close

Вход

Забыли?

вход по аккаунту

?

Microwave propagation characteristics under subtropical environment

код для вставкиСкачать
M IC R O W A V E P R O P A G A T IO N C H A R A C T E R IS T IC S
U N D E R S U B T R O P IC A L E N V IR O N M E N T
A T H E S IS
S U B M P T E f TO T H E F A C U L T Y O F S C IE N C E
G A U H A T I U N IV E R S IT Y
F O R T H E A W A R D OF T H E D E G R E E OF
D O C T O R O F P H IL O S O P H Y
IN P H Y S IC S
BY
S U K L A SEN
D E P A R T M E N T O F P H Y S IC S
GAUHATI
U N IV E R S IT Y
1999
M ICROW AVE PR O PAG ATIO N C H A R A C T E R IS T IC S
U N D E R S U B TR O P IC A L E N V IR O N M E N T
A TH E S IS
SUBMITTED TO THE FACULTY OF SCIENCE
GAUHATI UNIVERSITY
FOR THE AWARD OF THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN PHYSICS
BY
S U K L A SEN
DEPARTMENT OF PHYSICS
GAUHATI UNIVERSITY
1999
ProQuest Number: 10104589
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
ProQuest 10104589
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
C ertificate
jA/s is>to* (untfy tA-at/ tAts^ tAes/s is tAe x£S4t£i f teseatcA axotA (/one Ay
'h/t/, :>4uAAu.twiAex, my yai(£ance amA snjaetnision. fAfA/s tAesis, as a
wAoAe iwyjwui/; tAeieyf, isn/siyiuAAisAsxAox,S((Amit.texAto any otAer QAv/oexsityfox
any (fey tee. ^JfA/s- tAesibfuffA tAe* xsyuixsmen/sfmv tAe am wufftAey (feytee o f
fjo ctoi o f APAifoSoyiAy Ay tAe ^ ffaaAati Qfntto&tsity.
O f'tAewfne- Aanes tAes^fifeasate f x*>e.ommen<finy Aets tootA fo x tAe amax(f o f
tAis (feyxee.
Dated S P f - il- ‘ IS
P la c e : Gauhati University
/K Ji~
<D £ 02 '
(Minakshl Devi)
Professor,
Departm ent of Physics
G auhati University
ACKNOWLEDGEMENT
The author wishes to express her deep sense of gratitude to her guide
Prof. M Devi of Gauhati University without whose constant guidance and
unstinted support this thesis would not have materialised. The author is also
deeply grateful to Prof. A. K. Barbara (Professor Emeritus of Gauhati University)
for his theoretical and technical advice which had provided continuous en­
couragement during the tenure of this research work.
The author is also indebted to Prof. J. Das of I.S.I., Calcutta, for his invalu­
able discussion and suggestion leading to enrichment of this work.
A special thanks to Prof. S. Jois of Gauhati University and Mr. J. Bora of
IASST, Guwahati, for their help in development of computerised model which
has ensured that this research work stays in an even keel.
The author also wishes to acknowledge the critical co-operation ex­
tended by the Deptt. of P & T and N. F, Railways in allowing her to provide their
microwave links for her data which encompass this thesis. The author also re­
mains obliged to the D.O.E., Govt, of India, for rendering the vital financial
support for this work. An earnest thanks would also go to the Director, IMD for
providing the useful meteorological data.
In addition to the above, the author also acknowledges the assistance
and constructive criticism rendered by Dr. Sanjay Sharma, Dr. K. L Timothy, Mr.
Malay Kr. Barman and Mr. Deep Kamal Gohain.
She would also like to acknowledge the ceaseless motivation received
from Dr. Amalendu Sen, retd. HOD of AEI. The author also would like to appre­
ciate the moral support provided by her parents, brolher, parents-in-law, her
children Sayon and Sreya and other family members. Lastly, she would like to
extend her thanks to her husband Mr. Sanjib DuTtagupta, without whose con­
stant encouragement, the completion of this thesis would have been a dream.
ABSTRACT
The LOS microwave communication is still in existence in many parts of the
world. In a well designed link, the radio waves transmitted from the transmitter
reaches the receiver with no deterioration in signal quality, but sometimes it is
seen that the radio signal while passing through the medium gets degraded
and may at times be lost. As the reception quality of the radio signal is control­
led by the environment through which the radio signal propagates, it has been
an interesting topic of research since the first installation of LOS link. Though,
with the advent of optical communications this mode of propagation is be­
lieved to have lost its importance, the operation of LOS links are still very much
in existence in this country. So a look on the operational reliability anc predict­
ability of the microwave hops is worth noting. Degradation of such signal is
always measured in terms of attenuation and fading a, .a' prediction reliability
of such signal is very important, and as a result a number of models with these
aims have been developed, But even y lth all these models and volume of
information, there is no single model wKo can provide a flawless ”-'k prediction
at different terrain and environmental situations. Such studies also help in un­
derstanding the basic physics of the system and more so the coupling proc­
esses between earth and the near earth environment as the North Eastern
region of India has varied tropospheric and topological characters, a single
mode! will not be sufficient to study and analyse the attenuation or enhance­
ment of microwave signal energy at the receiving end. Very little theoretical
and experimental work has been carried out over this region, as a result the
necessity arises to develop models to improve the quality of signal reception.
The alms of this work Is based on these facts.
For this purpose, links under study are so selected, they have almost same
hop length, same topology but located at different situations.
1.
Milmilia-Durgasarovar link (frequency 6 GHz and hop length 41.2 km)
2,
L a o p a n i-H a b a lp u r link (fre q u e n c y 7 G H z a n d h o p length 55,8 km),
The thesis consists o f five chapters. A description o f th e ch a p te rs a re sum­
m arised as u n d e r :
C h a p te r I : This c h a p te r gives a ge n e ra l Introduction on th e role o f tro p o sp h e ric
environm ent a n d th e underlying terrain on th e p ro p a g a tio n o f m icro w a ve sig­
nal. A brief revie w o f th e work rele va n t to th e th e m e o f th e thesis Is also pre­
sented.
C h a p te r II : in this c h a p te r terrain profiles o f M ilm ilia-D urgasarovar link a n d
L a o p a n l-H a b a ip u r link un d er different atm ospheric c onditions (for k values 4/3
a n d 2/3) are presented. The necessa ry experim ental set u p for re ce ivin g th e
radio w a v e signal Is g ive n in this ch a p te r. The d e v e lo p m e n t o f instruments (fast
response th e rm o g ra p h a n d SODAR) for p ro b in g th e atm osphere is also dis­
cussed here.
C h a p te r H I: The fa d in g s o f radio signal w hile p ro p a g a tin g th ro u g h th e m edium
are a n a lyse d for b o th th e links a n d studied in details. It is seen th a t th e fa d in g
characteristics o f th e tw o links are d ifferent from e a c h other. W hile Miimilia un­
d e rg o e s m ore fa d in g in winter, L ao p a n i e xp e rie n ce s th e sam e in summer. Also,
fast fa d e s are m ore co m m o n In La op a n i w h e re a s it Is slow a n d d e e p in c a s e o f
Miimilia link. These d ifferences In th e fa d in g patterns a re c o rre la te d with m e te ­
oro lo gica l param eters a n d are p re se nte d in c h a p te r HI. For this pu rp ose th e
d a ta c o lle c te d b y radio so n d e , g ro u n d b a s e d te m p e ra tu re , pressure a n d hu­
m idity a n d satellite d a ta h a v e b e e n used.
C h a p te r I V : A n em pirical m o d e l is d e v e lo p e d in this c h a p te r associating fa d e
depths with different atm ospheric conditions u n d e r isotropic a n d anisotropic
situations. The m o d e l o u tp u t Is fe d into a n o th e r link to assess th e e ffic ie n c y a n d
reliability of the model, This chapter discusses details of framing of this model,
Chapter V ; In this chapter ray tracing Is adopted In designing of a model which
gives the angle-of-arrival a signal beam when encounters and Irregular me­
dium. This model enables one to realise the amount of deviation of signal path
at different atmospheric situations, for different path lengths and looking an­
gles. The effects of small scale irregularities are also discussed here.
u©
INDEX
CONTENTS
CHAPTER i
PAGE No.
PROPAGATION OF MICROWAVE IN THE
TROPOSPHERIC MEDIUM
A.
1
1.1
Background
I
1.2
Techniques for monitoring atmospheric
S
variabilities
CHAPTER li
1.3
Role of troposphere on wave propagation
7
1.4
Aims of the thesis
9
1.5
Conclusion
10
TERRAIN FEATURES, EXPERIMENTAL ARRANGEMENTS
12
AND CIRCUIT DEVELOPMENT
2.1
Introduction
12
2.2
Terrain features of the links
12
2.2.1'
MIImllla-Durgasarovar link
14
2.2.2
Laopanl-Habalpur link
14
2.3
Microwave link Information
14
2.4
Defection, measurement and calibration
18
of fadings of microwave signal
2.4.1
2.5
Fade measurement
Development of system and circuits
18
18
for measurement of associated
troposphere parameters
2.5.1
Introduction
18
2.5.2
Electronic dry and wet temperature
21
recorder
2.5.3
SODAR
24
CHAPTER HI
FADE CHARACTERISTICS ; SEASONAL DJJRINAL PATTERN
27
3.1
27
3.2
Introduction
Seasonal variation pattern between the
29
two link : Their characteristic differences
3.2.1
3.3
CHAPTER IV
Analysis
29
Classification of fades
34
3.3.1
Fade depth
34
3.3.2
Fade rate
39
3.4
Discussion
3.5
Conclusion
,
PREDICTION OF FADE THROUGH METEOROLOGICAL
48
55
58
PARAMETERS: MODEL COMPUTATION
4.1
Introduction
58
4.2
Analysis of fade data
59
4.3
Correlation of RR1 grad, with fade depth
60
4.3.1
Casel
60
4.3.2
Case II
67
4.3.3
Case III
72
4.3.4
Case IV
72
4.3.5
Case V
74
4.4
CHAPTER V
Discussion
74
PREDICTION OF FADE THROUGH RAY TRACING MODEL
81
5.1
Introduction
81
5.2
Techniques adopted
82
5.2.1
Description of the ray tracing model
83
5.2.2
Introduction of multipath parameter as
92
another input to the above model
5.3
REFERENCES
Discussion
94
99
CHAPTER I
PROPAGATION OF MICROWAVE IN THE
TROPOSPHERIC MEDIUM
The aim o f this c h a p te r is to present a general review o f th e work d o n e relevant
to the th e m e o f th e thesis w hich deals w ith prediction o f fades d u e to clea r air
effects through analysis o f tem poral variations o f these param eters (a nd asso­
c ia te d variables) over similar terrain but w ith d iffe ren t North-eastern atm os­
pheric situations o f India. A brief ba ckg ro u n d associated w ith this problem is
also described.
1.1 BACKGROUND
It was first p re d icte d by Hertz (1864), th a t e le ctric oscillations in a circuit pro­
duces e le c tro m a g n e tic w aves In the vicinity o f th e oscillator. M uch later, these
waves w ere actu ally p ro d u ce d -b y Hertz, as well as by Sir J. C. Bose. M arconi
(1901) was the first scientist to send such w aves across th e A tla n tic a n d th e
factors responsible for propagations o f these radio w aves soon b e c a m e a centre
o f research. It was proposed th a t th e basic mechanisms responsible for th e
propagation o f M arconi's trans-Atlantic radio transmission was through th e iono­
sphere (Kenne I1y a n d Heaveslde 1902) a n d it b e c a m e e vid e n t th a t th e upper
atmosphere consists of ionised layers w hich acts as reflectors for the radio waves.
The ionosphere then started draw ing the attentio n of scientific com m unity a n d
works h a d be e n d o n e within a d e c a d e o f their discovery. A volum e o f reports
on ionospheric studies with th e d a ta c o lle c te d over d iffe ren t stations o f the
globe have be en re ceived (Ellyett 1947, D iem inger 1952,55, Fejer 1955, Hirono
a n d M e a d a 1955, A p p le to n e t a t 1955, G regory 1956,Peino 1956, Bowles 1958,
Fejer a n d Vice 1959, Baynon a n d Brown 1959, Bowles 1961, Barrington a n d Throne
1
1962, Nelms 1963)
In India, pioneering work on Ionosphere w a s started by Prof, S. K. Mitra in 1926
at C alcutta University immediately after the discovery of Ionosphere. This w a s
followed b y m a n y workers (Rastogl & Rajaram 1971 Tyagl et al 1977, Devi and
Barbara 1978, Chandra et al 1981, Aarons an d Dasgupta 1984, Balan & Rao
1984) w h o m a d e significant contributions In the field.
ft w as quickly seen that w a v e s of frequency b e y o n d 30 M Hz w ere not reflected
by the ionosphere w hich b e h a v e s like a transparent m edium to frequencies
a b o v e a b o u t 30 MHz.
The com m unication system then took a revolutionary turn with the introduc­
tion of line-of-sight (LOS) m icrow ave (mw) link. The first LOS m w link w a s tried
successfully betw een D over a n d C alais at a h o p length of 40 km. The radio
w a v e w hich is like the light w a v e s in all respect u n d e rg o e s refraction. C h a n g e s
in the refractive index of the earth's atm osphere o ccu r with height or with other
m eteorological causes. A s a result the environm ent close to the e a rth 's sur­
fa c e (troposphere) has a key role to p lay in such propagation. Therefore, with
the introduction of LOS com m unication, it has b e c o m e a necessity to under­
stand the role of troposphere on radio w a v e p ro p ag atio n through near sur­
fa c e environment.
A s Is well known, troposphere Is a region of e a rth 's atm osphere from surface of
the earth to a b o u t 10 km a b o v e It at the poles a n d to 16 km a b o v e at the
equator. In this region the tem perature falls at the rate of 6.5°C/km a n d at the
upper b o u n d a ry the tem perature Is -5°C. A b o v e the troposphere is the strato­
sphere region w here the tem perature remains constant at -50°C.
