Middleton W.M. (ed.) Reference Data for Engineers: Radio, Electronics, Computer and Communications
Подождите немного. Документ загружается.
34-6
REFERENCE
DATA
FOR ENGINEERS
FREQUENCV
IN
MEGAHERTZ
Fig.
5.
Cosmic
noise
levels for
a
half-wave-dipole receiving antenna.
MAN-MADE RADIO NOISE
Man-made radio noise frequently limits the perform-
ance of receivers. This is particularly true for land-
mobile communications, television reception, high-
frequency radio, and other radio services below UHF.
Man-made radio noise originates from a wide variety
of
sources; some examples are noise from ignition systems
of gasoline engines, corona noise from high-voltage
power lines, gap noise from utility distribution lines,
noise from radio-frequency stabilized welders, and
noise produced by many other electrical devices found
in homes and businesses. Fig.
6
shows the frequency
range that is affected by several common types of
man-made noise. Sources that affect the performance of
low-frequency receivers generally do not produce harm-
ful noise at higher frequencies. The reverse is also true
in that sources that produce harmful noise at VHF and
UHF generally do not produce harmful noise at low
frequencies. Since noise amplitude decreases with dis-
tance from the source, the magnitude of noise at a
receiver site is dependent on spacial parameters, tem-
poral variations of the noise source, directional proper-
ties of the noise-radiating elements, and the directional
properties of the receiving antenna. Sometimes harmful
noise is inductively coupled or conducted from its
source into a receiver.
The average levels of man-made noise are higher in
urban and suburban areas than in rural areas because of
the larger number of sources in areas of higher popula-
tion. Fig.
7
shows median values of average noise power
in urban and suburban areas. In remote quiet locations,
man-made noise can be extremely low, and background
noise in most parts of the radio spectrum may be
determined by natural noise (atmospheric noise below
about
20
MHz and galactic noise above about
20
MHz).
In general, man-made noise levels decrease with in-
creasing frequency, although a specific source may not
comply with this general rule.
Near-Zone and Far-Zone Noise
Sources
Most cases of harmful man-made noise affecting a
receiver involve only one or two sources.
A
general
background level of man-made noise from numerous
sources is seldom encountered. Frequently, the sources
of harmful noise are relatively close to a receiver. The
propagation from the noise source to the receiver
sometimes involves the near-zone field rather than the
far-zone field of the source. For such cases, both the
electric- and magnetic-field components of the noise
source must be measured in order to define fully the
ability of noise to affect receivers. Table 1 provides a
general set of rules for the measurement of noise. When
the path from the noise source to the receiver is greater
than one wavelength, the far-field approximation ap-
TABLE
1.
GENERAL
RULE3
FOR
MEASUREMENT
OF
NOISE
Distance
Source
to
Sensor
Zone
Measure
Greater
than
A
Distant Either
E
or
B
Less
than
h/6
Inductive
Must
measure
both
E
and
B
Between
Mixed
Recommend
A16
and
A
E
and
B
RADIO NOISE
AND
INTERFERENCE
34-7
SUBFUNDAMENTAL
1
I
---
ATMOSPHERICS
-
-
-
-
-
_-
-
-
-
-
IGNITION
I
CORONA
.
.---
GAP
-----.
IMPULSES SYNCHRONOUS WITH POWER SYSTEM
I-----------
-
HARMONICS
OF
POWER SYSTEM
-
OF
POWER SYSTEM
3x
108
I
107
Ix
106
3x105
2
X
z
3x104
;.
x
103
-0
=
290
K
x
10
FREQUENCY
IN
MEGAHERTZ
Fig.
7.
Median values of average noise power expected from various sources (omnidirectional antenna near surface).
plies, and the noise can be measured with either an
electric-field or a magnetic-field antenna. The electric-
and magnetic-field strengths can be related by the
free-space impedance
of
120~
=
377 ohms. This
applies to measurements
of
man-made noise at frequen-
cies above about 30
MHz.
At lower frequencies, the
noise source may be electrically close to the receiving
antenna. At distances from the source to the affected
receiver
of
one sixth of a wavelength
or
less, i.e.,
<h/2rr,
the receiver is in the near field of the source. At
these short distances, the ratio of
E
to
H
is no longer
equal to 377 ohms, and both the electric and magnetic
fields must be measured in order to define the impact of
a noise source on a receiver.
