close

Вход

Забыли?

вход по аккаунту

?

el%3A19890311

код для вставкиСкачать
to 18GHz. To this end, a microstrip test fixture using K connectors has been developed with less than 0.7 dB/cm of insertion loss at 40GHz. A gain of 6 f 1 dB has been obtained
from 2 to 42GHz. These results are very close to the simulations. Less than 5.2 dB for the noise figure up to 18 G H z has
also been measured. These amplifiers exhibit the highest real
estate efficiency ever reported (3dB/mm2) in any monolithic
HEMT amplifiers.
0
5
10
15
20
25
30
35
LO
45
frequency, GHz
Fig. 5 Gain and matching of H E M T distributed amplifier
V, = 8.5V, V,,l = I/bS2= -0.1 V
. . . . . . . . calculated
measured
~
Conclusion: With a suitable design, wideband amplifiers with
a bandwidth larger than the cutoff frequency of the devices
can be obtained.
PROPAGATION RESULTS AT 11 GHz FOR
MICROCELLULAR RADIO
Indexing terms: Radiowave propagation,
systems, Radio links, Radiocommunication
Mobile
radio
Given the limited spectrum allocated to cellular mobile
radio, we consider the use of frequency bands where larger
bandwidth allocations may be feasible. A radio network of
‘microcells’has been proposed for such bands. We have conducted a propagation experiment at 11GHz to characterise
such communication channels in various environments.
Results are given here for rural and semiurban environments,
where propagation is found to be dominated by interference
between the direct ray and a roadway-reflected ray.
Introduction: A major goal of current efforts in cellular mobile
radio communications is to make portable telephone service
universally available with high quality at affordable cost.
Attaining the goal of universal availability may pose particular problems because the spectrum currently allocated to
cellular radio is not likely to be expanded. Accordingly, we
consider the alternative approach of going to microwave frequency bands, where larger spectrum allocations might be
feasible.
The physical architecture we assume is one consisting of
radio ‘microcells’.’~2That is, portable users (either in vehicles
or on foot) communicate via line-of-sight (LOS) or near lineof-sight paths with fixed compact canisters (base sites or
microsites). These canisters can be placed at relatively low
heights (e.g. lamp posts, etc.), with spacings of up to 600m,
along roadways or in other public areas.
The use of microcells is prompted by a number of factors.
For one thing, communication at microwave frequencies generally requires a relatively unobstructed portable-to-base site
path (a conjecture supported by our data). Moreover, the use
of microcells would facilitate the attainment of universal
coverage in high-density urban areas.
Experiment: We have measured 11 GHz propagation in a
10
11
12
13
1L
15
frequency, GHz
16
17
microcell environment. The experiment consisted primarily of
vehicle runs of 3 W 6 0 0 m along a road. We used CW transmissions, and recorded received-signal envelope power levels
(no phase). The antennas used were 1/4 vertically polarised
monopoles on ground planes.
Most measurements were made using LOS propagation
paths along a straight road (Fig. 1). At these locations, the
base site vehicle was parked at the roadside adjacent to the
traffic lane. Test sites were chosen where the mobile vehicle
could traverse a path of about 300 m on either side of the base
site.
18
j380/61
Fig. 6 Noisefigure of H E M T distributed amplifier
V,, = 8.5V, Vgsl = llgr2 = -0.1 V
This is demonstrated by monolithic millimetre-wave amplifiers using MOVPE 0.5pm HEMTs. The results show a real
estate efficiency of 3 dB/mm2 at 40 GHz, which represents the
state-of-the-art concerning monolithic distributed amplifiers,
and show that MMICs using heterostructure devices are very
promising for millimetre-wave applications.
Acknowledgment: We thank J. P. Andre and A. Briere for
supplying the MOCVD heterostructures, and P. Chambery
and M. Wolny for their work on the active devices.
P. GAMAND
8th February 1989
A. DESWARTE
J.-C. MEUNIER
I
I
k
I
I
I
X
-
l
I
I
X
1388’11
Fig. 1 Measurement geometry
x, and x I 2 300111
Our measured data reveal an interesting and important
dual-ray propagation mechanism for microcells which can be
explained by a plane earth model3
LEP-Laboratories d’Electronique et de Physique appliquee
3 avenue Descartes, 94450 Limeil-Brevannes Cedex, France
References
1 SCHINDLER, M. I., WENDLER, J. P., MORRIS,A. M., and LAMARRE, K.
I
XS
A.:
‘A 15 to 45GHz distributed amplifier using 3 FETs of varying
periphery’. GaAs IC symp., Oct. 1986, pp. 67-70
2 YUEN, c . , NISHIMOTO, c . , GLENN, M., PAO, Y. c., BANDY, s., and
ZDASINK, G . : ‘A monolithic 3 to 40GHz HEMT distributed amplifier’. GaAs IC symp., Nov. 1988, pp. 105-108
ELECTRONICS LETTERS 30th March 1989 Vol. 25 NO. 7
Rural area-macrovariations for vertical polarisation: Figs. 2