The water va p o u r content in this region is mainly d u e to evaporation of water
from seas, o ce a n s, rivers etc. Hence, the troposphere over the o c e a n is moist
w hereas tiie moisture content of the air over the desert region is very low. A n a lo ­
go u s to the variation of temperature, w ater v a p o u r content too d ecreases
with increasing height within the troposphere. A t a b o u t 1.5 km from the sur­
face, the water v a p o u r content is half of w hat it is at the surface a n d at the
2
upper boundary of the troposphere, it is few thousandth of the water vapour
content at the surface.
The important parameter in this study of tropospheric radio propagation is the
refractive index. Therefore, we define a quantity as refractlvity N given by
N = (n-1) x 106, - ......... 1.1
where n is the refractive Index.
The radio refractlvity or radio refractive index (RRI) N, depends on
Temperature T,
pressure P,
and water vapour pressure 0 a s
N=77.7 P/T + 3.73 x 10s e/T2-------- 1.2
(Smith a n d Welntraub 1963 an d Bean a n d Dutton J968) where,
T is In °K
e Is in mhar
P Is In mbar
Because of the absence of any water vapour content Of air the first term of
expression (1,2) is known as the dry component and the latter term as the wet
component of the atmospheric radio refractive index. Fig. '1.1 shows the varia­
tion of RRI of alf as q function of temperature and relative humidity.
The fluctuation in the RRI or Its variation with heightgives; another important
parameter for understanding fades in LOS link. This can, bb defined as
;,
r = 6370 (1-0.04665 exp (0.0057 Ns)):i km — - 1 - 1.3
It depends on the refractive index. For standard; atmosphere the effective
earth's radius is 4/3 times of the radius of earth which is 6370 km. The effective
* ■’
>
"
1i
,,
earth radius can go upto 9,000 or to 10,000 km in certain regions and in certain :
times, it is a function of the gradient of refractivity at and hear the surface. In ,
1933, Burrows, Shelling and Ferrel introduced a simple method, where radius of
3
X
‘5
J
100
2 0 0
300
400
500
0
4
-40
Fig 1
-30
-20
0
10
20
30
The R.RX of air as a function of temperature
and relative humidity
Temperature in °C
-10
40
curvature of the radio ray r, relative to the radius of earth rQmay be expressed
In terms of refractivlty gradient as
r/r0 ■ k a (l+CdN /dhJ/lS?)-1-------------1.4
where k is normally referred as the effective earth radius factor and dN/dh is
refractiviiy gradient in an unit per km. Fig. 1,2 shows the variation in k with dN/
dh as given by the equation 1.4, it is seen that k increases when dN/dh de­
creases from -200 N unit/km to -157 N unit/km. it is aiso seen that k becomes
large for dN/dh varying between -157 to 200 N units/km. K takes negative val­
ues for further increase of dN/dh beyond -200 N units per km. The effects of this
variation can be seen in fig. 1,3 (Stephansen 1981) in a particular terrain condi­
tion. This clearly shows that when effective earth radius factor varies from nega­
tive to positive values at different dN/dh conditions, it may obstruct the propa­
gation path in some situations or extend the radio horizon in some other situa­
tions. Hence when the gradient dN/dh is > -40 N/km will occur then such a
condition is termed as sub refractive condition. Here K becomes less than 1,
causing the radio wave to bend towards the normal. This produces increase in
the curvature of the earth giving rise to reduced ranges of radio wave propa­
gation.
1.2 TECHNIQUES FOR MONITORING ATMOSPHERIC VARIABILITIES
The parameters of the RRI can be determined by measuring the temperature,
pressure and humidity atthe surface and sensors installed at two or three heights
of a tall tower. Radio sonde observations give the values of RRI to a height
upto 16 km.
Irregularity structure m easurem ents:
The fluctuation of RRI is a good index of presenceof a tabulating medium and
, by measuring the variation of RRI with time (rapidTluctuations), the Irregularities
in the medium can be assessed. A signal passing through such a medium may
5
8
dN/dh
Fig 1.2 :
dN/dh K
314
157
0.33
0.5
°-157
L°
Nunits/km
Effective earth radius factor as a function
of Refractivity gradient dN/dh
K= 11
*
Fig 1.3
6
not arrive at the receiving antenna with the desired angle
(Crawford and
Sharpless 1946, Pari 1983, Webster 1987) or there may be multiple reflections
causing multipath propagation (DeLange 1952, Crawford and Jakes 1952,
Meadoweta! 1966)from these tabulating structures which act as scatterers or
reflectors (Inou andAklyama 1966, Benardinl 1977). One of the important pa­
rameter for defining these Irregularities Is Cn2 , Cn2 parameter (Ishlmaru 1978,
Reddy 1987) can be calculated by using relation
o n 2s 0.31 Cn2 k7/6 L1,/4-------- J.5
where c n2is the variance of the logrlthm of the Intensity Is given as
a n 2 = ln ( l+ S 42) ............1.6
This value can be calculated from fluctuations in the amplitude of the signal
received over a LOS path. S4 is the scintillation index. Such irregularities struc­
tures can also be received from remote probing of the atmosphere by SODAR
(Barbaraet al 1994, D. Narayan Rao et at 1992, Trlpathi et al 1992) upto about
1 km. The back scattered signal from SODAR will also be used for associating
atmosphere situations with fades over Milmilia path.
1.3 ROLE OF TROPOSPHERE ON WAVE PROPAGATION
The radio wave propagation which is largely affected by the meteorological
(Builington 1965, Dougherty 1968, Wait
1962, Livingstone 1970, Tatarski 1971, Crane 1981) while passing through the
parameters may attim e change rapidly
troposphere, radio waves may undergo the following effects.
a)
absorption by atmospheric gases
b)
attenuation and also scattering by hyderometeors
c)
refractive effect (changing direction, fluctuation in angle-of-arrival,
.
focussing and defocussing of the signal
d)
tropospheric scatter due to turbulence in the medium
• i
The effects of hyderometeors, viz., rain, fog, snow, hail etc. on the microwave
7
propagation are not serious at frequencies 6-7 GHz (in which present study is
made), but these factors have serious affects at frequencies above 10 GHz,
Another important factor that governs the microwave propagation is the ter­
rain features over which the radio wave passes, Such affects are reported right
from the first establishment of the LOS link about five decades back. (Epstein
and Peterson 1953, Bulilngton 1957, Millington 1962, Deygout 1966,Ttwarl 1990.
Levy 1990) Ho e ta i 1980observed that microwave propagation is affected by
the underlying terrain path, Large fluctuations are resulted due to the output of
an air conditioning plant near the receiving station. Sengupta and Dasgupta
1983 studied the effect of urban climate on microwave propagation by com ­
paring tw o links having almost similar frequency and path length but over dif­
ferent environment. A similar work carried out by Hall 7979 in London who re­
ported that the major factor which affects the signal strength is the degree of
urbanisation, in and around the receiving station.
So far the terrain effect is concerned, one of the most vital parameter needs to
be known Is the Fresnel zone which gives an estimation of the clearance be­
tween antennas situated well awqy>from the surface of earth and a reflecting
or absorbing objects In between, The first Fresnel zone is defined as,the surface
of the ellipsoid o f revolution., with the transmitting and receiving antenna at
the focal points. The first Fresnel zone Is calculated, by the following e q ua tio n:
E * 17.3 ( d ^ / f (d,+d2))1/2— — — - 1.7
where f Is the, frequency In GHz, d l, d2 are distances In kms from transmitter
and receiver to a particular point where Fresnel zorie radius is to be evaluated.
Considering the importance of understanding the changes of LOS ray path at
different dN/dh conditions (as described in article! l.T of the chapter), there
have always been efforts to understand these changes of path from its desired
angle-of-arrival of signals while propagating through medium when RRI shows
i
;'
i
1
large variations from its normal lapse rate (Webster^ 1982, 1987, Crawford and
Sharpless 1946, Pari 1983). In fa ct the RRI fluctuations are a fade character and
such fluctuations in fades again help In realising the atmospheric irregularities
8
parameters. So attem pts h a v e b e e n m a d e to associate dN/dh fluctuations
a n d C n 2 derived from m icrow ave fa d e d a ta (Shen a n d Vilar 1995) resulted
from small scale irregularities in the atm osphere.
Num erous m ethods are available for prediction of attenuation of m w signal
with different atm ospheric conditions. Schiavone a n d Mermiller (1986)d esigned
a m odel through regression analysis b y w hich the probability of fa d in g o f a
p articular f a d e d e p th c a n b e m a d e o n e d a y In a d v a n c e . Singh a n d
Parthasarathy 1976 form ulated equations to correlate refractive index gradi­
ent with path loss of a 2 G Hz m w link. In addition to these ray tracing m odels
are also d e v e lo p e d to find a relation b etw een the attenuation of a signal a n d
atm ospheric variability (Pickering a n d Derosa 1979, Sasaki a n d Akiyam a 1979,
Webster 1982) a n d also the roughness of the terrain (Blsceglia 1988).
1.4 A IM S O F THE THESIS
Besides the introduction given In this chapter, the thesis com prises of four more
chapters. C h a p te r II presents in details the terrain profiles of the tw o ne w links
with different K values. The chapte r also discusses necessary experimental set­
ups m a d e for receiving the am plitude of m w signal a n d also the circuits devel­
o p e d for monitoring tropospheric parameters. The output of the instruments
are also presented In the chapter.
The tem poral variations of the tw o links hqyjng similar h o p lengths a n d te rra in
profiles but with different environm ents are discussed In details in chapte r HI
a n d the contribution of m eteorological param eters (like fog) for c h a n g in g the
seasonal pattern of the tw o links are highlighted. For this purpose beside ground
b a se d a n d radio so n d e data, satellite observations h a v e also b e e n used.
In chapter IV, empirical m odels are d e v e lo p e d associating the o b se rve d fa d e
depth with RRI grad, values. The m odel aims to predict fa d e d epth through RRI
grad, for both isotropic a n d anisotropic situations for sub- a n d super-refractive
conditions prevailing in the atm osphere. M od e l outputs are then used for pre­
dicting fa d e depths for another link op erated in a similar frequency range. The
9
chapter also discusses the reliability of the model so framed.
Chapter V discusses ray tracing method applied for com puting deviation of
angle-of-arrival of the beam at the receiver while passing through an irregular
medium of different dimensions. The corresponding fade depth experienced
at the receiver because of the defocussing effect is calculated for different
hop lengths, various looking angles and also for irregular layers of different di­
mensions and thickness. The effects of small dimensional scattering in a turbu­
lent medium is also com puted.
1.5 CONCLUSION
Microwave propagation characteristics is very much influenced by the me-,
dium in which It passes. The troposphere is governed by a factor which is de­
fined as radio refractive index. The gradient of RRI is responsible for bending
the mw signal. If the normal lapse rate of -40N/km is maintained (where N is the
refractivity) then the main beam will reach the antenna as designed, but In
case if there is a deviation from normal Idpse rate then the beam will be devi­
ated. The radio horizon Is expressed In terms of product of RRI .gradient and
radius of earth. This again will have an influence on the-Fresnel'ellipsoid. This
chapter deals in details with the above characteristics.
i
10
!
;•
-100 > to >
-157
< -157
Super
refraction
Ducting
a km
e
X
r**
00
CD
V
00
Actual path
1A 1
fN
w
o
***
v
^
ID
t
CO
O
»“*
X
LO
ID
O
V
(D
A\
Tabular form of atmospheric refrection
o
i
A\
- 40> to >
-100
<
CO
'c
3
z
Normal
refraction
e
Sub.
refraction
j*:
refraction
Forms of
CO
O
T-
Equivelent
path
////////
CO
O
------- -
z
\f
\f
I
CHAPTER II
TERRAIN FEATURES, EXPERIMENTAL ARRANGEMENT AND
CIRCUIT DEVELOPMENT
2.1 INTRODUCTION
The traject in the LOS path microwave signal propagation is influenced by the
terrains as well as the environment (Craig and Kennedy 1987, D. Narayan Rao
et ai 1991 1992, Sen et al 1991, Jha et ai 1992, Shen and Vilar 1995). Further
even the artificial constructions or man-made changes in the environment may
also affect the signal significantly (Ho etal 1980, Sengupta eta! 1983).
The aim of this thesis is to frame model to predict microwave fades (for line-ofsight links passing over long flat terrain) by using atmospheric variabilities. For
this purpose the fade data collected over Milmilia-Durgasarovar and LaopaniHabaipur will be utilized. These two links are selected because of their similar
flat long terrain which have different background environment. Milmilia link
while passes over wet marshy land, the Laopanl-Habaipur hop basically passes
over dry area.
2.2 TERRAIN FEATURES OF THE LINKS
Fig. 2.1 shows the geographical locations of the two microwave links that would
be considered here. The information regarding the terrain features have been
collected from P&T Department, Govt, of India, and the remote sensing de­
partment of Geograpical Survey of India. A short description of the terrain fea­
tures of the links are given below.
12
Fig. 2.1
?3
2.2.1. M ILM ILIA-DU RG ASARO VAR LINK (transmitting frequency 6 GHz)
Fig. 2.2a, b show s the path profile of Mllmilla-Durgasarovar link. The transmitting
station is situated at Milmilia while the receiver is at Durgasarovar. The path
length of this link Is 41.2 km. The underlying terrain Is basically m arshy with tw o
major sw am p s (D eepor Beel a n d Kukurmara Beel) present in the path. M o re o ­
ver, the river Brahmaputra flows tow ards its north a n d river Kuisi tow ards its south.
There are build up iands of 3 km (about 15 km from Milmilia) a n d of 8 km (al­
most near Durgasarovar) in the hop. Rest of the link is c o v e re d b y agricultural
field a n d forest (Milmilia reserve forest). The reserve forest is a n evergreen o n e
a n d is p op u la te d with Sol, Ajhar, Sida trees. The Fresnel zone cle aran ce for
different "K " values are then determined. It is seen that for both norm al and
worse conditions, the Fresnel zone cle aran ce is sufficient.