Power
Line
Noise
The temporal and spectral properties of man-made
noise vary considerably from one type
of
source to
another. Noise associated with a power line usually
34-8
REFERENCE DATA FOR ENGINEERS
contains bursts of noise at intervals determined by the
fundamental frequency of the
power
line. The temporal
properties of gap noise (formed by the minute electrical
breakdown between tw pieces of metallic hardware
exposed to a strong electric field) are shown in Fig.
8.
Multiple bursts of noise occur each time the power-line
voltage reaches a maximum. Individual impulses within
a burst are spaced very close together
(less
than
1
ms
apart), while the bursts
are
spaced
8.33
ms apart for a
power
line operating at
60
Hz.
Individual bursts have
the same amplitude.
A
careful inspection
of
the
wave-
form in Fig.
8
shows that two gap-noise sources
are
present in the view.
PRECIPITATION STATIC
Precipitation static
is
produced by rain, hail, snow,
or
dust storms in the vicinity of the receiving antenna
and is important chiefly at frequencies below
10
MHz.
This form of interference
can
be
reduced by eliminating
sharp points from the antenna and its surroundings, and
also by providing means for dissipating the charges that
build up on an antenna and on its surroundings during
electrical storms.
-74
E
0
-44
I
n
w
3
-114
2
114
z
Fig.
8.
Temporal
structure
of
gap
noise
from
138-kV
trans-
mission line.
THERMAL NOISE
CALCULATIONS
Thermal noise is caused by the thermal agitation
of
electrons in resistances. Let
R
be
the resistive compo-
nent in ohms
of
an impedance
Z.
The mean-square
value of thermal-noise voltage is given
by
E2
=
4RkT
Af
whm,
k
is Boltzmann's constant
(1.38
X
T
is the absolute temperature in kelvins,
Af
is the bandwidth in hertz,
E
is the root-mean-square noise voltage.
jouledkelvin)
,
The equation given
above
assumes that thermal noise
has a uniform distribution
of
power
through the band-
width
Af.
In case
two
impedances
ZI
and
Z2
with resistive
components
RI
and
R2
are in
series
at the same
temperature, the square of the resulting root-mean-
square voltage is the sum of the squares of the root-
mean-square noise voltages generated in
ZI
and
Z2:
E2
=
El2
+
E;
=
~(RI
+
R2)kT
*
Af
In
case
the same impedances are in parallel at the
same temperature, the resulting impedance
Z
is calcu-
lated
as
is usually done for alternatingcurrent circuits,
and the resistive component
R
of
Z
is then determined.
The root-mean-square noise voltage is the same
as
it
wuld be
for
a pure resistance
R.
It is customary in temperate climates to assign to
T
a
value such that
1.38T
=
400,
corresponding to about
17
degrees Celsius
or
63
degrees Fahrenheit. Then
E'
=
1.6
x
R
AJ
NOISE MEASUREMENTS
Measurement
for
Broadcast
Receivers
For standard broadcast receivers the noise properties
are
determined by means of the equivalent noise side-
band input (ensi). The receiver
is
connected
as
shown in
Fig.
9.
Components of the standard dummy antenna
are
CI
=
200
picofarads,
C2
=
400
picofarads,
L
=
20
microhenrys, and
R
=
400
ohms.
The equivalent noise sideband input is
(ensi)
=
mE,(P',/P',)1'2
where,
E,
=
root-mean-square unmodulated carrier-input
m
=
degree of modulation of signal carrier at
400
voltage,
hertz,
RADIO NOISE AND INTERFERENCE
34-9
STANDARD
SIGNAL
GEhERATOR
STAYDARD
DUMMY
ANTENNA
BROADCAST
RECEIVER
UNDER TEST
OUTPUT
GROUND
Fig.
9.
Measurement
of
equivalent noise sideband
input
of
a
broadcast receiver.
P
Is
=
root-mean-square signal-power output when
P
In
=
root-mean-square noise-power output when
It is assumed that no appreciable noise is transferred
from the signal generator to the receiver, and that
m
is
small enough for the receiver to operate without distor-
tion.
signal is applied.
signal input is reduced to zero.
