Surface Acoustic Waves
(A.V. Ryzhkov; RADIOTEKHNIKA, No 10, 1979) 55
Some Aspects of the Design of Close-Range Noise Signal Radar Sets
- With Integrated Coverage
(V.V. Grigorin-Ryabov, et al; RADIOTEKHNIKA, No 10, 1979)...... 61
Some Features of the Depolarizing Properties of Radar Targets
(L.A. Zhivotovskiy; RADIOTEKHPIIKA, No 10, 1979) 67
Doppler Filtration in HF Direction-Finding in Combination With a
Method of Analytic Beam Separation
(Yu.M. Agafonnikov, et a1; RADIOTEKHNIKA I ELEKTRONIKA, No 8,
1979) 72
Distortions of the Radar Characteristics of Complex Objects When
Exposed to a Spherical Wave
(V.M. Shlyakhin; RADIOTEKHNIKA, No 10, 1979) 81
Synthesis of a Single-Pulse Discriminator for a Located Target
(N.Ya. Kuz'-; RADIOTEKHNIKA I ELEKTRONIKA, Vol 24 No 4, 1979)... 86
-b-
FOR OFFICIAL USE ONLY
~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
UDC 621.396.676:629.783
MARINE RADIO COMMUNICATIONS SHIP ANTENNAS
Moscow OBZORNAYA INFORMATSIYA, SERIYA PROMYSLOVAYA RADIOELEKTRONNAYA
APPARATURA I PODVODP7AYA TEKHNIKA, SUDOVYYE ANTENNY MORSKOY RADIOSVYAZI in
Russian No 2, 1979 signed to press 3 May 79 pp 1, 47
[Annotation and table of contents from book by S.F. Makarova
TshIITEIRKh [Central Scientific Research Institute of Information and
Technical and Economic Research of the Fish Industry], 47 pages]
[Text] The principles of organization of marine radio communications, the
requirements for equipment and the stipulations for its use have important
distinctive features, as compared with radio communications over land.
These f.eatures are caused primarily by the not too large dimensions of an-
tennas, the limited possibilities for placing them on vessels, and the need
for equipment to operate over a broad frequency range with a continuous
r.fiange in the coordinates of one correspondent (in ship to shore comaunica-
tions) or both correspondents (in ship to ship cocmnunications). The struc-
ture and types of ship antennas depend on the goals confronting marine
radio communications.
The International Radio Consultative Committee (MKKR) has assigned the
following frequency bands for marine communications: MF--405 to 535 kHz;
HF--1605 to 3800 kHz.and 4.063 to 25.600 MHz; and VHF�--156 to 174 MHz.
The 1605 to 3800 kHz frequency band is customarily called the "intermedi-
ate wave" (PV) band.
CONTENTS
MF-Rand Antennas
HI' Antennas
Some Methods of Improving the Effic:Lency of Ship MI' and HF Antennas
VHF Antennas
- Driven Antennas
_ Ship Antennas for Communications
Pr.oblems of installing satell_ite
vessels
Types of supports for antennas
ia Satellite
communications antennas on fj.shing
1
FOR OFFIClAT USE ONLY
Page
1
8
17
23
28
31
32
35
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
rucc urrtw-tu., uar, UNLY
Methods of regulating the position of an antenna 37 ,
Parameters or ship antennas 38
Experimental models of ship satellite communications antennas 40
f:onc 1 ust nns 43
B!h 1 i ogrlphy 44
COPYRI[;HT: Tsentral'nyy Nauchno-Issledovatel`skiy Institut Informatsii i ~
Tekhniko-Ekonomicheskikh Issledovaniy Rybnogo Khozyaystva SSSR, 1979
[98-8831]
CSO: 1860
- 8831
1
- 2 -
~ FOR OFFICIAL USE ONLY
- � la -
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
- UDC 621.396.677.83
INVESTIGATION OT THE FOCUSING PROPERTIES OF CERTAIN SECOND-ORDER SURFACES
= tJITH THE INCIDENCE OF A PLANE WAVE AT AN ANGLE TO THE FOCAL AXIS
- Moscow RADIOTEIaiNIKA in Russian No 10, 1979 pp 65-67 manuscript received
after completion, 4 Sep 78
- [Article by M.G. Gorlov and L.B. Tartakovskiy]
- [Text] In selecting the optimal focusing surface for scanning reflector
antennas it is necessary to know the dependence of the field strength at
the focusing point and of the amplitude-phase distribution af the field in
a focused spot near different surfaces on the geemetry of the reflector
and on the angle of incidence of a plane wave onto it. For the purpose of
cietermining the characteristics of the focusing spot, let us consider the
work ot the reflecting surface in the reception mode [1]. Diffraction cor-
rections for the close-range zone are not taken into account. Concretely,
the problem is formulated in the following manner. Given are a symmetxic
focusing surface with a square aperture, L X L(fig 1), its orientation
in space, and the dj.rection of arrival of tne plane wave. The focused
- Field of the antenna is calculated from points determined in the approxima-
tion 2[n, HPad [incidence]] '
i
y
rigure 1.
1x
1 X~
3
a_Z z,zr

ZZ
FOP, (lFi'IC I AZ. USE UNLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
eUx urElc;lAL U5E UNLY
_M
Results of Calculation
Numerical calculations were made for symmetric focusing surfaces in the
form of a paraboloid of revolution, a sphere and a parabolic torus, whose
aperture equals 50a X SOa.
In cross section through plane xoz (fig 1) the parabolic torus represents -
part of a circle, and in the cross section through plane yoz , a parabola.
Polarization is linear and vector E is placed in plane xoz .
In f.ig 2 are given calculated curves characterizing the shift in the focus-
ing point in relation to the geometrica.l focus in a parabolic reflector
when a plane wave strikes the reflector along the focal axis, as a function
of F/L for different L/a , where L is the linear dimension of the re-
flector's aperture. This shift is explained by the fact that when the ob-
servation point departs from the geometrical focus i.n the direction of the
focusing surface, the amplitude of the field reflected from the reflector
increases, but then a phase error is evidenced. For long-focus reflectors
with a low ratio of L/a the influence of the phase shift is slighter;
therefore the maximum field concentration is produced at a point shifted
toward the reflector; this shift increases with an increase in F/L and a '
reduction in L/a .
_AF ~ ~ �10 ~
0, 4 ~
~ 25
- g2~ / / 50
0,
ne-. ,
0,25 0,5 0,75 F,1L
rigure 2.
In f:ig 3 is given the relative variation in the zield at the focusing point
at different surfaces as afunction of F/L with @=; 0 (F is the dis-
tance From the focusing point to the vertex of the reflecting surface and
0 is the angle of incidence of the plane wave onto the reflector) . A
parabolic ref.lector with k'/L > 1 (curve 1) can be replaced by a sphere
(ctirve 2) or a parabolic torus (curve 3), and the energy at the Focusing
point drops by a total of about 0.4 dB. In order to maintain the same ra-
eio of F/L (i.e., for focusing to occur at the same d:istance from the
4
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
vertex of the surface) for the focusing surfaces considered, it is necessary
to select for each value of F/L its own radius for the sphere and radius
for the circle of the parabolic torus. Here we will assume that
F = F Curves 4 and 5
ar. or p ab lic orusJ ar oloi a Rrpar oloid] '
c~iaracler~z ~ie ~epend~ence of R s�~~ an~ t~ on F/L for a sphere
and parabolic torus, respectively. In terms gf its focusing properties
the parabolic torus is inferior to the sphere. This can be explained by
the fac[ that the raciius of tne sphere is always smaller than the radius of
the parabolic torus's circle upon condition that the distance from the
Cocusing point to the vertex is the same.