2.2.2 LAOPANI-HABAIPUR LINK (frequency 7 GHz)
This link falls in the N a g o a n district of A ssam a n d lies east of Guw ahati. The
terrain features of this link are mostly flat a n d plain a n d like Milmiiia terrain, the
Laopani-H abalpur terrain Is also c o v e re d b y agricultural field a n d forest. But
here, this h o p (55 km) Is passing mainly over dry a re a e xce p t for a small zone
w here river "Kopill" crosses It. The Fresnel ellipsoids over this link Is then c a lc u ­
lated a n d it is seen that th o u gh for K=4/3 situation the Fresnel zo n e cle aran ce
is sufficient (fig. 2.3a), the ea rth 's b ulge obstructs the m ain b e a m at K=2/3
condition (fig. 2.3b),
2.3 M IC R O W A V E LINK INFORMATION
The Miimiiia-Durgasarovar a n d Laopani-H abaipur m icrow ave link are installed
a n d m aintained b y P&T a n d Indian Railways respectively. The various link p a ­
rameters of the a b o v e tw o links are presented in table 2.1.
14
&a * '
fig 2.2 (a)
Fig. 2.2 {a) Path profile of IVlilmilia - Durgasarovar
link for K(b). Path profile of IVlilmilia - Durgasarovar
link for K- 2/3
is
K ” 4/3
3^
\*
Vt V *
-
fig 2.3 (a)
,
,
<
\
Fig 2.3 (a) Path profile of Laopani -Habaipur link
:
;
.
:
i
forK- 4 / 3 '
»
I
_
,"
i
,
! :■p . ‘.1 ■
|
. :
;
/
,
, {b) Path profile of Laopani-;Hab:aip,iir link
16
i;
t
i
5 5 .8
4 1 .2
6 .4
50
Distance (km)
Frequency (GHz)
Transm itting antenna
hight (m)
17
3 5 .5
54
91
4 3 .2
105
Receved antenna
gain (db)
Transm itter
HASL (m)
Recever
HASL (m)
Table : 2.1
40 db
Power w . r. t.
pv (signal level)
00
30dbm
3 9 .5
4 3 .3
Transm itting antenna
gain (db)
Receved antenna
hight (m)
80
80
7 .1 3
LAOPANI
MILMILIA
LINK _____ ^
INFORMATION
l
09
CM
CM
2.4 DETECTION, MEASUREMENT AND CALIBRATION OF FADINGS OF M ICROWAVE
SIGNALS
2.4.1 Fad e measurement
Schem atic diagram for recording fa d e ch aracter is shown in tig. 2.4. The set up
consists of IF, RF, A G C an d d etector stage along with differential amplifier an d
chart recorder an d d a ta logger.
To study the fa d e ch aracter of the a b o v e m entioned links, the A G C output of
the receivers are fed to the recorders through a differential amplifier or vo ltag e
to current converter system dep en d ing on the typ e of recorder. The normal
signal-level is p la ce d a t zero volts. Fluctuations over this signal level are m eas­
ured with the help of calibration chart.
The an alog output from d etecto r is fed to an AD card of 8 single e n d e d a n a ­
log input channels range -5v to +5v, 12 bit resolution a n d 25 m icrosecond co n ­
version speed with 0.915% a c c u ra c y an d is processed through com puter after
appropriate filter for noise suppression to receive fa d e d a ta (fig. 2.5).
The technique applied for calibration is by introducing a signal generator of 70
MHz a n d an attenuator p a d b e tw e e n the IF a n d RF section. The correspond­
ing output is amplified a n d goes, to the recorder. The dynam ical calibration is
done time to time.
2.5 DEVELOPMENT OF SYSTEM AND CIRCUITS FOR MEASUREMENT O F ASSOC1J
ATED TROPOSPHERE PARAMETERS
2.5.1 INTRODUCTION
To realise the e ffe ct of the medium on m icrow ave propagation the following
tropospheric parameters h a v e b een m easured ;
(a ) Ground b ased temperature, humidity a n d pressure.
(b ) Dry an d w e t temperatures a t different heights by a tow er of 25 meters
height from the ground.
18
fig : 2.4
Fig 2.4 :
Block diagram of microwave station
receving setup
19
8
6
a.
<D CD
“O
"O
4
“O
as
2
18
75? W T3 Id
25
Id
31
M 31
3U
55
19
8
6
4
2
fig 2.5
Fig. 2.5 :
Computed output of microwave fade data
20
(c) Dry temperature, humidity a n d pressure (upto 2 km) with the help of radio­
sonde,
(d) Elevated structure, thermal plum es a n d fronts b y SODAR.
The d a ta o n tem perature a n d humidity h a v e b e e n collected with the help of
conventional therm ohygrograph. But sensitivity of the system is p oor a n d fluc­
tuations of temperature only upto 0.5°C could b e m easured with limited a c c u ­
racy. Moreover, the time resolution w hich is 15 minutes, is also not a c c e p ta b le
for a m eaningful study In relation to m icrow ave fadings w here a c h a n g e of 30
d b signal level is d e te cte d within 2 to 3 minutes of observation. To ove rcom e
this limitation, four sets of electronic dry a n d w et tem perature systems a lo n g
with recorder h a v e b e e n d eveloped.
2.5.2 ELECTRONIC DRY A N D WET TEMPERATURE R ECO RD ER
The conventional therm ohygrograph used for routine tem perature humidity
m easurem ent is found to b e not fast a n d sensitive e n o u g h to record the quick
c h a n g e s in tem perature a n d humidity. There are c a se s w h e n in a d e q u ate re­
sponse of the tem perature recorder results to loss of information. To ove rcom e
this limitations a fast response electronic dry a n d w et tem perature sensing sys­
tem is d esigned (Barbara et a t 1992). The block d iagram of the circuit is given
in fig. 2.6. The circuit Is b a se d on LM 335 with a v o lta g e regulator of LM 336 a n d
associated amplifier circuit. The output of the tem perature sensor (V = Rl) is fed
to a sum m ing amplifier to g e t a n amplified output of the tem perature control­
led signal. This output is fed to the chart recorder. The output vo ltage is cali­
brated to the corresponding temperature. For w et temperature m easurem ents
the sam e circuit is used.
The sam ple records of the dry tem perature variation is show n in fig 2.7.
From the tem perature records It c a n b e clearly seen that the tem perature fluc­
tuations up to 0.1 °C c a n b e resolved. This resolution is conve nie nt for the micro
structure study of the tem perature parameter.
21
, Fig {2.6 a) : Block diagram of a fast response temperature recorder
Fig (2.6b) : Temperature sensor with amplifer
22
1
Fig. 2.7 : Sample record of temperature measurement.
23
2.5.3 SODAR
Among the remote sensing techniques, of the atmosphere, the sonic detec­
tion and ranging techniques (SODAR) can very effectively be used for con­
tinuously monitoring the tropospheric conditions, in an effort to Identify the
cause of fadings in microwaves, tw o SODAR units are installed. One in the
midpath between Milmllla-Durgasarovar microwave link and the other at the
Gauhati University, (Barbara et at 1991). The block diagram of the device is
shown in fig 2.8. The system consists of following units.
1. Antenna
2. Transmitter
3. Receiver
4. Recorder and monitor
The backscattered signal from SODAR gives the information about ABL height,
plume condition and also presence of irregularities, if any, in the medium (say 1
km). Backscattered signal of SODAR also carry information of precipitation and
rain events, So analyses of SODAR echograms help in understanding the pre­
vailing atmospheric condition. A sample record of the sodar echogram is shown
in fig. 2.9.
24
GATE
N.
PRE A M P
**
CONTROL
UNIT
<
V
-------------------------------- >-------------- ------------
Fig 2 .% : Block diagram of S O D A R
25
26
C H A PTER 111
FADE CHARACTERISTICS SEASONAL AND DIURNAL PATTERN
3 J Introduction
A radio wave transmitted from a station is received at a distant location
through ionospheric or tropospheric mode of propagation. However, for a sig­
nal in the frequency range of 300 MHz and above, the preferred mode of trans­
mission is by troposphere and therefore, for LOS communication, the activities
of the troposphere is of extreme importance to understand the role of this
medium on communication links.
It has been well received by theoretical models and experimental obser­
vations (Sasaki & Akiyama 1977, Taftersall& Cartright 1977 Blomquiest & Norbury
1978, Stephansen & Mogensen 1979) that such waves do not follow a strictly
straight line path but follows the earth's curvature to some extent. The path of
the signal gets curved due to refraction and the curvature is controlled by the
refractive index of the medium. This refractive Index again varies with time and
meteorological condition. It is seen that for a refractive index gradient dN/dH
> - 40 N (where N Is the refractlvlty ), the value of the effective earth's radius
factor "K" becomes less than unity, causing the effective radio horizon to re­
duce. On the other hand for dN/dH < - 100 N, this effective line of sight gets
enhanced. In such cases, there may be degradation in the received signal.
Further, there are cases when the direct ray and the reflected ray from the
ground or from the elevated layers (formed due to abrupt variation of radio
refractive index gradient) may cause serious signal fluctuations because of
interference between these rays. Fading caused due to such variation in re­
fractive index of the atmosphere becomes significant with increase in transmit­
ted frequency, since shorter wavelengths have more scope for interference.
Multipath fading is also called selective fading because its occurrence de-
27
pends on relative signal phase, which Is a function of frequency.
A survey of literature indicates that study on the aspect of such fading
started as early as 1930 (Epstien 1930, Bremmer 1947). Experiments were almost
simultaneously conducted (Buillngton 1947) to examine the influence of differ­
ent types of terrain on propagation of radio waves. Even five decades back,
studies were made by Ames et at (1955) over water bodies and by Day and
Trolese (1955) over desert to examine the role of humidity on microwave propa­
gation. Such experimental observation and theoretical calculations show that
the prevailing meteorological conditions have serious influence bn centimetre
radio signal Saxton et al (1964) pointed out that the meteorological param­
eters may undergo rapid fluctuations within minutes, and also these variables
show clear diurnal and seasonal changes. After these observations, attempts
were made to couple diurnal and seasonal fading of microwave signal with
meteorological parameters (CCIR 1978,Stephansen 1981 Schiavone 1982, 1983,
Sharma 1993). However, it is realised that such associations are not always sim­
ple and there are cases when correlation between microwave fades and
meteorological parameters does not even exist (Mazumdar 1976, Schiavone
1982).
In this chapter a detailed analysis of fading over Miimiila and Laopani links
has been hnade to bring out the fade features over these tw o links and to
highlight the distinctive differences In the fade character so received, over the
two hops. The justification In selecting these two links lies in the fact that, though
the two hops are passing over similar flat terrain, their environmental features
are different. While one passes mainly over marshy land (Milmilia), the other lies
over a dry path. So these tw o links stand as good candidates to understand
the role of environmental features on propagation of microwave signal. The
meteorological parameters received through radio sonde, ground-based
measurements of pressure, temperature and humidity a t different locations,
28
and. fog measurement through satellite have also been used. From these analy­
ses, the basic parameters that are Involved for the generation of fading have
been recognised and the causative mechanism has been highlighted (with
different environmental features). For this purpose, three years of continuous
data taken over the two stations have been utilized,
3.2 Seasonal Variation Pattern Between The Two Links (Milmilia-Durqasarovar
and Laopani-Habaipur lin k ): Their Characteristic Difference
The main aims of this section are as under:
1. To highlight the differences between the seasonal variation pattern of the
two links, and
2, To identify the parameters and causative mechanism responsible for con­
trolling the seasonal fade characters (in the two links).
3.2.1 Analysis
For analysis of fade data, the average fade time (for all fades with depth greater
than 1 db) for each day is evaluated and the monthly mean of the occur­
rence probability is estimated. The seasonal percentage occurrence of fades
isthen calculated after defining seasons in four basic categories, as under:
Summer
- June, July, August
Post monsoon - September, October, November
Winter
- December, January, February
Pre monsoon - March, April, May
The occurrence percentage of fading in different seasons is shown in fig
3.1 for the two links.
29
P C. of occurrence of fading
Seasonal probability of fading of both the
links
S
Post
W
□ Milmilia
Pre
BLaopani
Seasons
Fig:3.1SEASONAL FADING PATTERN OF THE TWO
LINKS
30
The p oints to b e h ig h lig h te d here a re that, w hile for Miimilla link, th e fa d in g s
a re m o re c o m m o n in w inter m o n th s, L a o p a n i e x p e r ie n c e s m a x im u m f a d in g in
su m m e r a n d m in im u m f a d in g in winter. T he s e a s o n a l f a d e p a tte rn s o f t h e tw o
links a re th e re fo re c o m p le t e ly d ifferen t from e a c h other. The h o p le n g th a n d
b a s ic terrain fe a tu re s (flat) b e in g t h e s a m e for b o t h th e t w o links, t h e e n v iro n ­
m e n ta l e ffe c ts n e e d to b e e x a m in e d in this co n te xt. It Is a lr e a d y p o in t e d o u t
th a t M ilm ilia link p a s s e s o v e r a m a rsh y la n d a n d runs p arallel t o t h e m ig h ty river
B ra h m a p u tra , w hile th e o th e r link b a sic a lly p a s s e s o v e r a relatively d ry a r e a
e x c e p t for c ro ssin g tw ic e a sm all rivulet Kopili in its p a th . T he critical e x a m in a ­
tion o f m e t e o r o lo g ic a l p a ra m e te rs In d ic a te s th a t w h ile w e t f o g Is a sig n ific a n t
p h e n o m e n o n o v e r t h e Milmilia link (in s o m e o f th e se a so n s), t h e L a o p a n i link
p a sse s o v e r a relatively fo g ie ss e n viro n m e n t, it is th e re fo re w o rth lo o k in g for th e
role o f f o g (a s o n e o f th e p a ra m e te rs) o n f a d in g o v e r t h e M ilm ilia link. Further,
a s f o g is b a sic a lly a p r e - d a w n o r e a rly m o rn in g p h e n o m e n o n , it is d e c id e d to
a n a ly s e t h e f a d e d a t a for p re -d a w n , a n d e v e n in g p e rio d se p arate ly. S o for this
analysis, f a d e s s e e n d u rin g m o rn in g p e rio d c o v e r in g 4 a.m . to 7 a.m . h a v e
b e e n a n a ly s e d in d e ta il for e a c h d a y o f a m o n th . T he m o n th ly m e a n v a lu e o f
p ro b ab ility o f o c c u r r e n c e o f f a d e Is t h e n c a lc u la t e d a n d th e s e a s o n a l a v e r ­
a g e o f p ro b a b ility o f o c c u r r e n c e o f f a d e is t h e n e stim a te d from this m o n th ly
m e a n . Fig, 3.2 g iv e s th e o c c u r e n c e p e r c e n t a g e o f d a w n tim e f a d in g o n s e a ­
s o n a l b a sis o v e r Milm ilia a n d th e figure a lso cle a rly in d ic a te s a w in te r m a x i­
m u m in th e o c c u r r e n c e o f fa d e .