Noise Factor
of
a Receiver
A
more precise evaluation of the quality of a receiver
as far as noise is concerned
is
obtained by means of its
noise factor.
It should be clearly realized that the noise factor
evaluates only the linear part of the receiver, Le., up to
the demodulator.
The equipment used for measuring noise factor is
shown
in
Fig.
10.
The incoming signal (applied to the
receiver)
is
replaced by an unmodulated signal generator
with
Ro
=
internal resistive component,
Ei
=
root-
mean-square open-circuit carrier voltage, and
E,
=
root-mean-square open-circuit noise
in signal generator. Then
E:
=
4kToRoAf‘
voltage produced
where,
k
is Boltzmann’s constant
(1.38
X
To
is the temperature in kelvins,
JIK),
SIGNAL RECEIVER INDICATOR CALIBRATED
GENERATOR UNDLR
TEST
TO READ RF POWER
Fig.
10.
Measurement
of
the noise factor
of
a
receiver. The
receiver
is
considered
as a
4-terminal network. Output refers
to
last
intermediate-frequency stage.
Af
is
the effective bandwidth of the receiver
If the receiver does not include any other source of
noise, the ratio
E:/E:
is equal to the power carrier/
noise ratio measured by the indicator:
(determined as below).
E:lE:
=
(E~14Ro)lk~Af’
=
SINi
The quantities
E;/4Ro
and
kToAf’
are called the
available
carrier and noise powers, respectively.
The output carrierinoise power ratio measured in a
resistance
R
may be considered as the ratio of an
available carrier-output power
Po
to an available noise-
output power
No.
The noise factor,
F,
of the receiver is defined by
Po/No
=
F1(PiINj)
F
=
(No/Ni)(PolPi)-l
=
E;l ,/4kToRoAf’
=
PJI
IlkToAf’
where,
Polpi
=
available gain
G
of the receiver,
PLlI1
=
available power from the generator
required to produce a carrier-to-noise ratio
of one at the receiver output.
Noise figure is the noise factor expressed in decibels:
FdB
=
10
log,#
Effective bandwidth
Af’
of the receiver is
where
Gf
is the differential available gain. Generally,
Af’
is approximated to the bandwidth of the receiver
between those points of the response showing a 3-dB
attenuation with respect to the center frequency.
Measurement
of
Noise Figure
With
a Thermal Noise Source
For the case where the spurious responses of the
receiver are negligible, receiver noise figure can be
conveniently measured by using the noise output of a
thermal noise source having an equivalent generator
resistance equal to that specified for use with the
receiver.
With the noise source off, but still possessing the
correct output resistance, receiver gain is adjusted for a
convenient amount of noise power output; then with the
noise source on, and still possessing the correct output
resistance, the noise power output is increased by a
34-
10
REFERENCE
DATA
FOR ENGINEERS
convenient power ratio
(N2/Nl).
The measured noise
figure is then given by
NF
=
(excess)dB
-
10
10g[(N2/N1)
-
11
For a thermal diode operating in the temperature
limited emission mode
(excess)dB
=
10
log(20RdZd)
where,
Rd
is the noise source output resistance,
Id
is the diode current in amperes.
When the receiver has appreciable spurious respons-
es, the correction factor that must be used with the
above simple equation is a complex function of the
spurious response ratios, and of the percentage of total
internal receiver noise produced by the circuits preced-
ing the mixer causing the spurious responses. For the
simple case of no preselection and a diode mixer having
negligible excess noise,
3
dB must be added to the
measured noise figure to obtain the true noise figure.
A
thermal noise source designed for a given generator
impedance
R,
can be used to measure the noise figure
of a receiver designed for a higher generator
R2
by
adding a resistor
(R2
-
Rl)
between noise source and
receiver input and using
Conversion of receiver noise temperature to noise
factor:
where,
TR
=
receiver noise temperature in kelvins,
F
=
noise factor of receiver (power ratio).
Conversely,
To
=
290
K,
Determination
of
effective noise temperature
of
re-
ceiving system (Le., antenna, transmission line, and
receiver):
TE
=
TA
+
(LF
-
1)To
where,
TE
=
effective noise temperature
of
receiving
TA
=
antenna noise temperature,
L
=
transmission line loss (power ratio),
F
=
noise factor of receiver (power ratio),
system,
To
=
290
K.