''rco. Ri~r 70P17 IEI �~6
F F 3 p s~E~MUn'C 2)
' 0=0; L 50
28J
z'
44t:5
4Zt S 5
~ ~
L~
q2 0,4 o,s qs ia
Figure 3.
Key :
1. R /F; R 2. 20 log (JEI/IEI dB
sf [sphere]~F p.t [para- max
bolic torus]
It should be mentioned that by varying F (F 0 F )
and R it is possible to selectparparafolic ~orustwit~h th~asame
ratio g? r'F~Erawhich will focus better than a sphere.
In fi.g 4 are given the trajectories of movement of the exciter, on which
, are plotted values of parameter 8for different focusing sur.Eaces. The
designations on curves 1, 2 and 3 are similar to the designatfons in figs
- 3-6. With an angle 0 oX from 0 to 8� the tra,jectories are close to one
another and pass almost perpendicularly to the focal axis.
The curves in fig 5 characterize the dependence of the dimensions of the
focusing spot on angle 9for the -3 dB level (curves 1, 2 and 3) and the
-10 dB level (curves 4, S and 6) in the scanning plane. The dimensions
of the focused spot vary chiefly in the scanning plane, and for a
5
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
rux urrtCtAL uSE UNLY
parabolic torus its dimensions in the + 14� scanning sector reiqain prac-
tically unchanged.
r_..,oo
,
1D 1
-P�

. _o
_f4o
F
--
> 1, the focusing propertfes of the sur!'aces discussed are approximately
identical.
7-
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
Bibliography ~
1. Rusch, W.V.T. and Ludwig, A.C. TIZANS. zEEE, Vo1 AP-21, No 2, 1973.
- (:OPYRICIIT: RADTOTEKHNIKA, 1979 .
[59-8831]
CSO: 1860
8831
8
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
UDC 621.317.757
SPECTRUM ANALYSIS OF ULTRASHORT WAVE SIGNAL LEVEL IN LONC--RANGE TROPOSPHERIC
PROPAGATION OVER AN OCEAN ROUTE -
" Moscow ELEKTROSVYAZ` in Russian No 10, 1979 pp 45-48 manuscript received
6 Mar 79
[Art"cle by B.S. Rybakov]
[Text] Introduction
Intermittent fluctuations in the signal level in the long-range tropospher-
ic propagation (DTR) of ultrashort waves (UKV's) are customarily divided into
rapid and slow, distinguishing between 24-h6ur and seasonal variations [1].
Such a distinction is, firstly, arbitrary, and, secondly, some of the cha-
racteristics, as 24-hour and seasonal variations, for example, only quali-
tatively characterize the spectral structure of fluctuations in signal le- _
vel. It would be interesting to investigate the spectrum of microwave sig-
_ nal fluctuations in DTR in terms of a long-duration realization (a few -
years) from the viewpoint of an ana.lysis of non-steady-state random pro-
cesses (NSP's) [2,7]. Unfortunately, the spectrum analysis of NSP's is a
difficult problem which to a great extent has not been solved either theo-
retically or from the equipment viewpoint [2,4,7]. The concept of the
energy spectrum of NSP's is defined variously [7,8].
- In this paper, within the framework of the correlation theory, an attempt
is made to estimate the spectrum properties of the level of a microwave
DTR signal over an ocean route 225 km long, in terms of its annual realiza- -
tion, based on the property of "local stationarity" of the process [2].
_ Method of Calculating Spectrum
Realization of the process can be characterized by a set of local probabil-
- ity characteristics [2):
_ ~I+1
8-= T ~ g[X(tl dr,t)=1Tg
. B t
1
0,1,2,..., N, (1)
9
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
. wliere .g[x(t)] is the estimate's operator; N is the whole part of the
ratio T/T ; T is the length of the realization, x(t) ; and T8 is
the segmed of time in which an estimate is made of 6*(Te) , which does
�ot exceed the interval of local stationarity of the process.
'I'he application of algorithm (1) for the purpose of calculating the prob-
. ability characteristics of rapid fluctuations is familiar [1]. The re-
_ sults of estimates in each segment (Te = 1 to 3 min) have furthermore been _
systematized in terms of some feature relative to the length of the real-
lzation.
in analyzing slow fluctuations, to initial process x(t) can be applied
- a smoothing operator [2]:
i ~
Cpyt [x (t)1= Tx S x(Y) dt', Tu < E' t T
_ t-Tx
, (2)
with the subsequent employment of algorithm (1).
I'}le algorithm f.or estimating the spectrum of fluctuations of a non-steady-
state process employed below is based on the ssccessive multiple applica- .
tion of (1) and (2) to the measured realization of x(t) . As functional
- operdtn.or g[x(t)] for the purpose of estimating the spectral power densi-
ty, C(f) , was selected the method in [31, based on an estimate of the -
correlation function with the subsequent employment of a Fourier trans-
Corm. The selection of this method was due to the fact that it provides
an asymptotically unbiased estimate of spectral density 31, and the inter-
- mediate results of calculation of correlation funetion i(t) are of in- `
dependent interest.
The procedure for spectrum analysis thus reduces tu the following. After
the j-th cycle for applying operator (2), the realization of "smoothed "
- process x.(t) , 0< t< T, is divided intR N. sectio s of length T8. ,
in each ofjwhich is made a calculation of K.i(4) and 8 .i(f) . Here ~
i= 1,2,3,..., and N. designates the numbe4 of the real~zation segment
eing processed. Tha3variance in the curves for correlation functions
~ (T) , of course, causes considerable inst bility over time in local
A ~
(l
nstantaneous ) spectra of fluctuations, ~~(f) . In this connection,
" -i.t has been considered [1,5J that it is advisAle for the purpose of prac-
tice to consider averaged (in terms of the realization length) character-
istics of the signal an3, consequently,
M Nl n N, ~'1 n
Kj ('1 = ' Kji (.0; Gi (f tijt (f 1 �
(3)
d
10
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
Then, to realization x,(t) is again applied operator (2) and process
x.+l(t) with 0< t < ~ is processed similarly. Algorithms (1) and (2)
cAn be implemented either with a computer in processing data, or by the
instrument method [7].
Estimates (3) characterize the averaged correlation and spectral properties
of the process in a definite frequency band, [f , f By the selection
Rf parameters T and T it is possible to ansunN that estimates of
C,(f) for adjacet~t cyclesiAf 3 and j+ 1 have re ons which overlap
w~th regard to frequency, fv i+~ > fn. . Then set ~3(f) ,j = 1,2,3,
...,k in the sense defined ~iSov~ will. t~haracterize the spectrum of fluc
tuations of the measured process.
Calculation Results
- Calculations were made according to the measurement data in [10], processed
and systematized with the traditional approach (24-hour and seasonal varia-
tions, etc.). The r_onditions for conducting the experiment, as well as the
characteristics of the measuring complex are described in the study indi-
cated. Of importance here is the routine for recording the results of
measurements, which were made in two-month cycles: February-March, May-
June, August-September and November-December 1973. Within a cycle the sig-
nal level was measured daily in the morning (0800 to 1200), daytime (1400
. to 1800), evening (2000 to 2400) and nighttime (0200 to 0600) hours.