Foa and Fadings:
To e x p lo re th e m e c h a n is m for th e d e v e lo p m e n t o f h ig h e a rly m o rn in g f a d e s In
t h e w in ter se a so n , t h e first c h o ic e wiil th e re fo re b e t o a n a ly s e f o g d a t a o v e r this
link. For this p u r p o s e satellite d a t a , r e c e iv e d from th e 1MD, h a v e b e e n u s e d for
f o g inform ation o v e r t h e link a n d its p e r c e n t a g e o c c u r r e n c e in d ifferent m o n th s
a re p re s e n t e d in fig 3.3. The u n s h a d e d r e g io n o f t h e p ie c h a r t g iv e s p e rc e n t­
a g e o c c u r r e n c e o f f o g in a p a rticu la r m o n th a n d , a s e x p e c t e d , th e w inters (D,
31
P.C. of occurrence of fading
Seasonal probability of fading of Milmilia link
(morning hours only)
S
Post
W
Pre
Seasons
Fig:3.2 SEASONAL PROBABILITY OF MORNING
HOURS FADING OF MILMILIA
32
PRESENCE OF FOG IN THE MONTH
OF OCTOBER
PRESENCE OF FOG IN THE MONTH
OF NOVEMBER
6.1 (6.1%)
21.1
(21.1%)
52.6
(52.6%)
84.6
(84.8%)
i
PRESENCE OF FOG IN THE MONTH
OF DECEMBER
16.1
(18.1%)
!
|
(26-3%)
PRESENCE OF FOG IN THE MONTH
OF JANUARY
I
22.5
43(43%)
52.5(52.5%)
(41.5%)
j
PRESENCE OF FOG IN THE MONTH OF
FIG : 3.3 PROBABILITY OF OCCURRENCE OF FOG FOR DIFFERENT MONTHS
33
J & F) show relatively high fog and therefore, It Is likely that the early morning
winter and equinoxial fades over Mllmilia have fogs as their source of origin.
Now to examine this possibility It Is essential to eliminate the effect of fog, if any,
from the average fade occurrence picture of winter and equinoxial months.
With that aim, the fade data are thoroughly screened and fades with "ho fog
days" are only considered and their probability of occurrence is once again
calculated. The monthly mean morning average fade values (during "no fog"
days) so obtained is shown in fig 3.4. Unlike the winter maxima in fade occur­
rence, the Important point to be noted from this figure Is that the probability of
fade to occur reaches maximum in summer. It is now Important to bring to the
notice the similarity in seasonal fade occurrence probability of this link (after
eliminating fog contribution) with that of Laopani link (fig 3.5). However, in this
analysis, as fogs over Laopani are rare, the contribution of fog to fade data are
eliminated only over Miimllia. Nevertheless, the study indicates that the fog
generated over the marshy plain hop is responsible for generating winter high
fades and this acts as a source for changing the seasonal pattern from that of
Laopani,
3.3 Classification of Fades:
The fading character of a signal reflects the signature of Its source. Hence, in
order to explore the possible causes of generating fades over the two links, it is
essential to know the types of fade, and also to know the preferential type, if
any, for a particular link. Therefore it is necessary to analyse the fade data with
respect to fade depth and rate, In details.
3.3.1. Fade Depth
Figs 3.6 and 3.7 give the percentage of occurrence of fade depth over the
tw o links and it is seen that the fade depth over the Milmilia link are generally
high, if 20 db fade occurs for 0.3% of the time over Milmilia, the same depth of
34
P.C. of occurrence of fading
Sum
Post
Win
Pre
Fig :3.4 Percentage of occurrence of
fade excluding fog days.
35
P.C. of occurrence of fading
Sum
Post
W int
Pre
□ Milmilia link (morning average)
■ Laopani
Fig:3.5 Percentage of occu rence of fades
of the two links.
36
P.C. of time fade depth is ^ ABCiSSA
Fade depth in db
Fig. 3.6: Probability occurrence of fade depth In Miimilla link.
37
P.C. of time fade depth Is sgABCISSA
0,00
20,00
40,00
Fade depth in db
Fig. 3.7 : Probability occurrence of fade depth in Laopani link.
38
60.00
fade occurs 0.08% of the time in Laopani. However, as fade depth increases
beyond 20db, th© percentage of tlm© 1hat a particular fad© Is detected, ap­
proaches the same value In both the tw o links and ultimately both links experi­
ence link cut-off for the same period of tlm© (about 0.008% of time).
3.3.2. Fade Rate
Fade rate, that is, the number of fades/time ( fades/second are also evalu­
ated depending on the type of fade ) is then examined separately for each
season over the two paths taken for this study. This (fade/hour) classification is
made for analysing the cumulative distribution of fade rates, but for receiving
parameters like Cn2 the fade rate per second (in the fast recording mode) is
taken. The fadings are classified either as fast or as slow, depending on the
number of fades seen in a particular specified period. When more than 20
fades/ hour are seen, we consider it to be a fast fade. The cumulative distribu­
tion of fade rates has then been made and Fig 3.8 and 3.9 give such distribu­
tion for the two links. It Isseen that fast fades are relatively higher In Laopani link
compared to the Mllmllla link. After receiving an overall picture of fade depth
and rate for the tw o links, fadings are classified In the following groups:
1. Scintillation,
2. Slow and shallow,
3. D e e p , .
4. Fast fades superimposed on slow fades.
Brief descriptions on the occurrence of these types of fades are given b e lo w :
3.3.2.1. Scintillation type of fading
Scintillation type of fading (fig. 3.10 a, slow mode and fig 3.10 b in fast mode
recording) is characterised of those fadings which exhibits fast fade rate and
fluctuate over the mean level. The fade rate in this type of fading may vary
39
Fade rate (Fades/ Hour)
□
Sum
+
Post
o
W in
a
Pre
F IG . 3 .8 : C u m u la t iv e d istrib u tio n of f a d e ra te in M ilm ilia U n k
40
Fade rate (Fades/ Hour)
Fig, 3.9: Cumulative distribution of fade rate In Laopani link
41
Fig. 3.10a : Scintillation type of fading (slow mode)
42
Fig. 3.10b : Scintillation type of fading (fast mode)
43
from 20 fades/hour to 180 fades/hour, The fade depth of scintillating type of
fading Is not very high. Both the links experience this type of fading In summer
and the probability of occurrence of this type of fade is more in Laopani than
'in Milmllla. It Isseen from the cumulative distribution of fade rate over these two
links, that, if in Laopani 70 fades/hour exceeds 15% of the time In summer, the
Milmilia link experiences fade rate of the same amount. In the same season,
only for 0.3% of time. In general, the average fade rate of Laopani link is more
than that of Milmilia link.
3.3.2.2 Slow and Shallow
In this type of fading, both fade depth and rate are considerably low. The
average fade depth of such type of fading is around 4 to 5 db and rates are
varying between 5 to 6 fades/hour (fig 3.11). Unlike scintillation type of fading,
this slow and shallow fades are;seen in both the two ,links with almost equal
!
■ '
i
probabilities mainly in winter season. The most frequent,occurrence of this type
of fading is pre-midnight to early morning hours.,1
‘
:
j
3.3.2.3. Deep fades
This type of fade pattern is seen In both the links (flg.3.12), but it is interesting to
note that deep fading of magnitude (say 30 db), when occurs for 0.08% of the
time in Milmilia link, the same depth of fade can be seen only in 0.05% of the
time in Laopani link. However, the probability of receiving large fade depth of
40 db is same for both the two links. The important feature to be noted here is
that, while in Milmilia link such deep fades occur in The early morning hours of
the day, Laopani shows such fades in the midnight to post-morning hours.
3.3.2.4 Fast-Fades Super-imposed on Slow Ones
This is a rare type of fading where fast fades are superimposed on slowly vary­
ing signal. Fig 3.13 shows this type of fading detected over Milmilia link.
44
Fig. 3.11 : Slow and shallow type of fading.
45
Fig. 3 .1 2 : Deep typ e of fading.
46
Fig. 3.13 : Fast fades superimposed on slow fades.
47
The analyses of the above-mentioned fade types therefore reveal the follow­
ing behavioural differences or similarity between the tw o links:
1.
Fast fades occur both at Milmllia and Laopani links at daytime,
2.
Fast fades are more frequently seen in Laopani link than in the Milmilia Sink,
3.
Fade depth is high in Milmllia link in comparison to Laopani link, though
the probability of occurrence of deep fades, about 40 db, approaches
the same value in the two links,
4.
The occurrence of deep fades Is more during early morning hours in Milmilia
link, while In Laopani link it is more often seen during midnight or postmidnight hours.
The study therefore points to the fact that even for links with similar hop length
and identical terrain structures, the environmental features play an important
role in controlling the fade character.
3.4 DISCUSSION
The fading In both the links Is basically a nocturnal phenomenon. The fade
depth as well as its occurrence percentage is observed to be significantly high
over the tw o links during night-time though the preferential time of develop­
ment of fades may vary from pre-morning to early morning hours, Further, the
time of occurrence or development of a fade depends on the terrain and
environmental features. For example, while the fast fades are observed in day
time in both the links during summer season, the deep fades are detected
either in the night or in the early morning hours, Interestingly, it is seen that fast
fluctuations may develop after sunrise. Fast fluctuations In signal In Milmilia link
seen during early morning hours summer time are then examined in associa­
tion with atmospheric condition seen through SODAR operation in mid path of
this link, and these fluctuations occur even in the absence of plume structure in
the SODAR echogram. The absence of plume structure is due to poor mixing
condition in the atmosphere, because of improper heating of the ground dur48
ing the period of heavy monsoon rains over the study period. On examination
of RRI parameter during this time, It Is seen that dN/dH shows variations be­
tween - 40 N/km to -60 N/km (Fig. 3.14).
So it is likely that the improper condition inhibits both the formation of a
stable layer, as well as development of plume structure and in such a situation
one can only expect microwave signal to suffer scattering from small scale
dense irregularities, If developed In the mixing process (Sarkar et al 1983), The
development of fast fluctuations in the microwave signal leading to large fade
rate with relatively small depth during sunrise period can therefore be associ­
ated with the scattering signal from such irregularities.
Further, as large water vapour content during summer leads to super-re­
fractive condition, the contribution of such atmospheric situation to summer
fading cannot be ruled out. The presence of super-refraction, situation though
may cause serious fading leading to link cut off, the scenario Is rather complex
and depends on the thickness and height of such super-refractive layers with
respect to the antenna. Considering a situation as shown (fig 3.15) when a
strong super-refracting layer Is formed just below the antenna height a diffrac­
tion type of fading is expected because of interference between the direct
and the reflected ray. Such interference will result in fast fading. As super-refraction is more often detected in the low reaches of the atmosphere below
100 m, the terrain of Laopanl (transmitter antenna height 91m and receiver
antenna height 54m) favours such type of fading and the effects of such el­
evated layers for generation of fast fades over this link will be more serious in
Laopanl-Habaipur linkthan.in Milmilia-Durgasarovar link
It is also noted with' interest that the link cut-off over Laopani-Habaipur is
often associated with fast fading, Fig 3.16 shows a case where fast fades are
49
Fig: 3.14 Percentage of occurence of RRI gradient for
0000 and 1200 hrs GMT.
□ Seriesl
% OF OCCURENCE
fl Series2
RADIO REFRACTIVE INDEX GRADIENT (N/km)
50
Super-refracting
layer
Fig. 3.15
(a ): A super-refractive layer formed below the antenna heights,
(b ): N-h profile.
51
Fig. 3.16 : Link cut-off event of Laopani-Habaipur link.
52
seen before link cut-off over this link. It is worth mentioning that though the
atmosphere over this region is not so favourable for generating strong fades,
the link cut-off condition is however as bad as in Milmilia and is also seen more
in summer. This may probably be associated with K=2/3 situation, when the
earth's bulge obstructs the first Fresnel zone In Laopanl link. It is ofcourse true
that the sub-refraction events are far too low compared to super-refraction
events during summer, but our analysis shows that there are at least 10-15 % of
the cases when atmosphere generates the K = 2/3 condition over this link.
Because such situation in Laopani leads to black out the link cut-off is control­
led by
1.
K = 2/3 situation and also by
2.
Super-refracting layer near antenna height.
m a result the link cut-off of both the links becomes almost equal.
in Milmilia though the high fade depth in winter is associated with sub
refraction because of fog, there is hardly a case where winter shows link cut­
off, because, even In situation like K = 2/3, the first Fresnel zone clearance is
■' 11
high enough to obstruct the main beam. The association of super-refraction
condition with link cut-off during post-monsoon (and also few In summer) Is
well evident over the link of Milmilia, and unlike the Laopanl-Habaipur link, this
i
link does not show development of fast fluctuations In signal before black-out
(Fig, 3.17).
The detailed analysis of fade character;of the two major links studied here
therefore shows a clear dissimilarity In the seasonal pattern between the links,
and these are coupled with the environmental features associated with ter­
rain. The effects of terrain on micro-wave propagation studied by many work. ers, are also available ( Hay & Poaps 1959, Schias/one 1982, Tiwarl & Jassal
1989, Narayan Rao e ta l 1991 1992). Jhe environmental aspects on controlling
fade pattern have been dealt by Hay & Poaps 1959 and Schiavone 1983, where
they have shown that in stations Palmetto (USA) and O ttaw a (Canada), which
are remote.from the sea, maximum fadings.were detected on 6 GHz, 57 km
53
Fig. 3.17 : Link cut-off of Milmilia-Durgasarovar.
54
link during summer and minimum were detected during winter (fig 3.18, fig
3.19). These fade characteristics are analogous to our Laopani-Habaipur link
which behaves in similar pattern and is comparatively a low water vapour,
content region.