Determination of the effective input noise power of
the receiving system:
Ni
=
kBTE
where,
N,
=
effective input noise power of the receiving
system,
k
=
Boltzmann’s constant
(1.38
X
joules/kelvin),
B
=
bandwidth in hertz,
TE
=
effective noise temperature in Kelvins.
dBm,
=
-198.6
+
10
lo@
+
10
logT‘
Calculation
of
Noise Figure
The active device can be defined for noise-figure
calculations as in Fig.
11.
The value of
Re,
can be obtained experimentally by
measuring, with a “zero impedance” generator, the
equivalent microvolts
of
noise
V,,
,
in a bandwidth
B,
in
series with the input terminals, with an almost-short-
circuit
on
the output terminals. Then
Re,
is given by
Re,
=
l&c/2/1.64
X
lO-”<BW>
(Eq.
3)
The value of
Re
is
obtained straightforwardly by
input impedance measurements with a short-circuit on
the output terminals.
The value of
p
can be obtained experimentally by
approximately open-circuiting the input terminals at the
frequency of interest with a tuned circuit of parallel
resonant resistance
R,,
and measuring the total equiva-
lent microvolts of noise produced across the input
terminals, with an almost-short-circuit on the output
terminals. Then, assuming negligible correlation
When the above characterized device is used with an
input transforming circuit of parallel resonant resistance
R,,
the resulting noise factor can be calculated as
follows: First calculate
R1
and
/3
from
CUR
R E
h
T
VOLTAGE
GENERATOR GENERATOR
~p14kTlBMilG,l)’’Z
[4kTIBCC)Req]’
z
T‘flv
NOISE~FREE
GAIN
Fig.
11.
Calculation
of
the
noise
figure
of
a
receiver
RADIO NOISE AND INTERFERENCE
34-11
In terms of the above quantities and the transformed
generator resistance R, seen by the input terminals of
the active device, the resulting noise factor is given by
F
=
1
+
2(R,,/R,)
+
(Req/R,)
+
(R,/R,)
X
[P
+
(Req/Rl)I
0%.
7)
It should be noted that to minimize noise figure the
input circuit should always be tuned
so
as
to
null any
part of the noise due to
PR,
,
which is correlated with
the noise due to
Req.
Equation
7
can be applied to this
best noise figure tuning case if
p
is obtained, by some
method, from only the uncorrelated part of the
PR,
noise.
This resulting noise figure is minimized when the
transformed generator resistance has the value
R,
opt
=
{(RiReq/PY[l
+
(Req/PRi)1}1’2
0%.
8)
and with this optimum source resistance
Fop,
=
1
+
2P(Re,/R1)I”
X
{[l
+
(Req/Rl)11’2
+
(Req/R1)l’*}
(Eq.
9)
Noise Factor
of
Cascaded
Networks
The overall noise factor of two networks,
a
and
b,
in
cascade (Fig. 12)
is
Fob
=
Fa
+
[(Fb
-
l)/Gal
provided
Ah
’
5
AA
’
.
The additional noise due to external sources influenc-
ing real antennas (such as galactic noise) may be
accounted for by an apparent antenna temperature,
bringing the available noise-power input to
kTaA
f’
instead of
N,
=
kToAf’
(the physical antenna resistance
at temperature
To
is
generally negligible in high-fre-
quency systems). The internal noise sources contribute
(F
-
1)N,
as before,
so
that the new noise factor is given
by
NOISE
NOISE
FIGURE
=
F,
FIGURE
=
Fb
AVAILABLE AVAILABLE
GAIN
=
G,
GAIN
=
Go
NETWORK
~1
NETWORK
b
Fig.
12.
Overall
noise
figure
Fob
of
two
networks,
a
and
b,
in
cascade.
F‘Nj
=
(F
-
1)Ni
+
kToAf’
F’
=
(F
-
1)
+
(T,/To)
The average temperature of the antenna for a 6-MHz
equipment is found to be
3000
kelvins, approximately.
The contribution of external sources is thus of the order
of
10,
compared with a value of
(F
-
1)
equal to
1
or 2,
and becomes the limiting factor of reception. At 3000
MHz, however, values of
Tu
may fall below
To.