and 3 show the measured (solid curves) received power. The
mobile antenna height, h,, was fixed at 1.8m. The base
antenna heights h, were at 3 m and 9m. In obtaining these
records, spatial averaging over a 0.4 m window was performed
on the data to smooth out the fluctuations due to random
scattering. These records are thus macrovariations.
453
Note the regular notch patterns of the measured data in
Figs. 2 and 3. The spatial frequency of the notches decreases
with increased mobile-to-base distance. Also note that notch
separation increases with decreased antenna height.
In Fig. 5, we show the worst case microvariations of the
received power at an interval starting at x 1 and ending at
x1 + 6 m. A spatial averaging interval of length 1/10 was used.
A random component, due to multiple scattering, exists and
becomes significant at power levels of - 50 to - 40 dB. In the
rural environment, the microvariations at these relative power
levels were negligible. Nonetheless, over the major portion of
the semiurban path, the random component is insignificant.
We can also discern a regular notch pattern which varies
more rapidly than does the two-ray pattern plotted in Fig. 4.
-2
ok.fT
&. -40
-50
projected distance x , m
-60
Fig, 2 Monopole-to-monopole measured and calculated received mean
power against projected distance x
Rural area, h, = 3 m, h, = 1.8 m
measured
---
E
-70
J
-80
I
calculated
Agreement with theory: The dash-dot curves in Figs. 2 and
J
/
/
/
0
3
represent the calculated received power in a dual-ray
environment, with the proper radiation patterns of the
antennas included. The vertical separation between these and
the measured curves was chosen to enhance clarity. Note the
striking similarity between the calculated and measured
results in all these Figures. The regularity of the pattern
attests to the dominance of dual-ray fading.
,
l
l
(
l
l
l
/
l
l
l
l
l
-300
projected distance x,m
l
I
i
I
-
150
rn
Fig. 4 Monopole-to-monopole measured and calculated received mean
power against projected distance x
Semiurban area, h, = 9 m, h, = 1.8 m
measured
---
calculated
In all our measurements we found that the average received
power does not decay with base-to-mobile distance r any
faster than l/r2.
e
-2
- l0o t
3 -3 0
(U
-50
2
!?
-60
-70
I
x,+6
I
x, + 3
projected distancex ,m
XI
L388151
Fig. 5 Measured microvariations of received power against projected
distance x
Monopole-to-monopole measurements in semiurban area
projected distance x, m
Fig. 3 Monopole-to-monopole measured and calculated received mean
power against projected distance x
Rural area, h, = 9m, h, = 1.8m
~
measured
calculated
The locations of some of the measured notches in Figs. 2
and 3 do not exactly match the theoretical ones. This is owing
to the fact that the roads are not perfectly flat. It can be
shown that a small undulation of the road can easily shift the
location of the notches, as we observed.
Semiurban area macrovariations f o r vertical polarisation: The
solid curve in Fig. 4 represents the measured received power
with h, = 9 m and h, = 1%m. A 0.8m averaging window was
used. The dash-dotted curve in this Figure is for the dual-ray
model, including the monopoles radiation pattern. Despite the
jaggedness of the experimental curve, we can discern a regular
pattern of notches corresponding to dual-ray interference
similar to the rural location. The jaggedness of the curve, as
well as the relatively shallow notches, indicate that there may
be more than just one reflected ray playing a role here. In the
regions of the deep notches, one reflected ray predominates
over all the other reflected rays.
454
Conclusions: In a microcell environment we have found that:
propagation is dominated by interference between the direct
LOS ray and a specular roadway-reflected ray; the random
(Rayleigh-distributed) components can become significant
with omnidirectional antennas, but only at low signal levels,
i.e. when destructive interference between two dominant rays
is present; the average received power does not decay faster
than l / r 2 with base-to-mobile distance.
A. J. RUSTAKO
N. AMITAY
G. J. OWENS
R. S. ROMAN
9th February 1989
AT&T Bell Laboratories
Crawford Hill Laboratory
Holmdel, N J 07733, U S A
References
and PRABHU, v. K . : 'Higher user density digital cellular
mobile radio systems', Proc. IEE F, 1985,132, (5), pp. 3 9 W 0 4
2 CHIA, s. T., STEELE, R., GREEN, E., and BARAN, A.: 'Propagation and
bit error ratio measurements for a microcellular system', J . Inst.
Radio Electron. Eng., 1987,57, (6) (supplement), pp. S255-S266
3 JAKES, w. c . , JR. (Ed.): 'Microwave mobile communications' (New
York, John Wiley & Sons, 1974)
1
STEELE, R.,
ELECTRONICS LETTERS 30th March 1989 Vol. 25 No. 7
Документ
Категория
Без категории
Просмотров
2
Размер файла
212 Кб
Теги
3a19890311
1/--страниц
Пожаловаться на содержимое документа