3.5 CONCLUSION
This chapter is thus concluded with the following points:
1. The fading pattern of Milmiiia - Durgasarovar and Laopani - Habaipur mi­
crowave links show distinctive difference, with respect to fade depth, rate
as well as seasonal variation.
2. In Milmiiia, fade depth is high while In Laopani fade rate is high.
3. The probability of fade occurrence Is more In winter In Milmiiia, whereas in
Laopani summer months show maximum fading.
4. The effects of fog on controlling fadings over Milmiiia are while significant,
sub-refraction generated cut-off and super-refracting layer near antenna
height control link cut-off over Laopani-Habaipur link.
nn
55
p.c. of annual fade time
Ottawa
Fig:3.18 SEASONAL FADE PATTERN OF
OTTAWA STATION
56
P C. of annual fade time
Palmetto
j
f
m
a
m
j
j
a
s
a
n
d
Fig:3.19 SEASONAL FADE PATTERN OF
PALMETTO STATION
57
C H A PTER IV
PREDICTION OF FADE THROUGH
METEOROLOGICAL PARAMETERS: MODEL COMPUTATION
i
4.1 INTRODUCTION
|
i
The variabilities of troposphere in certain cases pose serious problems in achievi
ing reliable microwave communication and there are situations when the LOS
link m ay deviate from its expected path of propagation. l!OS m icrowave links
are in general found to be very reliable and are widely used. But the tropo­
spheric medium through which the signal propagates may| create problems in
such links (Beam and Dutton 1968, White 1970, Liebe 1975, p it 1976, Andrianov
1982, Singhal 1981, Schiavone 1982,83, Robertshaw 1986). Bending of LOS path
from its desired direction is though not a rare phenomenon, the angle of devia­
tion of the beam leading to black-out should be a rare joccurrence in well
designed links. The effects of bending of such path (while traversing the tropo­
sphere) on the received signal Is analysed and reported byj m any earlier work­
ers (Webster 1982, Pari 1983, Webster and Scott 1987).
Researchers have already designed models correlating tjhe meteorological
parameters (like pressure, temperature and humidity) with fades of received
microwave signals, of different depths (Schiavone and Hdrmiiier 1984), as nu­
merical treatments of the problems. In the present work here| an effort has been
m ade to predict fade depth through a model, associating it with RRI grad.,
which is a function of meteorological variability such as ternperature, pressure
and humidity.
:
For this purpose the Milmilia-Durgasarovar microwave link has been selected
because of the availability of radio sonde data (radio refrpctive index gradil
58
ent) taken at the mid path of the hop. Further, meteorological parameters like
humidity, temperature and pressure data have been available from sensors
operated at Gauhati University (7 kms from Durgasarovar). The sodar echograms
have also been available from tw o sodars, one located at Gauhati University
and the other at 15 kms from the receiver site.
The aim of this chapter Is to develop a model to determine the maximum sig­
nal fade depth (at frequency 6 - 7 GHz) that may be generated by a particular
radio refractive index gradient and also to examine the validity of the model
by comparing the predicted output with the observed fade depth.
4.2 ANALYSIS OF FADE DATA
For the analysis, the amplitude data are examined for occurrence of fade
events basically for the following tw o periods:
■j
,
i.
(1) 4 a.m. to 7 a.m.
(2) 4 p m to 7 p.m,
j
I
These two periods are selected because of presence of RRI data from the IMD.
Here the fade depths are scaled at Id b Intervals for each easel' and the maxi­
mum fade depth for that particular sample is noted.
j
J
{
The RRI grad, is calculated twice a day from the surface to 500 meters from
data received through radiosonde, operated at the propagating path (MiimiilaDurgasarovar, as mentioned earlier). As the transmitting anteipna at Milmllia.
and the receiving antenna at Durgasarovar lie in the height range of 100 me­
ters to 300 meters, the RRI grad. Is evaluated upto the height of 500 meters, in j
>
!
cases where the RRtgrad. increases or decreases uniformly (say from surface >
upto 500 meters), then the slope is defined by a single parameter Nr But it is;
observed that in an anisotropic environment, this gradient loses its uniformity
and in that situation another slope N2 which measures the gradient between
100 meters and 500 meters (N, then gives RRI grad, from ground to 100 meters)
is calculated. These heights are taken from gradient calculation because ra69
dio sonde generally gives the temperature, pressure and humidity measure­
ments at the surface and at 100 meters and 500 meters heights. Fig. 4.1 and
fig. 4.2 give some representative RRI grad, profiles at sub, super-refraction and
normal atmospheric environment,
43 CORRELATION OF RRI GRAD. WITH FADE DEPTH
4.3.1. Case 1 : When RRI grad, (between 0 to 100 meters) indicates normal
lapse rate condition and RRI grad, (100 meters to 500 meters) gives subrefractive state or N 1 is in sub-refractive state and N2 gives normal lapse
rate.
From all the RRi profiles, the cases where one of the slopes (say from 0 to
100 meters or 100 to 500 meters) defines sub-refractive condition and
the other (100 to 500 meters or 0 to 100 meters) corresponds to normal
lapse rate, have been taken out. The corresponding fade depths of the
microwave signal over the path records are noted. As one these (MlimlliaDurgasarovar) from the amplitude slabs is In normal state (assuming no
fades), the relation between maximum attenuation (that is, the maxi­
mum positive and negative swings of the signal from the mean value)
and the sub-refractive gradient only has been examined. For this pur­
pose a model is m ade by looking into the variation pattern of fade depth
' with sub-refractive RRI grad, and the relation is expressed as
F * 57.B - 13.7 In {(-1) x } ------ -- 4.1
Where x is the sub-refractive gradient and F Is the fade depth in db.
From fig. 4.3 It is seen that when the RRI grad, changes from normal to
; sub-retractive conditions the fade depth changes in an exponential
pattern. As one approaches near to an acute: sub-refractive state (say
- lON/km), the fade depth becom es very serious (about 26 db).
60
I
i
Fig 4.1: Samples of R. R, I. profile (0-00 GMT)
61
Fig 4.2 Samples of R. R. I. profile (12 - GMT)
62
R.
R.
I.
grad (-ve)
Fig 4.3 - Best fit curve o f model equation No. 1
63
4.3.1 a: The lower slab atm osphere (0 to 100 meters) is in normal state a n d the
upper slab atm osphere Cl00 to 500 meters) Is In sub-refractive state.
In this case, events w hen the lower slab of the atm osphere (0 to 100
meters) is in normal state a n d the other level (100 to 500 meters) is in subrefractive state h a v e b e e n considered. The corresponding fa d e depths
of the 6 G H z signal are then received from am plitude record. A scatter
plot associating the m axim um fa d e d epth a n d thb sub-refractive gradi­
ent N2 is drawn, a n d is show n in fig. 4.4 a lo n g with the best fit line. The
fa d e d epth In d b Is given b y the relation:
F ■ 78.6 - 19.8 In {(-1) N 2 } — ------4.2
it is seen that for the worst refractive situation (say -10 N/km) the ex­
p e cte d d epth is 33 d b a n d for dN/dH = -15 N/km ,the fa d e d epth re­
ce ive d is 25 db. It is therefore clear that fa d e depths for those d ays w hen
N2 is In sub-refractive state are higher than in the, c a se s w h e n N2 is in
norm al state a n d N 1 in sub-refractive state.
4.3.1 b: The RR1 grad. N1 (0 - 100 meters) indicates sub-refractive state a n d N2
(100 to 500 meters) show s norm al condition.
Unlike the c a se of the fig. 4.4, here the c a se s w h e n lower slab (0 to 100
meters) Is in sub-refractive state a n d the higher slab (100 to 500 meters)
Is In normal state are considered. The sa m e a p p ro a c h a s taken earlier in
detecting fade depth for e a c h such c a se Is a d o p te d here too. The scatter
profile is then draw n b y associating corresponding fa d e d e pth a n d N 1
(the sub-refractive state) a n d its best fit curve Is given in fig. 4.5. The fa d e
d epth is given b y the relation :
F = 36.3 -9 .3 In {(-1) N 1 }............. 4.3
64
Fig 4.4: Best fit curve o f model equation No. 2
66
F = 36.3 - 9.3 In [(-1) x]
Fig 4.5 : Best fit curve of model equation No. 3
66
The fig, 4,5 shows that ©von for the worst atmospheric condition repre­
sented by dN/dH = -10 N/km, the fade depth may record only 15 db
and for dN/dH = -15 N/km, the fade depth of 11 db Is only received.
4.3.2, Case 2 : The RRI grad, of both the slabs (0 to 100 meters and 100 to 500
meters) indicate sub-refractive states.
The fadings observed In such situations are again examined in relation
to N 1 and N2 and the maximum fade depths when both the slabs have
sub-refractive grads, are taken out from amplitude records of 6 GHz link.
The maximum fade depth so received are associated with N1 and N2
for three different conditions:
a) average values of N 1 and N2 are taken without looking into weather
N 1 is more sub-refractive than N2 or vice-versa.
b) N 1 more sub-refractive than N2.
c) N2 more sub-refractive than Nl.
The results are presented b e io w :
4.3.2.a:
The RRI grad, (average of both the slabs are in sub-refractive state
(irrespective of either Nl > N2 or N2 > Nl.
in the first case here, the average values of the sub-refractive slopes and
corresponding maximum fade depths are sorted out and plotted to
examine the association between these two factors. This scatter plot of
fade with average RRI grad, is shown In fig. 4.6. The best fit relation be­
tween these parameters Is represented b y :
F ■ 76.5 - 16.7 In {(-1><N! + N2) / 2 } ------ 4 4
This indicates that for a maximum sub-refraction state, of dN/dH = -10N/
km, the fade depth may go up to 38 db and for a sub-refractive state of
dN/dH = -15 N/km the fade depth goes down by about 7 db from that
67
it
F = 76.5 -16.7 In [(-1) (N, + N2)/2]
Fig 4.6: Best fit curve of model equation No. 4
68
of maximum sub-refraction conditions.
4.3.2.b:The lower slab is in more sub-refractive than the higher level.
Here, only the events when N1 represents more sub-refractive condition
than N2 is considered. Al! relevant events are then dssociated by draw­
ing profiles between these two variables (N1 + N2)/2 with corresponding
fade depths recorded in the Mitmilia link. Such a profile is shown in fig.
4.7. The best fit curve for the scatter plot is given by :
F - 36.1 - 8.6 In {<-1)<Nl + N2)/2}
4.5
Here, it is seen that the fading does not exceed even 16 db, for RRI grad.
-lON/km and for a RRI grad. -15N/km, the fade dejpth of 13 db is fairly
i
low compared to the other case.
■
4.3.2.C: The upper ievei (100 to 500 meters) is more sub-refractive than the lower
level (0 to 100 meters)
This situation is just the reverse of the above case. Tnte relation between
fade depth with (N1.+ N2)/2 are examined againj with the same ap­
proach adopted above. It is seen that in such a situation the fade depth
can be expressed as:
F « 81.7 - 18.3 In {(-1) (N1 + N2)/2 } ............i4.6
!
Here, for a sub-refractive condition with RRI grad. [-10 N/km, the fade
depth is 40 db, and for dN/dh = -15 N/krn. the fade; depth is 32 db. Trie
1
fade depth of 40 db Is very significant as lt leads to Ijnk cut-off. All these
values are higher compared to what were received in all other cases. It
is therefore clear that the fade; depth increases wheri N2 represents more
sub-refractive situation (fig. 4.8).
!
69
F = 36.1 - 8.6 In [(-1) (Nt + N2)/2]
Fig 4.7: Best fit curve of model equation No. 5
70
R.
R.
I.
grad (-ve)
Fig 4.8: Best fit curve o f model equation No. 6
71
4.3.3. When the upper slab (100 to 500 meters) is in super-refractive state and
th© lower slab (0 to 100 meters) Is In normal state or the upper slab (100
to 500 meters) is in normal state and the lower slab (0 to 100 meters) is in
super-refractive state.
This event is seen predominantly during the post-monsoon period. Here
the following situations are considered:
Cl) NT gives normal and N2 shows super-refractive state,
(2) N1 shows super-refractive and N2 gives normal state.
The maximum fade depth for these days are screened out from the
amplitude records and are plotted along with the super-refraction gra­
dient. The scatter plot and the best fit line is shown in the fig. 4.9 and the
relation between these parameters Is expressed by
F = 47.3 ln<-1) (x) - 205.76------- 4.7
Where x is the super-refractive gradient. The fade depth of 17 db is ob­
served for a super-refraction condition of RRi grad. -110 N/km (fig. 4.9)
and the link cut-off of 45 db Is evident when the atmosphere represents
a high super-refraction situation (dN/dH = -200 N/km).
4.3.4. Normal gradient.
In order to examine the. changes in fade depth, if any, when the RRI
grad, shows a normal atmospheric situation from 0 to 100 meters and
also from 100 to 500 meters the RRI grad, (i.e., the normal atmospheric
situation) is associated with corresponding fade depths from amplitudes
data. It Is observed that for such a situation fading is extremely low. The
' numerical relation for achieving fade depth at such RRI grad, can be
expressed a s :
F = 0.13 In (-!) x-0.5...........4.8
where x is the value of normal RRI grad.
72
F = 47.3 In [(-l)x ]- 205.76
Fig 4.9: Best fit curve o f model equation No. 7
73
F = 0.13 In (-1) (N1 + N2)/2-Q,5
4.9
This Indicates that for the values of RRI grad, lying between -40 N/km to
-100 N/km. the fade depth of a microwave signal Is either zero or has
negligible value (less than 1 db).
4.3.5 Single sub-refractive slope (0 to 500 meters)
The atmosphere Is represented by a single sub-refractive slope upto 500
m. In that case the fade depth can be represented in association with
the RRI grad, by
F = 53.8 - 13.2 In (-1) x — ----- 4.10
Where x represents the single sub-refractive index. From fig. 4.10 it is seen
that for dN/dH = -lON/km, the fade depth may go up to 23 db. and for
dN/dH = -15 N/km, the fade depth goes down atleast by 6 db from the
maximum sub-refractive situation.
