el%3A19941047

respectively, may suffer from sidelobe contamination. We note,
however, that these cells account for less than 20% of the population of greater than 400 range-azimuth cells measured in this
experiment. They are included in the reported database to accurately represent the median NRCS (50% level); their impact on the
mean NRCS estimate is expected to be less than IdB.
Summary: Large bistatic angle vertically polarised NRCS (propagation factors included) exceeds horizontally polarised NRCS by
-6dB for rolling forested hills at low grazing angles, with average
values of -36 and 4 2 d B , respectively. Defoliation of the deciduous portion of the canopy does not significantly change the
NRCS. Additional measurements are needed to verify these results
and investigate any possible dependence of NRCS on cell size and
bistatic scattering angle. These measurement results should be of
interest to system designers requiring estimates of bistatic clutter
levels.
tion bands will naturally provide a limited operational range.
Our propagation studies were carried out as the prelude to the
design of a 6OGHz secure digital microcellular communication
system in rural environments. The measurements were conducted
to determine the effect of natural obstacles obstructing the line of
sight (LOS) path.
Several configurations were tested with a wideband sounder
using a network analyser to measure the channel transfer function.
Channel model: In the studied radio propagation environment, the
signals are received via a scattering mechanism. Therefore, multiple propagation paths occur (multipath) with different time delays,
attenuations and phases. This channel is classically modelled as a
complex time-varying linear filter associated with the following
complex impulse response [I]:
T(t)
h(t.
.A(t, rt)cxp(ja(t,T,))J(t- ~ ~ ( t )(1)
)
=
t=O
Acknowledgment: This work was supported by the US Air Force
Rome Laboratory, Hanscom AFB, MA, under contract number
F19628-91-k-0047. The authors are grateful to M. Sowa for his
support and R. Chanley for his assistance in performing these
experiments.
where N ( t ) is the number of multipath components received, and
A(t,.r,), @(t,.rj) and .r,(r) are, respectively, the magnitude, phase,
and delay associated with the multipath component i.
22 July 1YY4
0 IEE 1994
Electronics Lerrers Online No: 19941024
D.J. McLaughlin, E. Boltniew, Y . Wu and R.S. Raghavan (ECE
Department, Northeastern University. Boston. M A 02115, USA)
References
BOREL. C.C., MCINTOSH. R E., NARAYANAN, R.M., and SWIFT, C T : ‘File
of normalized radar cross sections (FINRACS) - A computer
program for research and the scattering of radar signals of natural
surfaces’, IEEE Trans., 1986, GE-24, (6), pp. 1020-1022
HANLE. w., FARINA. A., and PELL, c.: ‘Bistaticimultistatic radar
bibliography’, IEE Proc. F, 1986, 133, (7), pp. 666668
WILLIS, H.J.: ‘Bistatic radar’ (Atrech House, Nonvood, MA, 1991)
HSU. Y.s., and LORTI, D.c.:
‘Spaceborne bistatic radar - An
overview’, IEE Proc. F, 1986, 133, (7), pp. 6 4 2 4 8
SOWA, M.J., and MCLAUGHLIN. D.J : ‘A bistatic radar test bed for
signal-to-clutter enhancement studies’. Proc. 1993 Bistatic Radar
Workshop, Griffiths AFB, Rome, NY, 1993
JACKSON, M.c.: ‘The geometry of bistatic radar systems’, IEE Proc.
F, 1986, 133, (7), pp. 604412
MOYER, L.R.,MORGAN. c J . , and RUGGER, D.A.:‘An exact expression
for resolution cell area in special case of bistatic radar systems’,
IEEE Trans., 1989, AES-25, (4), pp. 58&587
ULABY.
PT., MOORE.
R.K. and FuNG.A.K.:‘Microwave remote sensing,
active and passive, Volume 11’ (Addison-Wesley Publishing Co.,
Reading. MA, 1982)
MEEKS, M.L.: ‘Radar propagation at low altitudes’ (Artech House,
Dedham, MA, 1982)
Outdoor millimetre-wave propagation
measurements with line of sight obstructed
by natural elements
N. Daniele, D. Chagnot and C. Fort
1
Fig. 1 60GHz channel sounding syAtem
Channel measurement technique: The measurement system (Fig. 1)
is built up around an HP 8753B vector network analyser (VNA)
which performs the frequency domain measurements of the channel transfer function (&, determination) between 59 and 61 GHz.
The impulse response can be calculated via the inverse Fourier
transform of the frequency data [2].
At port 1 of the VNA the frequency is swept between 2 and
4GHz and is kept as the reference for further calculations. A
200111 flexible coaxial cable allows mobility of the autonomous upconverter module. This module converts the 2 4 G H z I F signal to
the 59-61GHz sounding frequency which is radiated through a
biconical antenna positioned 0.5m above the ground.
At the receiver the down-converter is identical to the up-converter; the 2 4 G H z I F signal is amplified and fed into the return
signal port 2 of the VNA which displays the channel transfer function. Both LO frequencies are derived from a unique IOOMHz
temperature compensated crystal oscillator (stability 5 x 10 *). A
phaselocked oscillator is used to generate a I9GHz signal, which
is tripled to give the 57GHz LO signal.
A personal computer equipped with a GPIB card controls the
acquisitions, and computes the correction for the line of sight path
loss (free space loss) and for the antenna gains; finally it calculates
statistical values from the data.
Indexing terms: Radiowuve propago tion, Multiparh
The authors present widehand measurements of outdoor radio
channels operating within a 2GHz frequency band around a
60GHz carrier. Statistical analyses such as the RMS delay spread
and cumulative distribution function (CDF) of the received signal
envelope are presented. Investigations into the multipath effect
and data rate limitation are discussed.