The xesults of ineasurements for sessions lasting 1.5 to 3 h were recorded
on an NO-36 tape recorder (sessions of type I) and on an automatic recording instr.ument (type II) with an upper frequency limit of 30 and S Hz, respec-
L[vely. The data of sessions of type I were entered into the computer after
digitizing the process in terms of time with an interval of 'At = TO/(5 to
to 10) and in terms of level for 128 gradations of its dynamic range, and
for sessions of type II, after the formation of inean-m3nute sampling values
(T0 is the process's time correlation interval). _
- The nrocessing of sampled values in both cases was carried out in keeping
with the procedure described above, but in order to reveal, in particular,
the correlation and spectral properties of the signal level at intervals
frequently cited in the published data [1,5]. The parameters of some ave-
raging and processing cycles are illustrated in the table. The 0.5 to 3
min section corresponds to the interval for the stationarity of rapid fluc-
tuations [1], the 1.5 to 3 h segment was selected as the stationarity in-
terval (prior to the appearance o.f a 24-h variat3on) for mean-minute values,
and the 5 to 7 24-h-period segments, for mean hourly values, is contained
_ within the limits of the synoptic period [6]. T e was selected within
the limits indicated in the table for a specific4ealization segment, in
keeping with the stationarity test in [4].
The results of calculation of the averaged spectra, ~f1, for different
time intervals are given in fig 1, where the designat. (~dn� of the curves
11
FOR OFFICIEIL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FUIt UFFICIAL USE ONLY '
,
cor-respond to the number of the subscript, j, of the averaging cycle. The
spectral pnwer density, ~(f) , is a Fourier transform of the coYrelation _
funrtion (f.ifi 2d) compiiteA for a yearly realization of 83 samp ed values, _
~iver~~gcct For three 24-h periods, of. the gignal level. Curve ~~(f) is the
;iveY�zged spectrum of. Clucl-uations of inean 24-h values oL- the signal level ,
_ [n two-month measuring cycles: February-March, May-June, August-September
nnd November-December.
Tabl.e
- Length of Digitiza�� Samtpled Sampled Number of averaging
segment, tion in- values, values, segments, N.
T terval, x.(kAt.) x.
ej ~t. J J ]-1 ;
(kA tj_1)
0.5 to 0.05 to Measured - 2000
3 min 0.5 s (type I) '
1.5 to 1 min Mean-min- Measured 195 ~
3 h ute (type II)
5 to 7 1 h Mean hour- Mean minute 22
24-h ly periods
1 year 3 24-h Average Mean hourly 1
periods for 3 24-h ~
periods ~
The correlation functions for each cycle are given in fig 2a by curves 1,
9,3 and 4, respectively, and the correlation function averaged for them,
~(T) , is given in fig 2b. The correlation function for mean hourly values of the signal level in 5 to 7 24-h intervals, averaged for 22 segments of
ehe yearly realization, is given in fig 2c, and the corresponding spectral density is illustrated in fig 1 by curve 4.
The illustrations in fig 3 represent the results ot calculating averaged
correlation functions in terms of inean-minute samplc.d values in 1.5- to
3-h intervals for different months. The nismber c" aver.aging segments i
equals 26 in May, 19 in August, 29 in September, 68 in November and 53 in ~
December with a total number of sampled values greater than 20,000. The
averaged spectral density co responding p them is given in fig 1 by curve
3. The speceral densities, ~2(f) and U 1(f) , of rapid fluctuations in
- the signal level in minute time intervals obtained by averaging for 2000
segmenrs are illustrated by curves 2 and 1, respectively.
12
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
G
6
. S
f0 ~ - - -
~ A B
105 I
C
3
D
1p3
10
/
i)
10'6 fo-f 10'2
rigure 1.
- Key:
1. f , Hz M M
- The lack of agreement of curves G.(f) and G,+1(f) at the "boundary" is
_ explained firstly by the diffE;rent~amount of aaeraging for the set of seg-
ments and consequently by the~non-identical degree of reliability of the es-
timates obtained and, second~;, by the fact that the resolution of a spec-
trum analysis for different ~.(f) is non-identical (by virtue of the method
chosen for calculating the speltrum, and with a growth in j equals, re-
spectively, 2.5�10 1, 5�10 2, 2.08�10 4, 6.33.�10 6, 4..83�10 7 and 1.93�10 8
'Hz. As is obvious, at the "boundaries" the frequency resolution changes by
- approximately an order of magnitude, and consequently estimates of spectral
density relating to a single frequency but belonging to different curves
have not been computed under equivalent conditions. This fact is important -
for the purpose of isolating peaks of the spectral power density reflecting
"periodic" changes in the signal level.
_ The distinctly pronounced spectra.i, density peak with tip A(fig 1) corres-
ponds to a 24-h variation, and the peak with tip D, to its second harmonic.
The sharp-point nature of both spectral density peaks, of approximately
13
FOR OFFICIAL TJSE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
r'ux ur'YlUi.ai. U5t UNLY '
i.dentical stiape, characterizes the "regularity" of the 24-h variation over
the course of a year. Spectral density peaks with tips C and D, respec-
tively, at frequencies of 8�104 and 3.1�10 3 Hz, characterize changes in
the signal level with a period of 20.8 and 5.3 min. The position of peak
C in relation to the set of local spectral densities of fluctuations in ~
the mean-minute values of the signal level is sufficiently stable, but its
value and width vary substantially in relation to the length of the yearly
realization. The position of peak D, on the contrary, is unstable.
a,
Ozl
bj G=B,~d6
0,5
0 1 .2 3 4 5 Cym#u
Bl
o,s
' D 3 6 9 12 15 yaca
x(rl
g~ C'i =9,3d6
0,5
0
l' 10 20 30 40 50,CymHu
Figure 2.
Key:
l. dB 3. hours
2. 24-h periods
A physical interpretation of these peaks is difficult in view of the little
investigation of the spectral properties of ineteorological parameters, in
particular, the refractive index [9]. The expected spectral density peak
at a frequency of 3.22�10 8, reflecting seasonal variations in the signal
level, was not reproduced because of the limited nature of the length of
the measured realization.
14
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
X
~iau; ~ ~ a61
0,5
D
X
ABzycm; G=4, 2d6
0,5
D
N
CeHn~abpa;~=3,6d6
0,5
D
X
NoA6pa; ~7d6
45
O
X
DS TFI aeKQbpa; 61=3,7d6
0 5 10 15 ?D 25 � h', MlIN
Figure 3.
Key:
1. May, dB
2. August
3. September
4. November
5. Decemtier
6. T , min
The dependence of the sp ctral dens~.ty on frequency (fig 1) is approximated
well by power function ~(f) = f 7 5. The absolute value of the exponent
- here proye ; d to be somewhat lower than the similar parameter for approxima-
:
tion of tt`e spectral density of time-dependent fluctuations in the refrac-
tive index''[9]. However, the spectra analqzed in [9] were arrived at by
processing only individual realizations and characterize intermittent fluc-
- tuations in the refractive,index to frequencies not lower than 4.2�10 3 Hz.
Let us note that the spectral form f-3r representing the results of ineasure-
ments of the signal level makPS it possible to estimate quantitatively the
strength of fluctuations in the signal level in any frequency band of in-
_ terest to the investigator.
15
FOR OFFICIAL USE aNLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFEICIAL USE ONLY
In conclusion the author wishes to express his thanks to G.A. Kalinov and
A.A. Kashkan for their helpful discussions of this article's material and
for their assistance in computations and in processing experimental data.