4-4
DISCUSSION
Prediction of fades (In microwave links) through models based on meteoro­
logical parameters has been presented by many earlier researches (Weinstein
1980, Sweet 1980, Richter and Hitney 1976, Mazumdar 1976, Craig and Levy
1987) Schiavone and Hermiller<1983) have predicted with fair success the probability of occurence of fades from Pasquiil index (density stratification of the
ABL) and average temperature. Again studies were conducted over ocean to
determine the role of water vapour on microwave propagation, based on ob­
served propagation features, models are developed.
The model framed In this chapter Is aimed at to predict fading from the
RRI grad. It is seen that fade prediction is highly dependent on the atmos­
pheric situation specially In the height range of the transmltling and receiving
antenna. The model shows that if the heights of the transmitting and receiving
antenna lie above 300 meters, the fade value calculated from the RRI grad.
74
*
F = 53.8-13.8 In (-l) x
R.
R.
I.
grad (-ve)
Fig 4.10: Best fit curve of model equation No. 10
75
through a single slope b etw een 0 to 500 meters gives a lower fa d e value than
ob served In practical c a se (Table 4.1), The differences in the fa d e d e pth ca l­
culated through a single slope (0 to 500 meters) RRI grad, a n d b y using tw o
slopes (0 to 100 meters) a n d (100 to 500 meters) m a y at times b e very serious a s
show n in the table 4.1. It Is seen that while a n error of 5 3 % In fa d e predicted
values from a single slope RRI grad. Is observed, the fa d e ca lcu la te d b y tw o
slope m odels gives a n almost accu rate fa d e prediction. .
However, d e p e n d in g on the transmitting a n d the receiving a n te n n a heights, a
single slope RRI grad, (upto 500 meters) m a y som etim es well predict the fading
over a link. For a link like Laopani, It m a y not b e necessary to determ ine two
slopes as the transmitting a n d receiving a n te n n a heights’He within 50 meters
a n d 100 meters a n d a single slope RRI grad, should b e s u f f ic ie n t ^ estimating
fades. In such ca se s o n e c a n probably use grou n d -b a se d m eteorological p a ­
rameters for predicting fa d e s as these grou n d -b a se d param eters c a n effec­
tively b e used for estimating RRI grad. Such attem pts of defining RRI grad, from
gro u n d -b a se d d a ta w ere m a d e b y m a n y earlier researchers (Oyinloye 1987,
1979, Schiavone 1981) a technique for receiving RRI grad, from g rou n d -b a se d
m eteorological d a ta Is defined b y this group also. (Sharmq etal 1998). Here
the RRI grad, Is co m p u te d from pressure, tem perature at humidity m easured at
only ground level after a n d from 500 meters height observations. In this c o m ­
putation the RRI grad, is considered to b e linear (De etat 1994) i.e., the atm os­
phere is considered to b e Isotropic.
.
* .
In a n attem pt to calculate fa d in g over Laop ani through RRI grad, the m odel
fram ed b y this group has b e e n adop ted . The use of this m odel is justified b e ­
c a u se the transmitting a n d the receiving a n te n n a heights of the Laopani link
b e in g within 100 meters only a single slope RRI grad, m a y b e welt a d o p te d for
calculating fa d e depth.
Fig. 4.11 gives the fa d e values calcu late d b y this m odel for all the atm ospheric
situations (sub-refraction to super-refraction). The fa d e s exp e rie nce d b y micro-
76
O
W
H
a
lx*
<
Q
n
10
15
18
R.
20 , 25
25
R.
30
I
201
35
.
t
l
48
l
I.
40
h
45
&.
grad
50
72
fk $
(-ve)
100
105
% h &&
Fig 4.1 T Calculated fade depth from model output ana
observed fade depth at Laopani station _
i
i
a
ot
7 7
112
t
115
A A.
120
* *
125
140
n
X theoiitlcal values
• a observed values
I error
140
155
ft
160
Jzf
CM
VO
cn
9
vn
4.5
oo
r— W
1
7 8
4.3
4.3
4.7
oo
-57
a \
*— i
*-4
O
1
o
1
VO
««(
4.2
4.2
o
o
o
o
o
4.10
00
00
Tj1
4.8
"Sf’
-25
H
-50
-96
00
-20
t
-66
Os
-57
4.10
o
4.3
4.10
00
<?
4.3
Computed fade
depth(dB) for
single slope
o
-64
4.10
-36
Model adopted
equation No.
R R I grade
calculated
assuming
single gradient
from 0 - 500m
t-H
-42
CM
4.5
Fade depth
computed (dB)
«n
-22
cn
Model output
equation No.
TABLE-4.1
so
so
X|*
00
CM
ON
oo
I
i
T
wave signal during these conditions are also shown in the figure. It isvery clear
that the model values approach the observed fade values for all the cases
under study.
But the atmospheric scenario is not that simple as assumed in the models. In
fact, atmosphere isIn constant state of turbulence which may result to a change
in the RRI at a very localised region. Such localised disturbed region are very
difficult to locate and also to Identify the types of irregularities, though these
structures may cause fluctuations of the signal strength. One of the param­
eters that measure the strength of such disturbance is Cn2(Chapter I). Cn2can
be calculated from fluctuations In the RRI grad, or from the fade depth records.
In order to examine the effect of such turbulating region on fades, this param­
eter has been computed from fast fade record received over Milmiiia link. This
exercise is made to Identify the sources for generating fade depths which can­
not be explained from the model computed values. The Cn2values are there­
fore computed in two modes, one In 30 minutes in slow mode and the other is
5 minutes in fast mode and It is seen that there may be large changes of Cn2
within two minutes time. Fig, 4.12 a, b shows two such cases. It is also seen that
if the value of Cn2 goes upto the order of IQ*9 m*2/3
(Sharma et al 1996) an
additional fading of 4 to 5 db Is generated over the fading caused by the
presence of a medium in the LOS path. This is to some extent can explain the
discrepancy observed In the model computed output and the actual fading
experienced by the signal.
9SI
79
.. t.. .
23*06
13-30
7X10-13
a
o
8X10-12
2X1042
F ig 4 .1 2 ; C alcu la tio n o f c n 2
(a) S lo w m o d el
(b) F a s t m o d e l
80
CHAPTER V
PREDICTION OF FADE
THROUGH RAY TRACING MODEL
5.1 INTRODUCTION
The ideal free space communication assumes an atmosphere with perfect
refractive index condition and with normal lapse rate in RRI grad., and aiso
that earth's reflection co-efficient is negligible. But in practice the atmosphere
is far from ideal and even the clear air effects may degrade the signal quality.
However, such effects to a great extent have been taken care off by applying
techniques like frequency and space diversity systems. Further, introduction of
digital communication links has improved the reception quality, and phenom ­
enon like knife e d g e diffraction that may cause fading In analog com munica­
tions will not affect the digital links, in addition, the S/N ratio is improved in
digital systems as In this technique the signal at a repeater station can be re­
generated unlike the analogue-system where the transmitted signal at a re­
peater station is retransmitted after,, converting Into intermediate frequency
>»
etc. In this process noise is a dde d to the signal. But the advantage of ana­
logue modulation (basically with fM) Isthat the nonllnearlties in the system pro­
duce less distortion than the digital system and therefore this communication
m ode is still in existence. In this region where the study is conducted, a number
of analog links are still active and it Is seen from our earlier discussions and
analysis that the fadings experienced by m icrowave signal are fairly com plex
as links pass over different terrains and fade prediction in such links is still a
necessity. Further, even for similar terrain the environment may be different, for
example, a flat terrain may be marshy or dry or contains buildup areas. So it is
very likely that the atmospheric situation will differ from one environment to the
other, which ultimately is reflected In the RRI variations. The variation in the
81
refractive index gradient from its normal atmospheric condition results in fad­
ing of microwave signai. So if the prediction of microwave fading through RRI
can be made, it will help not only in the future link design but also performance
of existing links can be improved. With this objective in mind, it is planned to
develop a theoretical model by using ray-tracing technique for calculating
fade depths of microwave signal when the signai passes through a tropospheric
"medium". The medium defined in this mode! is that medium, where the ray
bends from Its normal path because of changes in refractivity of this medium
from its normal lapse rate. It is also assumed that this medium" or layer is con­
sidered to have a fair degree of stability (within the layer) and therefore scat­
tering of signals from such layers is considered to be absent.
5.2 TECHNIQUES ADOPTED
It has been shown by earlier workers that the fading caused by bending of the
main beam Is more serious than that of multipath fading
(Gilloi 1985, Ruthroff
1971 Webster 1983). However, there are also reports showing that multipath
interference (that between the main beam and its scattered components)
(Crawford and Jakes
1952, Delange 1952, CrawfordandSharpless 1946, Meadows etal 1966, Inoue
and Akiyama 1974, Benardlni ef al 1977, Mongensen 1977, Sandberg 1978,
Bolomqulst and Norbury 1978, Rummler 1979b, Sasaki and Akiyama 1979,
Sandberg 1980, WebsterandLam 1980). But even with multipath mechanisms,
the fading due to three multipaths (Ruthroff 1977) is observed to be more ef­
fective though seven or eight paths may reach the receiver point (Stephansen
1981). It has also been shown that, even there is an increase in attenuation
may sometimes degrade the signal quality significantly
with the number of paths taken by the signal (for example when the paths
increase from tw o to three. It has resulted in increase of fade depth by 10 db),
to generate the fade depth beyond 20db only tw o multipath interference may
be sufficient
(Martin 1980). Considering these aspects It is p la n n e d :
1. to develop a ray tracing model for evaluating fade depth through estima-
82
tion of th© deviation of the angle of arrival of the main beam at the receiv­
ing end. The bending of this ray will be calculated by considering the "me­
dium" (that the ray will enter on its path) to be In sub-refraction, super­
refraction conditions and also to have anisotropic properties in some cases.
2, to estimate multipath effects on fading by Introducing multlpaih factor as
another input to the model, described In (1).
5.2.1 DESCRIPTION OF THE RAY TRACING MODEL
The mode! is framed to evaluate the fade depth that will be experienced by a
signal passing through a "medium" having different dN/dh gradient from the
background environment. The Inputs to the model a re :
(1) Variable hop length.
(2) dN/dh within the "medium" to be either in the sub -refraction or in the
super-refraction.
(3) A point of disturbance "D" where the gradient Is assumed to change from
its normal lapse rate (-40N/km). This point of disturbance is taken at various
points of the HOP length.
(4) Different looking angles between the transmitter and the receiver.
THE RAY TRACING M O DEL: THE APPROACH AND THE MODEL OUTPUT
The model estimates the angle of arrival of the beam at the receiver when this
beam has to passthrough a medium having refractive index gradient different
from that of the ideal (normal) atmosphere. The refracted path that the ray
would take in such cases will be calculated from geometry (fig. 5.1). The fol­
lowing cases are considered for this purpose :
Case I : It Is assumed that the difference of antenna heights between the
transmitter and the receiver and the path length are constant but point of
83
Transmitt ng Antenna
T
84
disturbance "D " is variable.
So considering these basic parameters, the ray tracing is applied to calculate
bending of the link of the path on way to the receiving antenna. The following
are the inputs for this purpose :
Difference of
antenna height
Path length
(km)
Starting point of the "medium"
(from the transmitter)
(m)
(km)
100
40
20
100
40
0.10
100
40
10
100
40
39
it is assumed that the difference In the height between the transmitting and
the receiving antenna to be 100 meters and that the path length to be 40 km.
Also it is considered that the atmosphere shall have normal lapse rate from the
transmitting end upto a defined point "D " which is the starting point of distur­
bance in the medium and also that after point “D" the refractive index gradi­
ent changes from Its normal value and will remain so for the rest of the path.
The point of disturbance “D" is considered at the mid-point of the hop length.
In the first case of our study the disturbed condition is assumed to be present
upto the receiving antenna point that is the '‘medium" or the layer has an
extension from the midpoint of the hop to the receiving end. It is also consid­
ered that from the transmitter end up to this point of disturbance “D", the at­
mosphere has normal lapse rate as defined below:
RRI = 320 N at 600 meters height
RRI = 340 at 100 meter height
RRI grad. dN/dH = -40 N/km.
At the point of disturbance "D " the R.R.i. grad, changes to super-refraction
with values assumed
RRI = 340 at 100 meters height
85
(
RRI = 350 at 50 meters height
A signal of 5 cm wavelength is allowed to pass through the atmosphere with a
grazing angle of 0.143°, Then considering this looking angle and also the
changes In RRI grad. (In the "medium"), the angle of deviation of the main
beam at the receiving point from its original angle is calculated from a simple
ray tracing technique. A linear deviation of 3.49 meters isthen observed a t the
receiving antenna, when the point "D" lies at the midpoint of the hop. Now
the point of disturbance Is shifted very near to the transmitter, keeping other
parameters same as the above. The angle of arrival of the beam at the re­
ceiver is again calculated by the ray tracing technique and the linear shift of
the beam at the receiving antenna is estimated to be 6 meters. Similar calcu­
lation is done when the point of disturbance Is at 10 km from the transmitter
and as expected, the ray deviation is observed to be 5.2 m, which is less than
the previous case.
The deviation of the beam is observed to be minimum when the point "D" lies
very near (39 km) to the receiver (as expected) and the magnitude of this shift
is only 0.18 meters (fig 5.2) from its normal position.
Case 11: In this case it is proposed to keep the difference of antenna heights
constant and the point of disturbance "D" is considered at midpoint of the
hop length and the path length Is considered as variable. The inputs are as
under:
Difference of
antenna height
•
Path length
(km)
Point of disturbance
(from the transmitter)
(m)
(km)
100
20
10
100
25
12.5
100
30
15
100
35
17.5
100
40
20
86
6
Vertical deviation (m)
Difference In antenna
height = 100 m
Hop length = 40 km
0
10
20
30
Distance from transmitter (km)
40
Fig. 5.2 : Linear displacement of llne-of-slght for various distance of point of
disturbance, keeping path length and antenna height difference constant
87
In this case the deviation of the main beam at the receiver end is determined
by assuming fixed difference in antenna height (100 m) and the point of distur­
bance is considered to be always at the center of the hop length but the path
length Isconsidered to vary from 20 to 40 km. it is also assumed that the normal
atmospheric situation prevails from the transmitter-end to the mid path of the
hop.