Description of environments and experiments: Our measurements
where carried out at two different sites:
The first one is a large area with a water canal, rows of trees,
some bushes and tall grass.
U
b The second one is a large area with different kinds of ground
Introduction: Radio transmissions such as wireless communications, remote control, surveillance and local cellular radio networks require limited operational range. The use of a millimetre
wave frequency which lies within one of the atmospheric absorp-
ELECTRONICS LETTERS
1st September 1994
(grass, land, pebbles) and various topographies (hillocks, pits,
trenches).
For these two sites 25 channel transfer functions were measured. In each case the transmitter was placed on both sides of a
Vol. 30
No. 18
1533
-.
particular medium, the antenna being fixed 0.5m above the
ground. The line of sight was always obstructed by the environment. The separation of the transmit and receive antennas lay
between 2 and 150 m. For each measurement three graphs were
recorded:
C D F and the type of RMS delay spread that characterises this
particular environment. The results show that propagation in this
environment is possible in spite of scattering and diffraction by
natural elements. Moreover, to avoid the undesirable effect of fading (which increases the bit error rate) a diversity technique can be
used. These initial measurements show that a local cellular network can be established in this particular medium.
* + + +
+
+
+
+
+
+
f
+
t
~
+
l
+
t
t
.
t
t
+
+
f
+ ~. *
+
+
+
+
t
+
t
~
.
+ +
+
+
+ +
0
+
a
t
I
*
+
~
.
+
+
+
l
+
+
+
t
+
t
f
+
t
t
+
*
t
+ *
+
f
+
t
t
t
f
+
+
*
+
t
+
+
+
* * * * * + * * +
+ + + * * + * + *
. + + + + * + + + +
.>&-d.
* * + +
1.
,
3
ti&,ns
Fig. 4 R M S delay spreud
Fig. 2 Module
of
a typiral impulse response
(i) the medium tested channel transfer function between 59 and
61GHz.
Acknowdedgment: This work has been supported by DRET from
Direction Generale de I’Armement. The authors would also like to
thank GlAT Industries for their contribution in the experiments
scenarios.
(ii) the impulse response after inverse Fourier transform of (i)
(Fig. 2); two important parameters are calculated and displayed
on this graph: TOA (time of arrival) and spread (RMS delay
spread).
0 IEE 1994
26 July I994
Electronics Letters Online No: I9941047
N. Daniele, D. Chagnot, and C. Fort (LETI (CEA-Technologies
Avancies) CEN.G. 17, rue des Martyrs-F38054 Grenoble Cedex 9, France)
(iii) the C D F of the received signal envelope (Fig. 3); the straight
line on the graph represents the well known Rayleigh distribution
to be compared with the C D F of the different channels recorded.
References
-50
-40
-30 -20
-10
0
amplitude .dB
IO
20
30
LO
Fig. 3 CDF example
Results: Scattering and diffraction were the main propagation phenomena giving rise to multipath in these environments. For the
tested medium, reflection phenomena rarely occurred except in the
case of metallic elements near the link. In the tested environment
the vegetation loss difference is independent of the polarisation
(vertical or horizontal). Other results [3] confirm this phenomenon.
Sfatistical results: For 83% of the recorded channel transfer functions the cumulative distribution function of the amplitude components fits the log normal-Rayleigh distribution. This type of
distribution is thus a good representation of the effect o f scattering
generated by vegetation elements in obstructed sight (OBS) conditions. For 87% of the recorded cases the fades observed on the
received signal envelope are less than 20dB.
Fig. 4 shows the cumulative distribution of the RMS delay
spread for all the performed measurements. For 88% of cases the
RMS delay spread is less than 100ns.
Discussion/conclusion: Propagation measurements were performed
in a rural environment with natural elements obstructing the LOS.
The purpose of these experiments was to investigate the type of
1534
DAVIES, R., BENSEBTI. M., BEACH. M.A., and MCGEEHAN, J.P.: ‘Wireless
propagation measurements in indoor multipath environments at
I.7GHz and 60GHz for small cell systems’. 41st IEEE Vehicular
Technol. Conf., Gateway to the future, Technology in Motion, 1922 May 1991, (St. Louis, MO, USA), pp. 589-593
2 HOWARD, J.S., and PAHLAVAN, K.: ‘Autoregressive modelling of wideband indoor radio propagation’, IEEE Trans., 1992, COM-40, (9),
pp. 1540-1552
3 VIOLETTE, E.I., ESPELAND, R.H., and MITZ, A.R.: ‘SHF-EHF
propagation through vegetation on Colorado East slope’. Tech.
Report CECOM-81-CS020-F,June 1981
1
Demonstration of direct bonding between
InP and gadolinium gallium garnet
(Gd3Ga5OI2)substrates
M. Totoki, T. Mizumoto, K. Maru and Y. Naito
Indesing terms: Indium phosphide. Garnets
The authors demonstrate the direct bonding of InP/Gd,Ga,O,,
(GGG) and InPGaInSb on GGG without glue. After chemical
treatment, heat treatment was applied in an H, atmosphere
without a weight. This technique makes feasible the integration of
semiconductor and magneto-optic devices.
Introduction: To realise photonic integrated circuits, a technology
that is able to integrate different materials needs to be developed.
Although individual photonic devices may have satisfactory performance characteristics, the integrating technology faces many
problems when the physical properties of the constituent materials
are very different, especially in case of materials with highly mismatched lattice constants. Most photonic devices, including active
and passive components, such as optical switches and optical
multi/demultiplexers, can he fabricated by using 111-V compound
ELECTRONICS LETTERS
__-
1st September 7994
Vol. 30
No. 18
- -