Bibliography
- l. Vvedenskiy, B.A. et al., eds. "Dal'neye troposfernoye rasprostraneniye
UKV" [Long-Range Tropospheric Propagation of Uttrashort Waves], Moscow
_ Sovetskoye Radio, 1965.
2. Tsvetkov, E.I. "NestatsionarnyyP sluchaynyye protsessy i ikh analiz"
[Nonstationary Random Processes and Their Analysis], Moscow, Energiya,
1973.
3. Gribanov, Yu.I. and Mal'kov, V.L. "Spektral'nyy analiz slucaaynykh
_ protsessov" [Spectrum Analysis of Random Processes], Moscow, Energiya,
1974.
4. Bendat, Dzh. and Prisol, A. "Izmereniye i analiz sluchaynykh protses-
_ sov" [Measurement and Analysis of Random Processes], translated from
English, Moscow, Mir, 1974.
5. Shur, A.A. "Kharakteristiki signala na troposfernykh radioliniyakh"
[Characteristics of a Signal in Tropospheric Radio Lines], Moscow,
Svyaz', 1972.
6. Khromov, S.P. and Mamontova, L.I. "Meteorologicheskiy slovar [Meteo-
rological Dictionary], Leningrad, Gidrometeoizdat, 1974.
7. Vollerner, N.F. "Apparaturnyy spektral'nyy analiz signalov" [Instru-
ment-Assisted Spectrum Analysis of Signals], Moscow, Sovetskoye Radio,
1977.
8. Ceranin, V.A. "Spectral Representation of Nonstationary Random Pro-
cesses" in "Trudy IV Vsesoyuznoy shkoly-seminara po statisticheskoy
gidroakustike" [Proceedings of the Fourth All-Union Training Seminar
on Statistical Hydroacoustics], Novosibirsk, USSR Academy of Sciences
Siberian Division Institute of Mathematics, 1973.
9. Kazakov, L.Ya. and Lomakin, A.N. "Neodnorodnosti koeffitsienta pre-
- lomleniya vozdukha v troposfere" [Inhomogeneities of the Refractive
Index of Air in the Troposphere], Moscow, Nauka, 1976.
10. Korneyev, I.L. et al. "Variability of the Mean Level of an Ultra-
short Wave Signal in Long-Range Tropospheric Propagation Over the
Ocean," ELEKTRUSVYAZ', No l, 1979.
_ COPYRIGHT: Izdatel'stvo Svyaz', ELEKTROSVYAZ', 1979
[62-8831]
8831
CSO: 1860
16
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
UDC 621.371:621.396.96
DEPOLARIZATION OF RADIO WAVES IN SCATTERING ON VEGrTATIVE SOIL
Moscow RADIOTEKfINIKA in Russian No 10, 1979 pp 24-30 manuscript received
14 Dec 78
[Article by A.S. Skryabin]
[Text] Introduction
- In recent years a considerable amount of studies have been devoted to ana-
- lyzing the polarization properties of radio waves [1-4, etc.]. However,
the question of the depolarization of radio waves in scattering on actual
terrestrial soil has remained open, although it is a decisive question in
many practical problems in utilizing their properties. Below a theoreti-
cal and er.perimental investigation is made of the depolarization of radio
waves in scatte.ring on vegetative soil (a forest) and an estimate is made
of the relative influence of various factors on the amount of depolariza-
tion. Discussed is a radar case for a sighting angle relative to the nor-
- mal line to the mid-level of the base. As demonstrated by the experiment-
- al investigations in [5,6], vegetative soil causes the greatest depolari--
zation of radio waves. For theoretical investigations scatterers of this
source can be represented in the form of a three-dimensional structure
consisting of disks and linear reflectors. .
In the scattering of radio waves on vegetative soil, depolarization is
caused chiefly by the edge wave, anisotropy of the reflector and multiple -
scattering. By anisotropy will be meant a difference in the shape of the
reflector from a regular spherical. If it differs in one coordinate, then
the reflector.has the shape of an oblate ellipsoid, changing at its end
- into a circular disk. For plane reflectors there occurs a difference in
two coordinates, and the reflector can change shape from a round disk to
an elliptical one, changing at its limit into a linear reflector.
Theory
Let us investigate the depolarization of radio waves in scattering in a
three-dimensional structure consisting of circular ideally conducting
disks and linear reflectors. Let us consider the case of ka < 1 and
17
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
ka > 1 (where a is the radius of the aisk or half the length of the
linear reflector and k= 2ff/A is the wave number).
For a disk the case of ka < 1 is investigated on the basis of Eggimann's
study [7] and the case of ka > 1 by the method of edge waves [8]. The
geometry of scattering is repr.esented in fig 1. Axis Z is perpendicular
- to the surface of the ground and axis Z to the plane o9 the disk.
Zo } Eyo
r�~
Figure 1.
ror the purpose of estimating the amount of depolarization we employ the
depolarization coefficient,
s t
Dka) El ~~~.7, ka)I 1,
~
. E (`407ka) j ~
where Ei(~ ,Y.ka) and EII(~ ,Y,ka) are components polarized orthogonally
to and paraQlel with, respectp ively, the radiated wave.
Flrst were obtained dependences of the degolarization coe�ficient on as-
pects ~Q and y and dimension ka for the case of scattering on a
single refle�t r. Then was computed the mean value of the depolarization
coefficient,IG~, for an ensemble of identical disks oriented with equal
probability. Let us assume that the disks occupy a space with dimensions
much greater than X and are widely spaced so that it is possible to dis-
' regard their mutual influence ;shading and multiple scattering). Since
the signals from individual scatterers are put together as incoherent sig-
- nals, the equation for the mean value of the depolarization coefficient
has the form
18
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
.x T, t
l'
~
f El~ 2 slnYd?�dY i
o (T�. 7)
2: .
J J~ (T., 'f)f sin YdTod7 B
40=0 71
(1)
An experiment has demonstrated that the range of the aspects of reflectors
_ (leaves, needleG and branches) equals ~0 = 0 to 2n and Y=w/2 ; con-
sequently, the integration limits in equation (1), in calculating the mean
value of the coefficient of depolarization from a forest, must be taken
as from 0 to 2n for ~0 and from 0 to n/2 for 7. For a low-
frequency (LF) approximation, the integrals for ~0 and Y are reduced
to tabular and the solution is arrived at analytic$lly. Por a high-fre-
quency approximation, the integral for ~p is reduced to tabular and in-
tegration for Y is carried out on an M-Z20 computer. Th` results of
calqu ations are presented in fig 2, in which are illustrated dependences
of ID = f(ka) for 1) Y= 0 to 90� and 2) Y= 0 to 20� (the latter
case is characteristic of dry land in the absence of vegetation and of
an ocean in the absence of whitecaps). The preliminary results arrived
at by means of a model of an ensemble of randomly oriented identical disks
are given in [9].
Let us evaluate the accuracq of the theoretical procedure. Matson [10],
hav_'ng investigated the error of the Eggimann method as compared with a
strict solution, demonstrated that with ka = 0.5 the discrepancy equals
_ 0.1 dB. With ka = 1 can be expected an increase in the discrepancy to
a few tenths of a decibel. An estimate of the error of the high-frequency
(HF) approximation was made by M.G. Belkina [11], T.Ya. Ufimtsev [8] and
others for ka = 5, 3 and 1. It equals fractions of a decibel and has
a different sign with different sighting angles. Therefore in averaging
over the entire rangp of angles this error is compensated to a consider-
able extent. With ka � 1 the error of the theoretical procedure equals
less than 0.1 dB.
In the vicinity of point ka = 1 the graphs of the LF and IiF approxima-
tions are conjugated smoothly.
- Let us estimate the error of experimental methods. In determining the
effective areas of scattering (EPR's) of bodies of simple geometrical
shape, the accuracy is rarely higher than 30 percent (about 1 dB) [12].
_ The error of absolute methods of ineasuring the EPR's of terrestrial soil
19
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/48: CIA-RDP82-00850R000200064438-2
rUK UFFICIAL USE ONLY
equals 3 to S dB, and in calibrating with reference to a standard reflec-
tor it is reduced tW 2 to 3 dB. In comparative measurements of the depol-
arization factor,IDl, by the compensation method the error is about 1 dB.
D
�c ~
i
'04
j
f / 7~'~ 2
~ ~
~
e=_-
1Q S 10� S 117 S kQ
Figure 2.
Since the error of theoretical methods with any ka is considerably lower
than the error of experimental methods, then the theoretical procedure is
app].icable with a degree of precision sufficient for practice with any
ratios oE the dimensions of reflectors and the length of the radio wave.
The mean value of the depolarization coefficient for an ensemble of linear
reflectors oriented with equal probability ha.s been computed by disregard-
ing mutual shading and multiple scattering. The geometry of scattering
for a single linear reflector is illustrated in fig 3. .
- Figure 3.
- For a low-frequency approximation the expression for the mean value of the
depolarization coefficient for all orientations of linear reflectors has
- t}ie form [13]
20
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
:x *s
s!n' To cos' TodtP. ~ sin' 7dY
`�-o T=o
Ua.e ~T~ , 7) � ( � 4+c sf ~
l E~(~�7)~2 - t cos' cpody, S sln� 7d7
ro =0 1=o '
- ~ 0, 58,
� y'3
_ (2)
Consideration of the HF approxi.mation, when the length of the reflectar
equals R. � a and its diameter equals 41 � a results in the same final
- results for the mean value of the depolarizatian coefficient as in the
LF approximation (equation (2)). The results obtained for linear reflec-
tors are presented in fig 2(cf. graph 3, where the X's represent the low-
frequency approximation, tlie'diamonds represent the high-frequency appro-
ximation, and the plusses the results [16]).
According to fig 2(curve 3) IDl I = 0.58 and does not depend on ka and
Y. This was arrived at as theresult of the fact that the idealized case
was considered, when the diameter of the linear reflector is much less
than its length, 0 �t and a, in disregarding the influetice of
cross currents and the end surfaces. In taking into account the finite
value of the cross section an he influence of end surfaces, there should
be evidenced a dependence of ~D~ on Y and k~_land with an increase in
ka should be observed a drop in the value of ~D . According to fig 2
(curves 1P4 2) the magnitude of the mean value of the depolarization co-
efficient,IDI) , in scattering in a cluster of circular disks, depends on
ka and Y: dWith ka < 1 this dependence is slight, and with ka > 1
it is of an oscillating decrement nature.
The above makes it possible to estimate the relative influence of the edge
wave and of the anisotropy of the reflector on the value of the depolari-
zation coefficient for the most typical cases as a function of the shape
of the reflector, the range of aspects and the dimensions.
An increase in anisotropy--a change in shape from a disk to a linear re-
flector--causes an increase in the depolarization coef�icient during
sca t ring in this reflector and a change in the nature of the dependence
- of ~D~ on ka and ~0 and y.
In the case of a cluster of reflectors situated perpendicularly to the in-
cident field (y = 0 in fig 1 and Y= 90� in fig 3) and oriented with
21
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
cVL%. ~rt t~.leu~ UJZ V1VL1
,
equal probability relative to an angle of turn, from 0 to 27
depolar[zation is deter i ed total].y by the aniso?ropy of. the reflector.. _
[~or a clustTrloF disks ~Da A 0 for any ka and For a cluster of linear _
reflectors D1 0.58 for any ka .
- tn scattering in a cluster of reflectors distributed csith equal probabil- -
_ ity over all orientations, the magnitude of the mean value of the depolar-
ization coefficient is determined by the edge wave and the a isotropy of
the reflector. According to fig 2(curve 1) t e value of IDT caused by
_ an edge wave equals 0.23 in the LF r.egion a�d ~Dd ti(ka)1~2 In the HF
region. From a comparison of graph 1 and 3 it follows that an increase
in the anisotropy of the reflector (transition from a disk to a linear re-
flector) increases the mean value of the depolarization coefficient from
0.23 to 0.58 in the LF region and by YIK~ in the HF region.
Let us consider the more common case when the ensemble consists of disks
of different sizes and let us determine the mean value of the depolariza-
tion coefficient of radio waves in this case. For this it is necessary in
the numerator and denominator of equation (1) to integrate also for ka ,
taking into account the distribution density of this parameter,
y IE 1 ku)J', 1
Dn.
!"iU) - - I_
I 1..II~~oTkU)J~
l
_ ka, 2% x12 I
E1 (fo,7.ku)]'W(kn) sin Yd?,dYd (ka) Z
S S ~
ku~ po-0 7_0
= kQ, 2w A 1_ _
f Eu(yo,7~ ku)12 W (ku)sinYd?,d7d(ka)
~ l
fo=o i=o
ka, 1
~ alV (ka) d (ka) ?
kQ- ka, �
S BW (ka) d (ka)
ka,
(3)
The distributions of the dimens~ons of reflectors in real terrestrial soil
are rather complex: In fig 4a is given the distribution o� the dimensions
of the leaves of a birch tree, in fig 4b of an aspen and in fig 4c of an
oak. These curves were obtained experimentally for the woods in the Mos-
cow region in August, when the growth of leaves has stopped. Along the
X axis are plotted numerical values of the length of a leaf without the
stem, R, expressed in centimeters and in values of ka for wavelengths
of 3 and 36 cm, at wYiich the experiment was performed [6].
22
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-00850R040240060038-2
FUR OFFICIAL USE ONLY
~
Figure 4.
- It is a complicated af�air to arrive at a mathematical expression for the
distribution function, in view of which a numerical integration was per-
formed in terms of ka . In equation (3) the numerator and denominator
under the square root sign are proportional to the EPR's of the ensemble
of reflectors for orthogonally and parallel polarized components. Since
. these values are of independent interest, it is a good idea "to compute
them and compare them with the experimental data [4,6].
As an example was computed the matrix of EPR's of a birch forest.
For waves with a= 3 cm , ka > 1, and consequently it is necessary to
use the equations of the HF approximation: _
. .
a; e="Z= ~ al ABW (ka) n, oa~B -2
L(li B, !Y~ (ka) n,
l-1 tcl
where vl and QII are the effective scattering areas of the ensemble
of reflectors for tie vHF approximation, A and B are the numerator
and denominator under the square root in equation (Y) for the HF approxi-
mation, n is the number of reflectors in the ensemble and r is the -
number of division intervals for ka .
For waves with a= 36 cm , ka < 1; therefore the equations of the LF
approximation are valid:
23
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2
A-3cK 2,05 6,15 f0,75 1445 , ka`
36cy Q.'6~r ~4SZ 0,82 11~ ka

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200060038-2
FUIt Ur'FIC1AL U5E.ONLY
r r
40 ar A lw' (G a) n, 13
c r-~~~k C1c B�W ( ka) n,
` e� H- 13~~ z;..J i d a.~i ~
where oa n and oll n are the effective scattering areas of the ensemble
of reflecfors for t~ie LF approximation and A and B are the numerator
and denominator under the square root in equaPion (1) ?or the LF approxi-
mation.
In the calcuations it was assumed that in the spot irradiated there is a
single adult birch about 15 m high with a top diameter of approximately
15 m and with n= 104 . The results of calculati~ns of the EPR of the
top of a birch for parallel polarized component cs~~ are given in the
fiurth column of table 1, and for the orthogonally polarized component,
cs , in the second column. In the third and f3.fth columns are given
specific EPR's (UEPR's), in the sixth experimental v 1~�es of UEPR's from
[4] and in the seventh column, calculated values of ~DI .
Table 1.
 ~ 2 ~ s I a I 5 I ~ I 1
A~ c%A
I ol, ~i~ I
I
,yll u~ I
o0~
I Q0~� la) I
~pI
3 I
1,24 ~
0,027 I
27'6
0,6
I 0,1-0,8 (
0,212
36
I 9�10-1-
I 1,9�110-6 I
1,7�10-3 l
3,8�10-3 (
0,5-2 I
0,22
TYie calculated values of UEPR's for a wave with a= 3 cm are of the same
order of magnitude as the experimental from [4]. This demonstrates that
in the scattering of waves of this range on a forested area the major in-
Fluence is exerted by the leaves of the trees.
For a wave with a= 36 cm the cal.culated values of UEPR's are lower than
the experimental from [4] by approximately four orders of magnitude, i.e.,
the leaf cover practically does not play a xole in the process of the
scattering of radio waves of the 36 cm band on a forested area. A decis-
ive role in the process of scattering of radio waves with a= 36 cm is
played by the earth base. The results of experiments on the attenuation
of radio waves in passing through a forested area [14] also confirm that
vegetation causes relatively slight attenuation of radio waves with X_
=36 cm.
24
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2
~
FOR OFFICIAL USE ONLY
Experiment
Measurements were made of the mean values of depolarization coefficients
in scattering on a forested area for a sighting angle along the normal
line to the ground surface. The experiment itself and the apparatua were
= described in [6]; measurement results are given in secs 1 and 2 of table
2, where SD is the root-mean-square deviation in the depolarization co-
efficient and d(percent) is the root-mean-square value of the total
instrument error of the measurements.
Table 2.
n
TNn orpsMCareaA I
a, cr f
~
IpI (
SD
2I NCTOVNMK
I I 3)
3 I
O,S6I
0,14 I
9 I
161
I ~9ec C~teWBHHW~I C nFicr-
e
i
I
3 I
0,8 I
0,351
21 11151
3ncnep11nteH-
Taneewe

.9rc cNicwaiIewii c yac-
3 I
0,37
I U, UG
I 9
I 161
R8HH612
TN4H0 Ofl8B1l1c11 JINCTBOFI
4)
36 I
0.08 I
o,t+os
l 17
I16;
I
8)
5)
a 200 gradually increases
to 5 with probability of occurrence of 2, such rel ations as R;, (D;, (j=1,2,3,4) are possible where
B>D. Thus, no two-wave solutions exist. This inequality may
serve as one of the criteria of a situation in which the
measured interference field is composed of a greatc;r nuriber of
waves than the order of magnitude of the applied solution. The
physical soundne ss of the results obtained usually serves as a
practical criterion for the presence of a greater number of
- waves, for example, when there is an absence of any significant
variation in the results.
The simultaneous equations of type (7) for N>2 cannot be re-
duced to a simple equation. Thus, for N=3 we have
(0, exp (-iavl) -G,) exp (-iavZ) - (f!, egp (-iav,) -D',) X
(D': exp (-iav, ) -C/,) esp (-iav2) - (0, esp (-iav,) -f7,)
X e-;��,= 1,
(12) (D`, esp (-iav2) -0,) exp (-iav,) - (l1'3exp (-Savs) -II' b) X
(fJ: exp (-iav,) exp (-iav,) - (U4 exp (-iavZ) -0'e)
X e_rau, = 1,
~ (C7, exp (iav,) -C:) exp (--iav,) - (0, exp (=iav,) -U,) X
(l7: exp (-iav,) -t7``) esp (-i�v,) - (0, esp (-iav,) -t78)
X e_iau- = 1. .
Thus, we undertake other methods of deriving an analytic so-
lution. Redundant information is usually used for this-a num-
ber of required antennas greater than 2N. Therefore, in [10]
the proposed method leads to linear simultaneous equations in
which, for the c ases 2K and (2K+1) waves, 5K and (5K+3) an-
tennas, respectively, are required. The use of this method
makes it possible to find only the vertical angles of arrival,
assuming that the azimuth angles of the various waves are co-
incident.
76
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/48: CIA-RDP82-00850R000200064438-2
FOR OFFICIAL USE ONLY
P~ M 1 St
~ Z PZ M? sz
SC AD
4 3 ELP3 M 3 S,i
P 4 M 4 Sy
I II I!I ~ -
Fig. 1
I- recording of the field at transducers distributed in space;
II - multichannel coherent pickup; III - measurement of the
- complex amplitude; IV - spectroanalyzer; SC - special computer
- for calculating the wave parameters of each of the spectral
- components; AD - analysis and display of the information.
There is another analytic method for sepaxating the waves in
- which the number of antennas can be 1/T axe successfully filtered out. For eliminat-
ing the residual noise in those components worthy of attention
( that is, of sufficient amplitude) they solve simultaneous
equations of type (6), as in the method of analytic separation.
This is done not to solve for the instantaneous values of thE
complex amplii;udes of the field U(x,y), but for the complex
spectral amplitudes S(W)xY of the selected frequency w.
Allowing for prior poppler filtration, it can be assumed that
in the selected w-filter no more than two components axe inter-
fering. This requires four antennas, which is only one more
anterula than would be necessary in the most minimal configu-
ration for a direction-finder's antenna array. With the ar-
- rangement of antennas chosen in Paxt 2, the solution to the
simultaneous equations relative to parameters u,v can be obtain-
ed fron:� the relations -
3, (w)etp(-iau,)(t))
3`, (co)exp(-iau,)(~) exp(-iav:)=1,
(w) exp (-iattz)
(c~) exp (-iau.) e%p ~-iav,) =1.
The sequence of operations is depicted in fig. 1.
- As an example we will examine a typical Doppler spectrum of -
a reflected HF signal of 7336 kHz on a medium-width route of
4100 km C51 (fig. 2). The spectrum is obtained using a time
factor T=40 sec, 0.-"=0.025 Hz. In this case, for the applica- -
tion of the suggested method to determine the parameters of
- five interfering modes (N=5) the width of the spec tral line
is 0.025 Hz; solutions for the modes in the spectral lines were
_ not�derived for two modes, thus M=2. When utilizing a multi-
wave sirection-finder that measures the parameters according
- to instantaneous readings, the signal has to be received in
_ a frequency band OF>1.5 Hz. The resultant gain in the fre-
_ quency band is not less than AF/Af=60 times. The number of
transducers employed decreases from 10 to 4.
The essential feature is not only the reduction of the series
_ in system (6) as a result of prior poppler filtration and, cor-
- respondingly, a decrease in the number of antennas and a sim-
plification in the computational aspect of processing the sig-
, na1. No less important is the sharp increase in interference
78
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000240060038-2
FOR OFFICIAL USE ONLY
immunity of the measurement, provided that the latter is de-
termined now by the value Af=1/T�(f'~ax -fm,~ The stability of
L-he system's solution in (13) is also much greater than the
stability of the system's solution in (9), since (9) is solved
~ relative to the instantaneous values of the complei& amplitudes
U(x,y), , while (13) is solvEd relative to the complex spectral
amplitudes derived from integration over a sufficiently
large interval T(approximately 1000 readings U(x,y),
: 2FModes
( Lbwe
7; beam~~.
P' 400
~
- o
o a �Modes
i ~TPper
},.P be am )
ual
m
~ 0 200 rilusual
r-i
~ ~ ~_P ~
r-i+�, ,vModes
2fModes
v -3 -2 -1 0 1 2 3
Radio frequency
~ shif t , Hz
Fig. 2
`I'ho possibi.l i 1;y of achieving as analytic separation of the modes
~ iri addition to the Doppler separation permits shortening the
_ time of analysis to a value of ^-10-20 sec, which is pa.rticu-
larly important in practical applications and in investiga-
~ tions of the non-stationary iono sphere.
1'he authors are grateful to L. A. Lobachevskiy, V. S.
- LoUachevskaya and I. S. Volginaya for their constant atten-
'tion and assistance in this work.
BIBZIOGRAPHY
1. A.fraymovich, E. L., et al. In "Issledovanya po geomagnetizmu,
= aeronomii i fizike colntsa" LInvestigations of the Geomagne-
- tism, F.eronomy and Physics of the Sun], Vyp's 32, 33, 41, No's
33, 55, 50, Moscow, Nauka, 1974, 1975, 1977.
- 2. Bennet, S. M. US Patent, N 3991. 418, Nov 9, 1976.
3. Afraymovich, E. L. RADTOTEKHNIKA I ELEKTRONIKA, Vol 24, No 7,
= 1919, P 1444.
79
FOR aFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY '
4. Afraymovich, E. L., and Panchenko, V. A. "Dopplerovskaya
_ fil'tratsiya v KV-pelengatsii. Anp.'Logo-tsifrovaya -
obrabotka modulirovannogo signala" [Doppler Filtration in
HF Direction-finding. An Analog-numeric Treatment of the
Modulated Signal], IZMIRAN SSSR, No 18, 1978, P 217� '
5. Shepard, R. A., and Jomax, J. B. IEEE TRAN S. COM. TECHNOL.,
CoM-15, No 2, 1967, p 268.
6. Bayuklina, M. F., and Krasnov, V. M. In "Teoreticheskoe i
eksperimental'noe issledovanie r�asprostraneniya -
dekametrovykh radiovoln" [Theoretical and Experimental In-
vestigation of Decimetric Radio Waves]. Moscow, IZMIRAN
sssR, 1976, p 73. '
7. Boldovskaya, I. G. GEOMAGNETIZM I AERONOMIYA, Vol 19, No 2,
1979, P 251�
~
8. Kukes, I. S., and Starik, M. E. "Osnc~y radiopelengatsii"
[Principles of Rad?.o Direction-finding]. Izd. Sovetskoe
radio, 1964.
9. Baur, K. FREQUENL, `Jol 14, No 2, 1960, p 41.
1.0. Kel:;o, J. M. RADIO SCI., Vol 7, No 2, 1972, P 245�
11. I3aur, K. Bundesrepublik deutschland, Patentschritt N 2328720, -
z.1o.1975�
12. Agafonnikov, Yu. M.; Afraymovich, E. L.; and Polimatidi, V. P.
"A Method of Measuring the Parameters of a Multimode Wave ~
Field," in "Otkrytiya, izobreteniya, p-romyshlennye obraztsy,
tovarnye znaki" [Discoveries, Tnventions, Industrial Design s]
Authors License No 652489, Vol 10, 1979, P 172.
COPYRIGHT: Izdatel'stvo "Nauka," "Radiotekhnika i elektronika,"
1979 '
_ C23-9512]
9512
CSO: 1860
so
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007102/08: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
UDC 621.396.96.06
. DISTORTIONS OF THE RADAR CHARACTERISTICS QF COMPLEX OBJECTS WHEN EXPOSED TO
A SPHrRICAL WAVE
Moscow RADIOTEHIiNIKA in Russian No 10, 1979 pp 67-69 manuscript received
- after coinpletion 9 Jan 79
[Article by V.M. Shlyakhin]
_ [Text] A key objective in the further development of equipment for measur- -
ing the radar characteristics (RLKh's) of objects in models is the develop-
ment of a procedure for estimating and taking into account measurement
errors [1]. For this purpose, in this study an analysis is made of dis-
tortions of the statistical RLKh's of complex objects, representing a ran-
dom combination of interconnected reflectors, when exposed to a spherical
wave.
Ler there fall onto a system of interconnected reflectors (fig 1) a wave
spherical in terms of phase with a radius of curvature of its front of
R0 . Representing the field scattered by a complex object when exposed to
a spherical wave as the result of the superposition of fields scattered by
eacli elementary reflector, we determine
N
el [2krQ sin (a + pO iyn coss (a } 9n) -F Pn {al
L R .
n=t
where Q(a) is the monostatic diagram of the secondary xadiating element
of the nnth reflector; R(a) is the phase diagram ok the n-th reflector;
N is the number of reflectors; and r and ~ are the geoiqetrical char- -
acteristics of the system of reflectors; and *n = kr2 2R
n n~ 0 '
If the linear dimensions of each n-th reflector are small as compared with
= Ro ,,then the wave striking it is locall.y plane, and characteristics ctn(a)
and (3 (a) do not differ from the corresponding characteristics of re-
flectors in exposure to a plane wave.
I
81
FOR OFFICIAL USE ONLY .
-
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
a v�. va a tvi~au a~ua~ vl\L~ -
Z
n n7
6ra ~
' rN
O RO
tiN /al
R~ RY
Figure 1.
Since under real conditions the signal scattered by a target is of a random nature, in processing the results of ineasurements of the RLKh's of objects in models, the angle of the radiation, a, is assumed to be random and
statistical RLKh's are computed. We will assume that the distribution law '
for random parameter a is given by the equation
exp [d CO5 (ao - n)]
W ~a~ 270l0 (d) , To < "(2)
wherc Y is the averaging interval, I(d) is a modified Bessel function,
and a0 ~and d are characteristics of~the mean value and the variance of
random parameter a .
In investigations of the RLKh's of objects one is usually interested in ,
their averaged values over the entire possible range of angles of exposure
(-n to +ff) or in a fairly narrow sector in relation to the selected direc- '
tion. Therefore we will restrict ourselves to discussing two practically
important cases: y0 =7r and yo < O.l7r . -
From (1) it follows that with random paramete~ a the quadrature compon- ~
ents of the harmonic signal, z(t) = Re {EOe1W scattered by a system of
- r.eflectors in exposure to a spherical wave, are also random. Furthermore,
N . _
x0 _ x" - - 2~' (a) sn c052 an) sin [?krn sin an ~n cos= an Pn (a)j I ~
, n=1
N
yo - YR � 2~j 1' ~n (a) Sln (,.',in COSZ an) COS [2krn Sfitta~ -F YR CoS= an �rn (a)],
ncl
(3)
82
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
whrre xa , yo , xp and yp are the quadrature components of a signal
- scattered hy a combination o,f r.e,flect4zs in exposure to a sphericul and
plane wave, and a n = a-h ~ n .
With a sufficiently great number of reflectors, according to the condit6ons
oE the0 central limit theorem, the distribution law for random values x
and y is asymptotically normal with a mathematical expectation of
(X�) (.rn), (Y�) (V,0
n=1 n=1
and a variance of
N N .
D!.r") =~pIxnI, D{yo}�- ~U{y'n
n-1 fl_1
' Averging (a) in keeping with 62), we determine that generally # -
# ~ ;"and D{x } # D{xP} # D{yU} . For example, _
~ 'n~ ~ ~ ~n ~a)~ COS'~n {Jo ~  b~ (cos I'Ln + Pn ~a)~) - J2 ~ l b~ ~g+n (sln T pn (�)~)Y,
(9nj = Co9 `{'n lJo (  L) (sin I~n + Pn + Js ~ i -f b~ 1f;'yn (co: + P. (�)])b
(XI) 7:- (a)) COS +n SIO C (2kr,7.) I(COS I~Ln 'i 9n - ig +n ,Sin (`~a f' Pn (a)])},
70 ~ O, l a, -
- (4)
~ where J(x) is a Bessel function; b= 4k2r2 - d2 ; and sin C= sin
n x
- Gin x/x.m
From this it is obvious that the cui-vature of the front of the irradiated
wave # 0) results in intensification of differences between like char-
acteris~ics of the signal's quadrature components. In addition, it follows
Erom (4) that with wide averaging intervals the errar in measuring statis-
tical characteristics of a Signal's quadrature comp9nents increases with a
reduction in the value of 4k2rn - dz . Therefore, with krn = const and
a uniform distributio*t of c~ (d = 0) , the error in tveasuring these charac- -
teristics is minimal as compared wfth other distribution laws for a random
angle of er_posure (d > 1). With reduction of the averaging interyal to
yo : O.ln , the differences between characteristics of quadrature compon-
ents are intensified and there is an increase (of approxitnately an order
of magnitude) in:the error in measuring these characteristics, caused by
curvature of the front of the irradiating wave. But, if the distances
_ between reflectors (in wavelengths) are greatiar than the inverse value of
the averaging interval, ~
_ 83 -
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
t d�
[Or rn1 X j 470 co` (ao ~n) + n Sin (ao Rn) ~ Qe 4n ~
-
- the stat13tical characteristics of the signal's quadrature components prac-
- eically do not depend on the radius of the front of the irradiating wave.
Reasoning in a similar manner, it is possible ta demonstrate that the in- ,
tercorrelation coefficient between components, R{x0 ,y0} , generally is not -
equal to zero and also depends on the radius of the front of the irradiating
wave.
Taking the above into account, we establish that generally the statistical
structure of the signal scattered by a combination of interconnected reflec-
tors in exposure to a spherical wave is described by a generalized probabil-
istic model of the envelope and phase of random signals [2,4]. For practi-
cal calculations it is more convenient to use another type of approximation
of the real distribution of the amplitude of random signals--the Nakagami
d is tribution: *
m
�
lv(A 2 (A"-')c.' J- 2 A'
, -
a~ r (m) p m
4 Qn
' (5)
where A is the amplitude of the signal, a E is the effective scattering
surface (EPR) of the object and r(m) is a gamma function.
Utilizing the results of [2-4], it is not difficult to establish the rela-
tionship between the dis6 ributionOparameters in (5) and characteristics of
quadrature components x and y and at the same time to determine their
dependence on the geometrical characteristics of the system of reflectors
and the radius of the front of the irradiating wave.
By employing statistical methods, it is possible to estimate also the error ;
in measuring instantaneous values of the characteristics of a signal scat-
tered by a complex object when exposed to a spherical wave [5]. _
Conclusions
1. The statistical structure of a signal scattered by a cpmbination of
interconnected reflectors when exposed to a spherical wave is described by
- a generalized probabilistic mode7, of the envelope (amplitude) and phase of
random signals.
*Probability density W(cris determined in (5) with an accuracy of a
constant. -
_ 84 ,
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICr
I.AL U5E ONLY
2. Curvature of the front of the irradiating wave results in the intensi-
fication of differences between like characteristics of quadrature cosn-
ponents of the scattered signal.
3. With an increase in distances between reflectors (in wavelengths) to
r /a � (1 + d) , YO _ n and rn/J1 ti [1/4YQ cos (a0 + ~n)] , Y0 < O.ln [
(c~ and a are the distribution parameters of the random angle), the
statistica4 RLKh's of a complex scatterer are determtned only by the pro-
perties of elementary reflectors and practically do not depend on the rad- -
ius of the front of the irradiating wave. �
Bibliography
1. Mayzel's, Ye.N. and Torgovanov, V.A. "Izmereniye kharakteristik ras-
seyaniya radiolokatsionnykh tseley" [Measurement of the Scattering
Characteristics of Radar Targets], Moscow, Sovetskoye Radio, 1972.
2. Pozdnyak, S.I. and Melititskiy, V.A. "Vvedeniye v statisticheskuyu
teoriyu polyarizatsii radiovoln" [Introduction to the Statistical
Theory of the Polarizatian of Radio Waves], Moscow, Sovetskoye Radio,
1974.
3. Nakagami, M. "lhe m-Distribution--a General Formula of Intensity Dis-
tribution of Rapid Fadings" in "Statistical Methods in Radio Wave Pro-
pagation," Fnrgamon Press, 1960.
4. Iieckmann, P. and Spizzichino, A. "The Scattering of Electromagnetic
Waves from Rough Surfaces," Pergamon Press, Oxford-London-New York-
Paris, 1963.
5. Kuznetsov, Yu.A., Melititskiy, V.A. and Shlyakhin, V.M. RADIOTEKHNIKA,
" Vol 33, No 12, 1978.
COPYRIGHT: RADIOTEKHNIKA, 1979
[59-8831),
8831
CSO: 1860
85
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000200060038-2

APPROVED FOR RELEASE: 2007/02108: CIA-RDP82-00850R000200060038-2
FOR OFFICIAL USE ONLY
UDC 621.396.96:621.391.26
SYNTHESIS OF A SINGLE-PULSE DISCRIMINATOR FOR A LOCATED TARGET
- Moscow RADIOTEKHNIKA I ELEKTRONIKA in Russian Vol 24 No 4, 1979 manuscript
received 29 Mar 77 pp 852-854 . ,
[Article by N. Ya. Kuz']
[Text] Various methods can be used to increase the angular resolution of
a monopulse radar [1, 2]. One of them is based on the use of angular
strobing which pernu ts the corresponding coarsening or complete shutdown
of the tracking system automatically with the presence of several targets
in the resolved volume or when there is flickering noise.
Z'he data on the composition (nature) of a located target (single or group),
_ required to control the strobing device, can be obtained by different
methods. For example, a method based on creation of an energy contrast of
targets by using two transmitting and two receiving devices, is considered '
in [4]. Besides the complexity, a significant disadvantage of the given
method is the low efficiency when operating, for example, in the passive
mode by radiation sources. In this regard it is of interest to determine ;
- methods of distinguishing the composition of the located target which fol- '
low directly from statistical decision theory.
Let us limit ourselves to consideration of the two-dimensional case, assum-
ing that signal reception is accomplished by two antennas, the radiation
patterns F1(4() and F2(0() of which are shown in Figure 1. Let us assume
that signal A' (t) received from angular direction o( 1, read from an
equisignal, corresponds to hypothesis H1 and that mutually uncorrelated
signals t 1(t) and cf 2(t,) received from directions a(1 - ao( and a 1+ '
+8x , respectively, cori~espond to hypothesis H2.
Bearing in mind the relatively small values ~f