The ray tracing is done for different path lengths from 20 to 45 km. (at a step of
5 km) to calculate the angle of arrival of the main beam for different hop
lengths. Fig 5.3 shows the deviation of the signal beam in meters for such situa­
tions, It is seen from the graph that as expected, with the Increase in path
length deviations also Increase. For a path length of 20 km., the deviation is
1.75 m. when the path length Increases to 25 km, the deviation also increases
to 2.12 m. With further increase of path length by step of 5 km, the beam devi­
ates by 2.5 m, 3 m, 3.49 m, 3.8 m respectively.
Case HI: In this case the path length and the point of disturbance (at the mid­
point of the hop length) are considered to be constant, whereas the looking
angle is considered to be varying. The Inputs accordingly are :
Path length
(km)
Point of disturbance
(from the transmitter)
(km)
Difference of antenna
height
(m)
40
20
100
40
20
80
:
In this case, to introduce a change In the looking angle, the difference In an­
tenna heights are varied. Here too it is assumed that the disturbed condition
persists from midpoint of the hop to the receiver and the ray tracing technique
isthen applied at different looking angles, say, from 0.14° to 0.11° and the cor­
responding difference In angle-of-arrlvai of the beam at the receiver are cal­
culated out. These deviations of 1 meter and 3.49 m for 0.14° and 0.11° looking
88
03
to
—•
i- - - - - - - !- - - - - - - 1- - - - - - - r
o
Vertical deviation (m)
&
5-
0
20
25
30
35
40
45
Path length (km)
Vertical deviation (m)
1.75
2.12
2.5
3
3.49
3.8
Fig. 5.3 : Deviation of the signal beam for various hop lengths keeping the point of
disturbance at the mid point of hop length and difference of antenna height constant.
89
angles are shown In fig, 5,4.
Case IV : Here, the difference In antenna height Is taken to be constant but the
zone of disturbance is assumed to be more than one, and the path length is
also considered to vary.
So in this case, the atmosphere is supposed to be in a complex environment
compared to the earlier cases, in our earlier cases, it is considered that the
microwave signal first enters into normal atmosphere and then goes to a "me­
dium or layer" where refractive index changes from its normal value to sub/
super refractive values and it is also assumed that the disturbed situation per­
sists always upto the receiving end. But in this case we will consider that the
disturbed layer may not extend upto the receiving end, and there may be
localised layer/layers of disturbances of different lengths. So unlike the earlier
cases it is now required to consider the presence of at least three RRI gradients
in the atmosphere, as described below ;
(1) from the transmitter end to the point of disturbance
(2) in the disturbed layer itself
(3) from the rare edge of one of the disturbed layers (nearest to the receiver)
to the receiving end.
So considering this situation, the following inputs are given to the model
Difference antenna
Point of
Extension of the
Path length
height
disturbance
layer
(km)
(m)
(km)
(km)
100
1
6-
20
100
1
9
30
100
1
12
40
Assuming the difference between the transmitting and the receiving antenna
height to be constant (100m), the hop lengths as well as the extension of the
90
4
03
K5
—<
Vertical deviation (m)
Hop length 4 km
Point of disturbance 20 km
0
80
100
Difference in antenna height (m)
Vertical
deviation (m)
Fig. S .4 : Linear deviation of iine-of-sight for different looking angles, for fixed hop
length and point of disturbance.
91
Irregular medium/layer are then varied to calculate the deviation of the main
beam at the receiver. The medium is considered to be In sub-refraction/superrefraction condition at a fixed distance from the transmitter (say 1 km). The
corresponding deviation of the beam for each case is then calculated by
adopting the same ray tracing approach and these are shown In fig. 5.5. For a
path length of 20 km., the deviation is calculated to be 0.78 m. When the path
length increases from 20 km to 30 km, the deviation also increases to 1.26 m.
For a 40 km hop length, deviation is 1.9 m.
5.2.2 INTRODUCTION OF MULTIPATH PARAMETER AS ANOTHER INPUT TO THE
ABOVE MODEL
In this case the effects of scatters (when the scattered reflections are present
along or in an elevated portion of the main LOS path) on the beam path will
be considered along with the bending of the main beam (as described above).
Then the fading observed is the resultant of the deviation of the main beam
and the reflected components of the signal from the scatters. The signal strength
of the scattered rays is generally very low in comparison to the main beam,
and therefore the deviation of the main beam generally dominates the fade
depth
(Websteretai 1987), But the presence of scattered may sometimes cause
large signal fluctuations because of the Interference between the direct ray
and scattered signal, especially when the size of the scatters are comparable
to the wave length of the signal and they are situated within the first Fresnel
zone of the ray path
(Stephansen 1981Martin 1980). It istherefore aimed here
to see the effects of such scatterers along the line-of-slght path and to exam­
ine and evaluate the situation when the main beam and the scattered rays
will interfere to cause fading. For this purpose the deviation of the main beam
at the receiving end Is first estimated by ray tracing technique for the link pa­
rameters as defined below, and then the effects of the secondary rays from
the scatters on the generation of fades will be found out. The link parameters
92
<N
Vertical deviation (m)
CO
Difference of antenna height 100 m
Point of disturbance 1 km
Thickness of the layer 25 m
O
20
30
40
Hop length (km)
Vertical
deviation (m)
0.78
1.26
1.9
Fig. 5.5: Linear deviation of iine-of-slght for different path lengths, and the disturbed
region is confined within a limited zone.
93
are the follow ing:
Hop length
"D" point
RRI at
RRI at "D"
RRI grad.
(km)
(km)
transmitter (N)
point (N)
N/km
41
20
320
311
-180
The deviation of the main beam in presence of a "medium" (transmitter and
receiver antenna height at 278 m and 173 m respectively) at the receiving end
is first determined. Now, considering the presence of scatterer inside the first
Fresnel lobe (say at a height of 180 m), the ray tracing is made to examine
whether the scattered ray will meet the receiver and will interfere with the
signal defocussed by the medium present in the ray path. The results (fig. 5.6)
show that the scattered ray wil! intersect at the receiver with the main beam
which is deviated by the medium.
in such situations, an additional fading of magnitude of 4 or 5 db may be
generated because in the presence of small Irregularities in the medium, the
parameter Cn2 has a high value of magnitude 10'14to 10*9 nrr2/3 (Sharma et ai
1996). The fadings so generated may to some extent explain the relatively large
discrepancy seen between observed fading calculated through only bend­
ing of ray by the "medium".
5.3 DISCUSSION
Ray tracing Is an attractive method for generating images because it not only
produces realistic optical effects such as reflection, refraction, defocussing
(Hodges ondAdelson 1995) but also determines grazing angle, ground range
and slant range for higher altitude propagation paths (Robertshaw 1986).. With
this view in mind, and also to ca lcu la te fadings through estimation of
defocussing of the main beam of a microwave link, the ray tracing model has
94
Fig. 5 .6 : Linear deviation of the iine-of-sight along with the introduction of multipath componen
jeAjaoey
95
been framed In this chapter, There were earlier efforts on evaluation of fading
In Une-of-slght link by modelling. Webster (1983) has shown that by measuring
the delay time of the angle of arrival of the signal, while It passes through an
atmosphere of different meteorological variabilities over the link path, fading
can be calculated. He has also shown In his model that significant fading may
occur due to the deviation of the main beam when an irregular layer devel­
ops between the transmitter and the receiver (within the main link). The present
model also has this as a basic approach. To assess the reliability of this model,
the estimated fade depth from the model output is compared with the ob­
served fade values, and for this purpose fade detected over Milmilia is taken.
This link is selected because of the availability of RRi parameters as taken by
the 1MD over this link path. The RRI grads, are selected for the days when the
parameter gives either a super-refraction or a sub-refraction condition. The;
point of disturbance is then assumed to be at the mid-path of this 40 km hop.
This assumption may not be objectionable as the IMD is situated almost at the
mid-path of this link. The table (5.1) shows the model output and the observed
fade values and it is clearly seen that the model estimates the fade depth with
fair accuracy, However, If the point of disturbance is shifted to the transmitter
side, the model output overestimate the fades. On the other hand, if the po in t ‘
of disturbance Is shifted to the receiver side then the model output underesti­
mate the fades (table 5.3). it may thus be concluded that a precise estimation
of the fade depth can be calculated if the RRIgrad. is available at the trans­
mitter end as well as at the midpoint of the hop and the point of location of
the disturbed layer is known. Even in situation where a number of radio sonde
data over a path is not available or, difficult to obtain, ground based meteoro­
logical, parameters (temperature, pressure and; humidity) can effectively cal­
culate the RRI grad, (refer to Chapter !V) which; in turn, can facilitate the pre­
cise estimation of fade depth.
96
T A B L E - 5. 1
Calculated
deviation
(m)
Calculated
fading
(dB)
Observed fading
(dB)
1
2,03
8
8
2
0,7
2,5
3
3
1.04
4
■4
T ABLE - 5.2
I
Calculated
Calculated
deviation from
fading
R R Ia t
MBS
M ID P O IN T
8
Calculated
deviation fro m
•
2 .0 3 m
n
Calculated
Calculated
deviation from
fading
R R Ia t
(dB)
M ID P O IN T
- 2 .5 dB
“ 0 ,7 m
.
R R 1 a t 2 km
from
transm itter
- 6 .6 m
Calculated
deviation from
R R ia t4 k m
■ from
transm itter
- 1 .8 9 m
Calculated
Calculated
Calculated
Observed
fading (dB)
deviation fro m
fa d ing
fading (dB)
R R I a t 4 0 km
- 1 . 8 dB
8 dB
- 2 6 dB
fro m tra nsm itte r
- 0 .4 m
Calculated
Calculated
Calculated
Observed
fading (dB)
deviation fro m
fa d ing
fa d ing (dB)
R R I a t 4 0 km
- OdB
-3 d B
- 6 . 6 dB
fro m transm itter
- 0 .0 8 m
97
# Include < stdlo,h>
# include < mat.h>
main ( )
{
float theta, r2, rl, x, y, xd;
float intercept, d theta, il, 12;
float eta mu, h, reta, emu;
float p;
print f ("\m Enter x y rl r2 xd h & m u:\x");
scan f ( “% f % f % f %f % f %f %f", &x, 8cy, &rl, &r2, &xd, &h, &eta);
mu=90-eta;
\* Conver into radian * \
reta=(3.1416*eta)\ 180; rmu=(3.1416*mu)\ 180.0;
theta=atan(x/y);
d th e ta -18Q,Q#theta/3,1416;
i 1=(x-xd)/sin(theta);
i2»(90-dthetaHasln(Crl/r2)#cos(theta)))#180/3.1416;
intercept= 1*(i2#3.1416/180);if(h!=0)
P-(h-y)/tan(reta)+h*( 1/tan(rmu));
printfCx-%.2f y=%.2f xd=%.2f rl=%.2f r2=%.2f intercept=%.2f\ntheta=%.2fp=%
if (p>x)
printf("rays will not intersectin'');
also
printf("rays will intersect\n");
1
98
Aarons J,
Klobuchar J.A, Whitney H.E,
Austen J, Johnson A.L,
Rino CL (1983)
Radio Sci,18,421
Aarons J
Dasgupta A (1984)
Radio Sci 19,731
Adeniyi J.D. (1997)
J. atmos, terr, phy, 59,1205
Ames L.A.
Newman P,
Rodgers T.F (1955)
" VHF tropospheric over water
measurements for beyond the
radio horizons" Proc IRE
Vol. 43 P.P. 1369 - 1373,
Andrianov (1982)
Raddio Sc. vol.- 17, No.- 5,
Pp. 888-895 Sep/Oct.
Appeleton EV,
Lyon AJ,
Pritchard AG, (1955)
J. atmos, terr, phy,
7,292
Balan N,
Aiyer K. N. (1983)
J. geophy. Res, 88,10259
Balan N,
Rao P.B. (1984)
J. geophy. Res, 89,9009
Balan N,
Rao P.B.,Aiyer K. N (1986)
Proc of India Academy of
Science
pp 409 - 416
Barbara A.K.,
Devi M. (1992)
Paper presented at the
international Beacon Satellit
symposium,
MIT Boston USA
Barbara A.K.
Devi M. (1991)
TR-I
Barbara A.K.
Devi M. (1992)
TR- n
Barbara A.K.
Devi M. (1994)
Technical report
final edition
Barrington RE,
Thrane EV (1962)
J. atmos. terr. phys,
24, 31.
Basu S,
Basu S, Mackenzie E (1988)
Radio Science,
23, 363.
Bean BR,
Dutton EJ (1968)
" Radio meteorology ”
Dover publications Ink
New York
Belrose J.S,
Burke MJ, Coyne TNR.
Reed IE (1972)
J. geophys. Res,
33, 321
Benardini A,
Diblasio G, Mardini P,
Meucci G.F. (1977)
"Channel model and experimental
evaluation of its parameters by
direct observation by the input
response for a line - of - sight
radio link”. Paoer presented at
commission F, Symposium,
Union radio Sci Inst.,
La Boule, France
Beynon WJG,
Brown GM (1959)
J. atmos. terr.
Phys, 14, 138
Bisceglia B,
Franceeschetti G,
Mazarella M, Pinto IM (1988)
"Compuer Program Description"
IEEE Trans, ant & Prop. Vol 36
No. 10
100
Bolomquist A,
Norberry J. R (1978)
"Multipath effect on a terrestrial
path, Influence on atmospheric
conditions on electromagnetic
wave propagation at frequencies
above 10 GHz"
EUROCUP - COST
Final report proj.
25/4, pp 141 - 173,
comm of the Eur,
Communities, Brussels.
Bowles KL (1958)
Phys. Res. Lett. 1,321
Bremmer H (1947)
"The propagation of EM waves
through a stratified medium and
its WKB approximation for
oblique incidience" Physica,
vol -15 pp - 594 - 608.
Buffington K (1947)
"Radio propagation of
frequencies above 30 MHz"
Proc IRE, vol- 351,
pp- 1122 - 1136,
Buffington K (1957)
Bell system Tech J
(USA), 36, 595.
CCIR (1978)
Vol V, Recomendation 369-3
International Telecom. Union,
Comite.