INTERFERENCE FROM
SIGNALS
OF
OTHER SERVICES
Although the usage of the radio spectrum is subject to
the detailed system of regulations described in Chapter
1,
there are many circumstances in which a service
operating in compliance with the radio regulations can
cause interference to another service. This most often
results from sharing of an allocated band by more than
one service. For example, the Fixed service (communi-
cations between fixed stations on the Earth) often shares
a band with the Mobile service (communications be-
tween vehicles, sometimes including a fixed station).
These two services also share some bands with a
satellite service. Even if the systems have been designed
to avoid interference most of the time, it is difficult to
prevent interference when, for example, a nongeosta-
tionary satellite moving close to the horizon passes
through the horizontally pointing beam of an antenna of
the Fixed service, or when a mobile station is close to
such an antenna. Interference can also be caused by
transmitters operating in other bands, if out-of-band
transmissions such as harmonics are radiated. The
extensive out-of-band radiation that results from unfil-
tered transmitters using the direct sequence form of
spread-spectrum modulation can cause harmful inter-
ference to very sensitive systems such as those used in
radio astronomy, even though such side lobes are low
enough to
go
undetected by most other services.
For any particular situation, it is possible to deter-
mine by calculation whether harmful interference will
occur. First it is necessary
to
know the EIRP (effective
isotropically radiated power) of the particular transmit-
ter in the direction of the victim receiver. If the
direction does not fall within the main beam of the
transmitter, one may need to know the radiation pattern
of the antenna. In the case of large parabolic antennas,
the reference radiation pattern recommended by the
CCIR”
may
be useful. This model applies
to
large
paraboloids of diameter greater than
100
wavelengths in
the frequency range
2
to
30
GHz.
The envelope of the
side-lobe pattern corresponds to a gain, in dB relative to
an isotropic radiator, of 32
-
25
log
4,
where
4
is the
direction in degrees measured from the axis of the main
beam. This relationship applies for values
of
4
from 1”
-
*
Reference
1.
34-
12
to 47.9’. For greater values of
4,
the gain is
-
10 dBi.
For other types of antennas, if specific details of the
power pattern are not available, it may be necessary to
use a “worst-case” estimate based on general informa-
tion about the type of antenna. After the transmitted
EIRP has been estimated, the next step is to find the
power flux density (PFD) at the receiver location. This
requires determination of the propagation
loss,
for
which methods are described in Chapter 33. For
free-space propagation the situation is simple, but
propagation between two widely separated points on the
surface of the Earth may involve diffraction around a
spherical model earth, diffraction over one or more
knife-edge ridges, or tropospheric scatter. Curves and
formulae for such cases are given in Chapter 33. From
the estimated PFD, the power received in the antenna of
the victim receiver can be calculated if the gain of the
antenna in the direction of the transmitter is known. As
in the case of the transmitting antenna, this may involve
an estimation of the side-lobe gain. Finally, it is
necessary to compare the received power with the power
level that corresponds to the threshold of harmful
interference at the input of the particular receiving
system involved. This harmful threshold can be calcu-
lated from the theory of operation of the receiver, or
determined from laboratory measurements on the re-
ceiver. The harmful threshold is a measure of the
vulnerability of a system to interference. The most
vulnerable service is radio astronomy (see for example
reference 2) followed by activities such as Earth explo-
ration by passive sensing and reception of information
from spacecraft at deep-space ranges. At the other end
of the vulnerability scale, systems using spread spec-
trum techniques or digital modulation with error-
correcting codes can function in the presence of inter-
fering signals as much as
100
dB greater than those
harmful to radio astronomy.
The extensive literature of the CCIR,* in the form of
Reports and Recommendations of the various CCIR
study groups, contains many studies of interservice
interference, sharing of frequency bands, and related
topics. This material provides technical input to the
conferences at which the international radio regulations
are revised, and is a highly authoritative source of
reference.
To illustrate the consequences of overlaying an active-
ly
transmitting service on a sensitive passive service,
assume a typical spread-spectrum personal communica-
tion system, radiating
0.1
watt isotropically over a
10-megahertz band centered at about 2.0 gigahertz. It
would produce an interfering power density of the order
of
3
X
wattihertz in the far-out side-lobes of a
receiving antenna one kilometer away. This would
represent an increase in the system temperature of the
service receiving interference of 20
000
kelvins. As a
-
*
Reference
3.
result, the typical performance of a modern communi-
cation system would be degraded by a factor of
200.
Such a penalty would reduce the maximum distance for
communication with a spacecraft by a factor of
14.
It
would reduce the volume of the Universe from which an
astronomical radio telescope could receive useful data
by a factor of
200.
Terrestrial services would suffer the
same type of degradation from overlaid spread-
spectrum services. Further, in addition to this interfer-
ence to the overlaid services. the “direct-sequence”
type of spread-spectrum system typically emits interfer-
ence well outside the band it utilizes for communication
unless well designed bandpass filters are employed at
the output
of
the transmitter.
REFERENCES
1.
CCIR Recommendation
509-1,
Recommendations
of
the CCIR,
Vol. 2. Geneva: Int. Telecommunication
Union, 1990.
2.
Thompson, A. R., Moran, J. M., and Swenson,
G.
W., Jr.
Interferometry and Synthesis in Radio
Astronomy.
New York: John Wiley and Sons, 1986,
and Melbourne, FL: Krieger Press, 1991, Ch. 14.
3.
Recommendations of the CUR,
Vols. I-XI1 and
Annexed Reports. Geneva: Int. Telecommunication
Union, 1990.
General References
Morrison, Ralph.
Noise and Other Interfering Signals.
New York: John Wiley and Sons, 1992.
White, Donald R.
J.
A
Handbookseries on Elecromag-
netic Interference and Compatibility.
German-
town, MD: Don White Consultants, Inc.
Vol.
I.
Electrical Noise and EMI Spec$-
cations,
I97
1
.
Vol.
11.
Electromagnetic Interference Test
Methods and Procedures,
1973.
Vol. 111.
Electromagnetic Compatibility
Control Methods and Techniques.
1973.
Vol. IV.
Electromagnetic Interference
Test
Instruments and
Systems.
197 1.
Vol. V.
Electromagnetic Interference Pre-
diction Techniques,
1972.
Crane, Patrick
C.,
and Hillenbrand,
L.
A. “Estimating
Harmful Levels of Radio-Frequency Radiation,”
in Crawford, David
L.,
ed..
Light Pollution, Radio
Interference and Space Debris,
Astronomical
So-
ciety of the Pacific Conference Series,
Vol.
17.
International Astronomical Union Colloquium
Pakala,
W.
E.,
and Chartier, V. L. “Radio Noise
NO.
112, 1990, pp. 258-266.
RADIO NOISE
AND
INTERFERENCE
34-13
Measurements
on
Overhead Power Lines from
2.4
to
800
kV.
IEEE
Trans on Power Apparatus and
Systems,
PAS-90, 1971,
pp.
1155-1165.
Skomal,
E.
N.
Man-Made Radio Noise.
New York: Van
Nostrand-Reinhold,
1978.
Skomal,
E.
N., and Smith,
A.
A.,
Jr.
Measuring the
Radio Frequency Environment.
New York: Van
Nostrand-Reinhold,
1985.
Cudak, M. C., Swenson, G.
W.,
Jr., and Cochran,
W. W.
“Airborne Measurements
of
Incidental
Radio Noise from Cities.”
Radio Science
26,
1991,
pp.
773-781.
35
Broadcasting, Cable
Television, and
Recording System
Standards
F.
M.
Remley,
J.
F.
X.
Browne, andS.
N.
Baron
Broadcast and Cable Transmission Systems
35-4
Standard Broadcasting
Frequency-Modulation Broadcasting
Television Broadcasting (VHF and UHF)
Cable Television
Other Television Services
Network Distribution
of
Broadcast Program Signals
Auxiliary Broadcast Services
International Broadcasting Service in the United States
Sources for Federal Communications Commission Documents
Program Production Standards
35-26
Sound Recording Systems
Television Recording Systems
Selected Lists
of
Television Standards
Digital Television Systems
35-34
Introduction-The Basics
Resolution
Sampling and Spectra
Quantizing and Dynamic Range
Composite and Component Signal Coding
Bit-Parallel and Bit-Serial Interfaces
Composite Encoded Signals: SMPTE
244
(NTSC and PAL)
35-
1