Int. des. radio
comm, vol V, "Propagation in
non ionized media", Geneva
pp- 68- 69
Chandra H,
Sethia G, Vyas G-D,
Despande MR and
Vates HD (1981)
Indian Journal for radio and
space Phys, 10,45
101
De Lange D.E (1952)
Devi M,
Barbara A K (1978)
"Propagation studies of microwave
frequencies by means of very short
pulse", Bell syst. Tech J,
31(1), 91- 103
Indian J Radio space
phys. 7, 142
Craig K.H.,
Kennedy G.R.(1987)
"Case study of LOS mw multipath
propagation." Electron Lett (GB)
V ol-23 No 10, p 483 - 4
Craig KH,
Levy MF
"A forecasting system using the
parabolic equation application to
surface to air propagation in the
presence of elevated layers".
AGARD Conf, Proc. No. 453
Crane R.K.(1981)
"A review of trans horizon
phenomenon,: Radio Sci, 16(5),
649 - 669,
Crawford AB
Sharpless W.H.(1946)
"Furthem observation of A-O-A
of microwaves" Proc IEEE,
34(11), 845-848.
Crawford A.B.
Jakes W.C. (1952)
"Selective fading of microwaves,"
Bell syst. Tech. J, 31 (1), 68 - 90
Day JP &
Trolese'lB (1955)
"Propagation of short radio waves
over desert terrain" Proc IREE
Vol 38, pp 165 - 175.
Deygout J (1966)
IEEE Trans. Ant & Prop (USA),
14,480
Dieminger W (1952)
J. atmos. terr. phys, 2, 340
102
Dieminger W (1955)
"Physics of the atmospher1
pp 53, phy. Soc. London.
Dougherty HT (1968)
"A survey o f mw- fading
mechanism, remedies and
applications", COM - 71- 50288,
National Tech. Inf. Service,
Springfield,
VA, 22161,
Etas - Unis - Amerique.
Ellyett CD (1947)
J. geophys Res, 52, 1
Epstien P S (1930)
"Reflection o f wave in an
inhomogeneous, absorbing
medium" (A Proc Nat. Acad. Sci,
Wash -1 6 , pp - 627 - 637,
Epstien G,
Peterson DW (1953)
Proc IEEE (USA), 41, 595
Fejer FJA,
Vice RW (1959)
J. atmos terr phys, 291
Fitzenreiter RS,
Blamle U (1964)
J. geophys, Res. 69, 407
Gilloi HG (1985)
profiles caused by multipath
"A study of field strength height
fading "IEEE trans Ant & Prop.
Vol pp- 33 No. 12 Dec. 1985
pg. 1378 - 1385.
Gregory JB (1956)
Aust. J. Phys, 9, 324,
Hall NPM (1979)
"Effect of the troposphere on radio
communication"
London Peregrinus Inst.
Elect. Eng. EM Waves. Series
103
Hay DR,
Poaps GE (1959)
"Prolonged signal fadeout on a
short mw path" Com.
J. phy. 37, 318 - 321
HOKL,
Cole RS, Mavrakoukoulakis (1980)
"Spectral analysis of anamolous
mm wave amplitude scintillations
in a town environment" IEEE
Trans Ant prop, vol - 28
pp 941 - 942
Hodge LF,
Adelson SJ (1995)
"Generating exact ray-traced
animation frames by reprojection"
IEEE Computer graphics &
Application.
Hirono M,
Meada H (1955)
J. geophys Res, 60, 241.
Inou T (1966)
IEEE Trans AP 22, 557 - 565.
Inou T (1974)
Akiyama T
"Propgation characterestics on LOS
over sea paths in Japan"
IEEE Trans Ant Prop AP 22(4), 557 - 565
Ishimaru A (1978)
"Wave propagation and scattering
in media, Academic Press,
New York
Ch 17 and 20.
JhaKK,
Verma AK, Tiwari R.K.(1992)
"Radio wave attenuation due to
formation of tropospheric thermal
lenses " URSP
vol 21 pp 1 - 4
,104
Levy MF (1990)
"Parabolic equation modelling of
propagation, over irregular terrain"
Electron Lett, (UK)
vol - 26, N o .-15 p 1153 - 5
Liebe MJ (1975)
IEEE Trans mw theory
Tech vol. MTT - 23 pp 380 - 386,
Livingston (1970)
"The physics of microwave
propagation" (Prentice Hall
Inc. New Jersey), 1970
MaSY,
Xu Y, Yeh K (1995)
J atmos. terr phys. 57, 1037
Martin L (1980)
Etude de la seleetivite des
evanouissements dus aux
trajects multiple paper
presented at Commission
F Sym, Union Radio, Sci,
Int Lexonville, Que.
Mazumder SC (1976)
"Prediction of transmission loss
over trans horizon tropospheric
propagation paths in the Indian
sub continent” URSP Vol- 5
pp 152 -157
Meadow RW,
Lindgren RE, Samuel JC (1966)
"Measurement of m ultipath
propagation over a LOS radio link
at 4GC/S using frequency sweep
technique, proc, Inst.,
Elect.
Eng. 113(1), 41 - 48
Mogensen G (1977)
"Experimental investigation on
radio wave propagation in the
13.5-15 GHz frequency bands",
Rep. 2D- 30,203 pp. Electromagn.
Inst., Tech. Univ of Denmark.
-■es
105
Millingtan G,
I-Iewith R, Immerzi FS (1962)
Narayan Rao,
K Krishna Reddy, K S Ravi
SVB Rao, MJ Kesava Murthy,
SK Sarkar, HN Dutta .
BM Reddy. (1991)
Proc IEEE (USA)
16, 419.
"Effect on Fresnel Zone clearance
of propagation characteresties of
mw links in hilly terrain"
URSP
pg, 2 9 -3 8
Narayan Rao, K Krinshna Reddy
KS Ravi, SVB Rao,
MJK Murthy, HM Dutta,
SK Sarkar, MV SN Prasad,
BM Reddy. (1992)
"Microwave propagation studies
and correlation with SODAR
structrurees observed over a hilly
terrain in S. India" URSP
Vol. 21 p 329 - 333
Nelms GL (1963)
Can. J phys, 41, 202
Ott RH (1978)
Prop. Vol AP - 26 pp 329 - 337
Oyinloye J (1987, 79)
"Hie troposphere in tropical and
sub - tropical latitude"
handbook on radio prop, for trop.
& sub - trop countries.
(INSDOC, New Delhi)
Pari SA (1983)
"Characterization of multipath :
parameters for LOS mw prop."
IEEE Trans. Ant & prop.'
vol. AP - 31 vol 6,
Pickering LW,
De Rosa JK (1979)
"Refractive multipath model for
mw LOS link," IEEE Trans,
commun. vol COM - 27,
pp 1174 - 82,
i
Pieno VC (1956)
J geophys. Res, 61, 165.
Quian TP
Nisbeth JS (1965)
J geophys. Res, 70, 113
106
Rastogi RG,
Rajaraman G (1971)
Indian J. Pure and
Applied phys, 9, 531
Reddy BM (1987)
"Physics of the troposphere",
handbook on radio propagation for
tropical and sub - tropical countries
(INSDOC)
New Delhi, India 1987.
Robertsaw A (1986)
"Effective earth radius for
refraction of radio waves at
altitude above 1 km" IEEE Trans.
Ant & Prop. Vol AP - 34 No. - 9
pp 1099 - 1105.
Richter JH,
MV (1976)
"IREPS, A refractive effects Hitney
assesments system" present at 7th
Tech, exchange conf.
EL Pas TX Nov 30 - Dec 3
Rishbeth H (1964)
J. atmos. terr. phys, 26, 657.
Rurmnler WD (1979 b)
"Extension of the multipath fading
channel model", paper presented
at International conf. in comm.
Inst, of Electr. and Electron. Eng.
Boston.
Rpthruff CL (1971)
"Multipath fading on LOS mw
radio systems as a function of
path length and frequency"
Bell sys. Tech. J.vol. 50,
pp 2375 - 2398,
Sandberg J. (1978)
"Multipath parameters and
multipath propagation related to
PSK communications"
Rep 20 - 34 (2 Parts), 259 pp.
Electromagn. Inst. Tech.
Univ. of Denmark, Lyngby,
Denmark.
107
Sandberg J. (1980)
"Extraction of multipath parameters
from swept measurements on
a LOS Path" IEEE Trans Ant.
Prop AP - 28 (6), 743 - 750.
Sarkar SK,
Dutta HN, Reddy BM, (1983)
Proc. of Int. Conf of Ant. & Prop.
(UK), 219 - 294.
Sasaki 0 ,
Akiyama T (1977)
"Characteristics o f fade out and
fade spread in mw radio link" Rev
Elect. Commun.
Lai, 25 (3 - 4), 315 - 323.
Sasaki 0 ,
Akiyama T (1979)
"Multipath delay characterestics on
LOS microwave radio systems",
IEEE Trans, commun,
COM- 27 (12), 1876 - 1886.
Saxton JA,
Lane JA, Meadows RW,
Mathews PA (1964)
"Layer structure of the troposphere"
proc. Inst. Elect. Eng.
111(2), 275 - 283
Schiavone JA, (1982)
"Microwave radio meteorology :
fading by beam focussing".
IEEE Conf. Rec. vol - 3,
pp 78-1-5
Schiavone JA, (1983)
"Microwave radio meteorology,
*"'isonal fading distribution"
Radio sci. vol- 18 No.3,
pp. 369 - 380.
Schiavone JA,
Hermiller SM (1986)
"A regression model for forecasting
mw radio fading at Palmetto",
GA, IEEE Trans. Ant. & Prop,
vol AP - 34 No.7
Sen AK,
Karmakar PK, Das TK,
Rahman M, Devbarman S, (1991)
"A theoretical estimate of relative
contribution of refractivity forwards
group delay",
IJ RSP p 39 - 43.
Sengupta N,
Dasgupta MK (1983)
"A comparative study of mw
signal charaeterestics observed
over two LOS links"
IEEE Trans. Ant. & Prop,
vol AP -31 No.- 6.
a
Sharma S (1993)
"Study of scinlillation and fading
of mw signal with respect to
tropospheric irregularities”,
Ph D Thesis, Gauhati University.
Sharma S,
Timothy KI, Barabra AK,
Devi M (1998)
"An emperical model to estimate
RRI grad, from ground based data
sources" DRSP Vol - 27.
pp 11-14.
Sharma S,
Timothy KI, Barabra AK,
Devi M (1996)
"Dynamics of the lower atmosphere
through analysis of fade
charaeterestics of mw signal"
IJRSP Vol - 25. pp 187 - 196.
Shen XD,
Vilar E (1995)
"Anamalous trans horizon
propagation and metereological'
procesaes of a multi link path".
Radio sci vol - 30 No. - 5
pg 1467 - 1479
Singh,
Parthasarathy (1976)
"Effects of refractivity Index
gradient layers on tropospheric
links" DRSP vol - 5.
Singhal (1981)
Journal for industrial research
40, pp - 765 - 77
Smith EK,
Wientraub S (1953)
"The constants in the equation for
atmospheric refractive index at
radio frequencies" Proc
IEEE, 41, 1035 - 1037.
109
Taylor (1965)
Planet space sci, 13, 507.
Tiwari RK.,
BS (1989)
"Propagation studies over irrgular Jassa
terrain in Doon Valley."
URSP Vol -IS p 14 - 18.
Tripath S,
A.K. and Das J.(1992)
"Computer analysis of atmospheric De
plume structures"
URSP vol -21 pp 321 - 328.
Tyagi TR, Singh L,
Devi M, Borbara AK (1977)
DRSP,
16, 240, 244.
Wait JR (1962)
J. Res. natn. Bur. stand,
67, 617.
Webster AR,
AM (1987)
"A-O-A and tropospheric multipath Scott
microwve propagation"
IEEE Trans Ant. & Prop. (USA)
vol. AP - 35 No. 1, p 94- 9.
Webster AR (1982)
"Raypath parameters in tropospheric
multipath propagation"
IEEE Trans. Ant. & Prop,
vol AP - 30 No.- 3,
pp 796 - 800.
"A-O-A and delay times on
terrestrial LOS mw links"
IEEE Trans. Ant. & Prop.
Vol. A P -3 1 N o . 1. ;
• Webster AR (1983)
"Microwave A-O-A measurements Lam WI
under anamalous tropospheric
j
propagation conditions.!” paper
;
presented at commission F symp,
Union Radio, Sci, Int,,
Lennoxville, Que.
;i
Webster AR.,
(1980)
\
110
> I
.
! \
i "
■
i .
i •
1
Stephansen ET,
Mogensen GE (1979)
"Experimental investigation o f some
effects o f multipath propagation on
a LOS path at 14 GHz". IEEE
Trans, commun COM 27(3),
643 - 647.
Stephensen ET (1981)
"Clear air propagation on lineof-sight radio paths. A review.”
Radio sci vol. 16 No. 5
pg 609 - 629.
Stubbe P (1964)
J. atmos terr. phys, 26, 1055.
Swarap S.,
Tiwari PK (1974)
"Scale model technique for radio
wave propagation studies" URSP
Vol. 3, pp 21 - 27.
Sweet W (1980)
"Assessment/forcasting of
anamlous mw propagation in the
troposphere using model output"
Naval environ, predic. Res. faculty,
Monterey, CA,
NEPRE Tech Rep. TR 80 - 85.
Talukdar PH,
Devi M, Barbaraa AK (1979)
URSP 8, 318
Tatarski VI (1971)
"The effects o f turbulent
atmosphere wave propagation"
Nanka, Moscow.
Tattersall RLO,
NE (1977)
"Multiple propagation data Cartright
collected from test on LOS radio
paths at 20 and 37 GHz during the
period 1972, 1975, Rep - 594, 105
pp - Res, Brit, postoffice,
Martlesham Health,
Ispwich, England.
Ill
Wienstein AI (1980)
"Forcasting anomalous mw prop, in
the marine boundary, layer" paper
presented at 25th Joint Electron
Warfars conf. May 6 - 8 .
White RS (1970)
Space phys, Gardon & Preach
pp 61 - 64.
Yuen PC.,
Roelofs T.H. (1967)
J. atmos. terr. phys. 29, 321.
m
Документ
Категория
Без категории
Просмотров
0
Размер файла
4 859 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа