INTERNATIONAL MARITIME ORGANIZATION
E
IMO
SUB-COMMITTEE ON SAFETY OF
NAVIGATION
50th session
Agenda item 10
NAV 50/10/4
26 March 2004
Original: ENGLISH, FRENCH & SPANISH
ITU MATTERS, INCLUDING RADIOCOMMUNICATION
ITU-R STUDY GROUP 8 MATTERS
Recommendations ITU-R M.1372-1 and ITU-R M.1461-1
Efficient use of the radio spectrum by radar stations in the radiodetermination service and
Procedures for determining the potential for interference between radars operating in the
radiodetermination service and systems in other services
Note by the Secretariat
SUMMARY
Executive summary:
This document contains the text of Recommendations
ITU-R M.1372-1 and ITU-R M.1461-1 adopted by Study Group 8
concerning the efficient use of the radio spectrum by radar stations in
the radiodetermination service and procedures for determining the
potential for interference between radars operating in the
radiodetermination service and systems in other services
Action to be taken:
Paragraph 2
Related documents:
None
Introduction
1
Annexed are the texts of Recommendation ITU-R M.1372-1 (annex 1) and
ITU-R M.1461-1 (annex 2), which have been adopted by Study Group 8 concerning the efficient
use of the radio spectrum by radar stations in the radiodetermination service and procedures for
determining the potential for interference between radars operating in the radiodetermination
service and systems in other services and brought to the attention of IMO, ICAO, CIRM
and WMO.
Action requested of the Sub-Committee
2
The Sub-Committee is invited to consider the annexed information and take action as
appropriate.
***
For reasons of economy, this document is printed in a limited number. Delegates are
kindly asked to bring their copies to meetings and not to request additional copies.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
RECOMMENDATION ITU-R M.1372-1*
Efficient use of the radio spectrum by radar stations
in the radiodetermination service
(Questions ITU-R 35/8 and ITU-R 216/8)
(1998-2003)
Summary
This Recommendation provides some of the methods that can be used to enhance compatibility
between radar systems operating in radiodetermination bands. Several receiver post-detection
interference suppression techniques currently used in radionavigation, radiolocation and
meteorological radars are addressed along with system performance trade-offs (limitations),
associated with the interference suppression techniques.
The ITU Radiocommunication Assembly,
considering
a)
that the radio spectrum for use by the radiodetermination service is limited;
b)
that the radiodetermination service provides essential functions;
c)
that the propagation and target detection characteristics to achieve these functions are
optimum in certain frequency bands;
d)
that the necessary bandwidth of emissions from radar stations in the radiodetermination
service are large compared with emissions from stations in many other services;
e)
that efficient use of the radio spectrum by radar stations in the radiodetermination service
can be achieved by reducing transmitter unwanted emissions and utilizing interference
suppression techniques;
f)
that methods to reduce spurious emissions of radar stations operating in the 3 GHz and
5 GHz bands are addressed in Recommendation ITU-R M.1314;
g)
that the inherent low duty cycle of radar systems permits the use of interference
suppression techniques to enable radar stations in close proximity to use the same frequency,
recommends
1
that interference suppression techniques such as, but not limited to, those contained in
Annex 1, should be considered in radar stations to enhance efficient use of the spectrum by the
radiodetermination service.
*
This Recommendation should be brought to the attention of the International Maritime Organization (IMO),
the International Civil Aviation Organization (ICAO), the International Maritime Radio Committee
(CIRM), and the World Meteorological Organization (WMO).
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 2
ANNEX 1
Interference suppression techniques
1
Introduction
As spectrum demands for radiodetermination bands increases, new radar systems will need to
utilize the spectrum more effectively and efficiently. There will be heavily used areas throughout
the world where radiodetermination systems will have to operate in high pulse density
environments. Therefore, many radar systems may be subjected to pulsed interference in
performing their missions. The incorporation of interference suppression circuitry or software in
the design of new radar systems will ensure that system performance requirements can be
satisfied in the type of pulsed interference environment anticipated.
Interference suppression techniques, are generally classified into three categories: transmitter,
antenna, and receiver. Receiver interference suppression techniques are more widely used.
Receiver interference suppression techniques are categorized into predetection, detection and
post-detection.
The following is a brief discussion of several interference suppression techniques currently used
in radionavigation, radiolocation and meteorological radars. System performance trade-offs
(limitations), are also addressed for many of the interference suppression techniques.
2
Antenna beam scanning suppression
Interactions between two radars of different types almost always involve asynchronism between
the scanning of the two antenna beams. Consequently, the situations that are normally of
concern are limited to:
–
radar side lobe/back lobe to radar side lobe/back lobe;
–
radar main beam to radar side lobe/back lobe;
–
radar side lobe/back lobe to radar main beam.
The antenna side-lobe and back-lobe levels are generally determined by the radar antenna type
(e.g. reflector, slotted array, or distributed phased array). Reflector type antennas typically have
average antenna back-lobe levels of –10 dBi. Consequently, back-lobe-to-back-lobe coupling is
typically 70 to 80 dB weaker than main-beam-to-main-beam coupling. Slotted array antennas
and distributed phased array antennas can achieve back-lobe levels of approximately –30 to –
40 dBi resulting in back-lobe-to-back-lobe coupling typically 90 to 120 dB weaker than mainbeam-to-main-beam coupling.
The power coupled between two radars (radar 1 and radar 2) is proportional to the sum of the
gain of radar 1 antenna in the direction of radar 2 the gain of radar 2 antenna in the direction of
radar 1. The sum of the two antenna gains (G1(dBi) + G2(dBi)) is commonly referred to as the
mutual antenna gain. As the two antennas rotate, the mutual gain fluctuates rapidly by large
amounts. Since the rotations of the two radar antennas are asynchronous, i.e. since their rotation
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 3
rates are not rationally related, any one point on each radar’s antenna’s pattern lies in the
direction of the other radar shifts progressively through every point on that other radar’s pattern.
Eventually, the main-beam peak of each antenna will point toward the other radar at the same
time. However, that event will be exceedingly rare and fleeting. The vast majority of the time,
illuminations of each radar by the other radar’s main beam will occur when the other radar
illuminates the weak side lobe of the other radar.
This is especially the case when 3-dimensional radars, which use pencil beams scanned in
elevation as well as azimuth, interact with 2-dimensional radars, which almost invariably scan
only in azimuth. Thus, the pencil beams of 3-dimensional radars normally spend much of the
time searching regions above the horizon, where they cannot couple strongly to the surface-based
radionavigation radars. Furthermore, some 3-dimensional radars often use electronic steering
and scan in deliberately pseudo-random patterns or patterns that are quasi-random because they
adapt to the target environment. In such cases, the main beam of the 3-dimensional radars revisit
the direction of 2-dimensional radars only at irregular intervals instead of periodically. The fact
that main beams of all radars are narrow causes the fraction of time during which main-beam-tomain-beam conjunctions prevail to be extremely small.
Figure 1 shows a temporal pattern of mutual gain between two planar-array radar antennas with
both radar antenna beams scanning the horizon. Figure 2 shows the temporal pattern of mutual
gain between two planar-array radars with one of the radars beam scanning 45° above the
horizon. Figure 3 shows a mutual antenna gain distribution for two reflector type antenna radars
with gains of 27 dBi on the horizon. The Figure shows that only three per cent of the time the
mutual antenna gain exceeds 0 dBi, and fifty per cent of the time the mutual antenna gain is
below –19 dBi. Figure 3 also shows mutual antenna gain curves for two planar array type
antennas with both radar main beams on the horizon, and with one main beam elevated 45°.
FIGURE 1
Sample of mutual-gain pattern for planar-array RL and RN radar antennas with RL beam on horizon
(spans 7 scans of the RL radar antenna)
50
Mutual antenna gain (dBi)
40
30
20
10
0
–10
–20
–30
–40
–50
0
5
10
15
Time (s)
I:\NAV\50\10-4.doc
20
25
1372-01
NAV 50/10/4
ANNEX 1
Page 4
FIGURE 2
Sample of mutual-gain pattern for planar-array RL and RN radar antennas with RL beam elevation 45°
(spans 7 scans of the RL radar antenna)
50
Mutual antenna gain (dBi)
40
30
20
10
0
–10
–20
–30
–40
–50
0
5
10
15
20
25
Time (s)
1372-02
FIGURE 3
60
50
Mutual antenna gain (dBi)
40
30
20
10
0
–10
–20
–30
–40
–50
–60
0.01
0.1
0.5 1
2
5
10
30
50
70
90
95
98 99 99.5 99.9
99.99
Per cent of time exceeded
Two reflector type antennas
Two planar-array type antennas with both
mainbeams on the horizon
Two planar-array type antennas with one
mainbeam elevated at 45°
I:\NAV\50\10-4.doc
1372-03
NAV 50/10/4
ANNEX 1
Page 5
3
Integrator
The process of summing the echo pulses from a target is called integration. Integrators are
generally used in radars for two reasons:
–
to enhance weak desired targets for plan position indicator (PPI) display,
–
to suppress asynchronous pulsed interference.
The principle of the radar video integrator is that radar signal returns from a point target consist
of a series of pulses generated as the radar antenna beam scans past the target, all of which fall in
the same range bin in successive periods (synchronous with the radar’s transmitted pulses). It is
this series of synchronous pulses from a target which permits integration of target returns to
enhance the weak signals. The integrator also suppresses asynchronous pulsed interference
(pulses that are asynchronous with the radar’s transmitted pulses) since the interfering pulses will
not be separated in time by the radar period, and thus will not occur in the same range bin in
successive periods. Therefore, the asynchronous interference will not add-up and can be
suppressed.
Basically two types of integrators have been used in radar systems. The most common type of
integrator is the feedback integrator shown in Fig. 4. A binary integrator shown in Fig. 5 has also
been used in a few radionavigation radars.
Figure 6 shows a simulated output for a desired target return (pulse width = 0.6 µs, pulse
repetition frequency (PRF) = 1 000) without integration for a signal-to-noise ratio, S/N, of 15 dB.
Figure 7 shows a simulated output of radar without integration in the presence of the desired
signal and three interference sources (interferer 1, pulse width = 1.0 µs, PRF = 1 177;
interferer 2, pulse width = 0.8 µs, PRF = 900; interferer 3, pulse width = 2.0 µs, PRF = 280) with
interference-to-noise ratios (I/N) of 10, 15 and 20 dB, respectively.
FIGURE 4
Feedback integrator block diagram
ein
Input
limiter
eout
Output
limiter
Σ
Delay
K
TD = 1/PRF
1372-04
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 6
FIGURE 5
Binary integrator block diagram
ein
0, 1 binary
Threshold
comparator
Binary counter
or PROM
D/A
converter
eout
Shift register
(range gate)
1372-05
Clock
FIGURE 6
Simulated output of radar without integrator for
S/N = 15 dB
8
7
6
1 V/cm
5
4
3
2
1
0
0
5
10
15
20
25
5 ms/cm
I:\NAV\50\10-4.doc
30
35
40
45
50
1372-06
NAV 50/10/4
ANNEX 1
Page 7
FIGURE 7
Simulated output of radar without integrator in presence of interference
8
7
6
1 V/cm
5
4
3
2
1
0
0
5
10
15
20
25
30
35
40
45
50
5 ms/cm
Desired S/N = 15 dB
Interferer 1 I/N = 10 dB
Interferer 2 I/N = 15 dB
Interferer 3 I/N = 20 dB
3.1
1372-07
Feedback integrator
The feedback integrator shown in Fig. 4 consists of an input limiter, an adder, and a feedback
loop with an output limiter and a delay equal to the time between transmitter pulses (1/PRF) in
radars using non-staggered pulse trains. The overall gain, K, of the feedback loop is less than
unity to prevent instability. The input limiter serves as a video clipping circuit to provide
constant level input pulses to the feedback integrator, and is a necessary integrator circuitry
element to suppress asynchronous pulsed interference. The input limiter limit level is usually
adjustable, and controls the transfer properties of the feedback integrator. Figure 8 shows the
radar output for the same interference condition shown in Fig. 7 with feedback integration for an
input limit level setting of 0.34 V. The asynchronous interference has been suppressed by the
feedback integrator.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 8
FIGURE 8
Simulated output of radar with feedback integrator in presence of interference
8
7
6
1 V/cm
5
4
3
2
1
0
0
5
10
15
20
25
30
35
40
45
50
5 ms/cm
Desired S/N = 15 dB
Interferer 1 I/N = 10 dB
Interferer 2 I/N = 15 dB
Interferer 3 I/N = 20 dB
3.2
1372-08
Binary integrator
The binary integrator shown in Fig. 5 consists of a threshold detector or comparator, binary
counter or programmable read-only-memory (PROM) logic (adder/subtractor circuit), a multi-bit
shift register memory, and a digital-to-analogue (D/A) converter. Each inter-pulse period is
divided into range bins. Each time a pulse of a target return, noise, and/or interference exceeds
the comparator threshold level, the binary counter or PROM is bumped up to the next level. For
this simulation, a PROM logic with non-linear state progressions of 1, 2, 4, 8, 16 and 31 was
used. If the successive pulses of the target return pulse train continue above the comparator
threshold in the given range bin, the PROM is advance to the next highest programmed state until
a maximum integrator level of 31 is reached. If in any PRF period the signal fails to exceed the
comparator threshold, the PROM logic is bumped down to the next lowest programmed state
until a state level of zero is reached. The subtraction provides the target return pulse train signal
decay required after the antenna beam has passed the target, and also enables the suppression of
asynchronous interfering signals. The voltage amplitude at the integrator D/A converter output is
determined by the binary counter or PROM level (0 to 31) for the particular range bin times
0.125 V. Therefore, for a binary counter level of 31, the maximum enhancer output voltage
would be 3.875 V (31 × 0.125). Figure 9 shows the radar output for the same interference
condition shown in Fig. 7 after binary integration. The asynchronous interference has been
suppressed by the binary integrator.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 9
FIGURE 9
Simulated output of radar with binary integrator in presence of interference
8
7
6
1 V/cm
5
4
3
2
1
0
0
5
10
15
20
25
30
35
40
45
50
5 ms/cm
Desired S/N = 15 dB
Interferer 1 I/N = 10 dB
Interferer 2 I/N = 15 dB
Interferer 3 I/N = 20 dB
3.3
Trade-offs
Target azimuth shift:
Angular Resolution:
3.4
1372-09
0.9°
(0.7 beamwidth) for feedback integrator
0.2°
(0.2 beamwidth) for binary integrator
1.2°
(0.9 beamwidth) for feedback integrator
0°
(0 beamwidth) for binary integrator.
Desired signal sensitivity
Approximately 1 dB decreases when the integrator is adjusted to suppress pulsed interference
with the normal video mode and with moving target indicator (MTI) mode in the 2 and 3 pulse
canceller mode without feedback. However, in the MTI mode with feedback, the sensitivity loss
can approach 2 dB due to the need to adjust the integrator input limiter to limit the interference
level below the receiver inherent noise level.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 10
4
Double-threshold detection
The double-threshold detector, sometimes referred to as sequential detection, is a post detection
signal processing technique used in radionavigation and search radars. The function of the
double-threshold detection circuit is to extract or identify targets from radar target pulse returns.
However, the double-threshold method of detection also has an inherent capability to suppress
false alarms caused by asynchronous pulsed interference. Figure 10 shows a simplified block
diagram of a double-threshold detector.
FIGURE 10
Double-threshold detector block diagram
ein
First
threshold
Up
Shift register
(sliding window
of N PRI's)
Counter
Down
Second
threshold
(M out of N)
Target
No target
1372-10
The “double-threshold” detector consists of establishing a bias level, T, the “first threshold”, at
the output of the radar detector or Doppler filter and then counting the number of pulses whose
amplitude exceeds the bias level, T, in a “sliding time window”. The sliding window consists of
N successive repetition periods in a given range bin. Where N is approximately equal to the
number of pulses emitted as the beam scans through an angle equal to the half–power antenna
beamwidth. If in any given range bin the number of pulses exceeding T in the sliding window is
greater than or equal to a preassigned number M, the “second threshold”, a target is declared to
be present in that range bin. The values of the first threshold, T, and second threshold, M, are
chosen to meet a particular probability of false alarm, Pfa, and probability of detection, Pd.
There are also more complex double threshold detection criteria than discussed above. For
example, a fixed window size with separate leading and trailing edge first threshold levels can be
used. Also, a variable window size with separate leading and trailing edge first threshold levels
can be used.
Intuitively, the double-threshold technique should be useful in reducing the effects of
asynchronous pulsed interference. Target echoes received as the beam scans past a target will
occur in the same range bin. However, interfering pulses, occurring at random in the repetition
period, will be unlikely to occur in any given range bin more than a few times in N repetition
periods, unless the interfering pulse density is extremely high.
4.1
Trade-offs
The double threshold detector has a slightly poorer target probability of detection performance
than the integrators which sum the target return pulses. The performance (Pd and Pfa) of the
double threshold detector in suppressing asynchronous pulse interference depends on both the
first and second thresholds.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 11
5
PRF discriminator
Figure 11 shows a simplified block diagram of PRF discriminator, sometimes referred to as a
pulse-to-pulse correlator. The PRF discriminator utilizes a threshold comparator, delay (shift
register) and a coincidence circuit (AND gate) to suppress asynchronous interfering pulses that
do not have the same PRF (interpulse period) as the desired signal. The discriminator usually
operates at video, target pulses above the threshold are passed by the comparator; one pulse
repetition period later, a second target pulse arrives at the input to the coincidence circuit just as
the first leaves the shift register. In this scheme, all except the first pulse in the target return
pulse train are processed. The threshold level of the comparator is generally set at a 6 to 8 dB
threshold-to-noise ratio. More complex PRF discriminators can be designed to suppress
multiples of the desired signal PRF.
5.1
Trade-offs
The PRF discriminator does not enhance the desired signal as the feedback and binary integrator
circuits. Also there is a loss in desired signal sensitivity which is a function of the comparator
threshold setting.
FIGURE 11
PRF discriminator block diagram
ein
Threshold
comparator
Delay
(interpulse
period)
AND gate
eout
1372-11
6
Pulse width discriminator
If the pulse width of the interference differs from that of the victim radar, it may be used to
provide a means for discrimination. One method of implementing a pulse width discriminator is
shown in Fig. 12. The input pulse is differentiated and split into two channels. In one channel
the differentiated pulse is delayed a time corresponding to the width of the desired pulse τ, while
in the other channel the differentiated pulse is inverted. If the input pulse were of width τ, the
differentiated trailing edge inverted pulse would coincide in time with the leading edge pulse
delayed in time τ. The coincidence circuit permits signals in the two channels to pass only if
they are in exact time coincidence. If the input pulse were not of width τ, the two spikes would
not be coincident in time and the pulse would be rejected.
Pulse width discriminators are generally not effective against off-tuned interference due to the
inherent receiver IF output impulse response on the leading and trailing edge of an off-tuned
pulsed signal. The leading and trailing edge impulse response of an off-tuned pulsed signal are
each typically similar to the desired signal full pulse width because of the matched radar IF filter.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 12
6.1
Trade-offs
The utilization of pulse-width discriminators generally results in reduced receiver sensitivity and
probability of detection.
FIGURE 12
Pulse width discriminator block diagram
ein
Differentiator
Delay (T )
Coincidence
circuit
(AND gate)
eout
Inverter
1372-12
7
Pulse amplitude discrimination
Pulse amplitude discrimination can be used to suppress asynchronous pulsed interference if the
interfering signal levels are several dB above the receiver noise or clutter level. In one pulse
amplitude discrimination technique, the signal level in the same range bin is added for several
consecutive radar pulse periods. The voltage magnitude is then stored and the average voltage
computed. The voltage in each range bin is then compared with 4 or 5 times the average. If any
range bin exceeds this number, it is replaced by the average of the range bins. When there is
interference in only one of the range bins and noise only in the other range bins, asynchronous
pulsed interference with a peak I/N greater than 12 to 14 dB (depending on the criteria of 4 or
5 times the average) will be eliminated from further processing in the radar.
Many different algorithms can be developed to suppress asynchronous pulsed interferences based
on pulse amplitude discrimination. The radar mission and type of radar signal processing must
be taken into consideration in determining an appropriate pulse-amplitude discrimination
algorithm.
7.1
Trade-offs
Desired signal trade-offs should be minimal with proper choice of algorithms. Pulse amplitude
discriminators do not suppress weak interfering signals, and they do not work well in the
presence of strong clutter unless they include additional features.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 13
8
Asynchronous-pulse suppressor
In Doppler radars, individual pulses lose their identity in the Doppler filtering process, so direct
suppression of asynchronous pulses can only be done prior to Doppler filtering. This is
accomplished by implementing a local averaging and threshold process, for each range bin, that
spans all the PRIs or “sweeps” in each coherent processing interval (CPI) (instead of spanning
several range bins within a single PRI, as is done in a cell-averaging detection CFAR background
window). Since asynchronous pulses are normally absent from all but one of the PRIs in such a
group of samples, the average of the voltages, powers, or logarithms of voltage in each such
background window tends to be lower than the value in a particular range cell in which an
asynchronous interference pulse is sampled. As in a local-average-and-threshold CFAR process
used in the main detection flow, sensing threshold is set at a suitable multiple of the average over
the background window, and asynchronous pulses that cross that threshold, or detections
associated with those pulses, are excised.
9
Constant false alarm rate (CFAR)
It is virtually standard in modern radars to use some form of local-average-and-threshold CFAR
process. CFAR circuitry is used in both non-Doppler and Doppler radars. In Doppler radars, the
CFAR process is performed at the output of the Doppler filter bank. CFAR is performed to
provide a detection threshold that adapts to the clutter (and interference) level in the immediate
vicinity of each range/Doppler/azimuth cell that is being tested for target presence. Localaverage-and-threshold CFAR processes operate by constructing a sliding window for each PRI.
Each such window spans the range cell for which a first-detection decision is to be made plus
roughly 10 to 30 adjacent range cells (usually half of them at shorter range and half at longer
range). In local-average-and-threshold CFAR processes, the signal amplitudes in those adjacent
cells (often called the background window) are averaged and the average value is multiplied by a
factor such as 4 or 8 to establish the local detection threshold.
Low-duty cycle asynchronous pulse interference will not affect the threshold until I/N ratios are
in the order of 30 dB or greater. Also, in cell-averaging CFAR processes, a technique can be
used that excludes an individual cell that contains the strongest signals among the adjacent range
cells from the averaging (see § 8). This prevents isolated asynchronous pulses from
contaminating the threshold value and producing inappropriately elevated threshold levels.
However, continuous-wave like unwanted signals (BPSK, QPSK, etc.) will affect all
range/Doppler/azimuth cells, and thus raises the detection threshold resulting in loss of desired
targets. Other CFAR techniques, based on ranking the signal amplitudes in the cells of the
background window, are sometimes used. The signals in the highest-ranking cells are used only
to establish the rankings and are effectively discarded, so their actual levels do not affect the
threshold even via the average of all the cell values. These techniques therefore have a similar
mitigating effect on narrow unwanted pulses. All CFAR techniques also tend to prevent wide
unwanted pulses from producing false alarms. This is desirable when the duty cycle of the
unwanted signals is low, but degrades detection probability when high-duty-cycle unwanted
signals are received.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 1
Page 14
10
Doppler processing rejection
Even if the asynchronous pulse suppression techniques discussed in Doppler radars above are not
implemented, asynchronous pulses incur integration loss, relative to a synchronous pulse train, in
Doppler filtering. For example, Doppler filters generally use approximately 10 pulses per CPI,
but may have as low as 4 pulses per CPI. For the first case of 10 pulses per CPI, isolated
asynchronous pulses are rejected, relative to the synchronous return elicited by a valid target, by
roughly 18 dB (with allowance of 2 dB made for data-window weighting), while in the case of 4
pulses per CPI, they are rejected by roughly 10 dB (with similar allowance made).
Because Doppler radars have a multiplicity of Doppler passbands, another opportunity exists to
recognize isolated asynchronous pulses by virtue of the fact that a single pulse amounts to an
impulse input to each Doppler filter. Since an impulse has a uniform spectrum; i.e. since its
spectrum spans all frequencies, it evokes equal outputs from all the filters. Some Doppler
processors sense occurrences of simultaneous outputs from multiple Doppler filters and use such
occurrences to flag the presence of isolated (asynchronous) pulses. This technique can
complement asynchronous-pulse suppressor processes (see § 8) that operate prior to Doppler
filtering or it can be used in the absence of that process.
***
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
RECOMMENDATION ITU-R M.1461-1*
Procedures for determining the potential for interference
between radars operating in the radiodetermination
service and systems in other services
(Question ITU-R 226/8)
(2000-2003)
Summary
This Recommendation provides guidance and procedures for determining the potential for
interference between radars operating in the radiodetermination service and systems in other
services.
The ITU Radiocommunication Assembly,
considering
a)
that antenna, signal propagation, target detection, and large necessary bandwidth
characteristics of radar to achieve their functions are optimum in certain frequency bands;
b)
that the technical characteristics of radars operating in the radiodetermination service are
determined by the mission of the system and vary widely even within a band;
c)
that the radionavigation service is a safety service as specified by the Radio Regulations
(RR) No. 4.10 and harmful interference to it cannot be accepted;
d)
that considerable radiolocation and radionavigation spectrum allocations (amounting to
about 1 GHz) have been removed or downgraded since WARC-79;
e)
that some ITU-R technical groups are considering the potential for the introduction of
new types of systems (e.g., fixed wireless access and high-density fixed and mobile systems) or
services in bands between 420 MHz and 34 GHz used by radars in the radiodetermination
service;
f)
that representative technical and operational characteristics of systems operating in bands
allocated to the radiodetermination service are required to determine the feasibility of introducing
new types of systems;
g)
that procedures and methodologies are needed to analyse compatibility between radars
operating in the radiodetermination service and systems in other services,
*
This Recommendation should be brought to the attention of the International Maritime Organization (IMO),
the International Civil Aviation Organization (ICAO), the International Maritime Radio Committee (CIRM)
and the World Meteorological Organization (WMO).
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 2
recommends
1
that the procedures in Annex 1 provide guidance for determining the potential for
interference between radars operating in the radiodetermination service and systems in other
services;
2
that those radar characteristics contained in appropriate ITU-R Recommendations be
used for the frequency band under study.
NOTE 1 – This Recommendation will be revised as more detailed information becomes
available.
ANNEX 1
Procedures for determining the potential for interference between
radars operating in the radiodetermination service
and systems in other services
1
Introduction
Analysis procedures have been developed. Because of the high transmitter output power (50 kW
to several MW) and antenna gain (30 to 45 dBi) of radars operating in the radiodetermination
service (hereafter simply referred to as radars), compatibility between radars and systems in other
services is largely determined by analysing the effects of the emissions from radars on receiving
functions of other services. Therefore, this analysis procedure primarily addresses the methods
to assess the potential for interference from radars. In addition, potential desensitization of radar
receivers by emissions from modulated continuous-wave (CW) systems in other services is
briefly discussed.
By the nature of the missions of radars, many are mobile and cannot be constrained to prescribed
areas of operation. Also, the mission of radars often requires frequency agility and utilize the
entire allocated band. But when radars are anticipated to operate in certain areas in proximity to
other systems, the potential for interference can be assessed using the procedures contained in
this Recommendation.
2
Interference from radars to systems of other services
Investigations of several interference cases have identified two primary electro-magnetic
interference coupling mechanisms between high power radar systems and other services. These
interference coupling mechanisms are receiver front-end overload and radar transmitter
emissions coupled through the receiver IF passband. Discussion of the interference mechanisms
are provided below.
2.1
Receiver front-end overload
This interference mechanism occurs when energy from the fundamental frequency (necessary
emissions) of an undesired signal saturates the victim receiver front-end (low noise amplifier
(LNA) in some systems), resulting in gain compression of the desired signal sufficient to degrade
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 3
receiver performance. Receiver front-end overload is typically a result of inadequate
RF selectivity in the front-end of the victim receiver.
2.1.1
Assessing the potential for receiver front-end overload
The input threshold at which receiver front-end overload occurs is a function of the 1 dB gain
compression (saturation) level and the gain of the receiver front-end or LNA. Specifically:
T=C–G
(1)
where:
T:
C:
G:
input threshold at which receiver front-end overload occurs (dBm)
output 1 dB gain compression (saturation) level of the receiver front-end or
LNA (dBm)
gain of the receiver front-end or LNA at the radar fundamental frequency
(dB).
For example, if the receivers use LNAs with gains of 50 to 65 dB and they have an output 1 dB
compression level of +10 dBm, the range of values for T is –55 dBm to –40 dBm, depending on
the gain of the LNA.
A potential for interference from receiver front-end overload will exist whenever:
IT = T – FDRRF
(2)
where:
IT :
T:
FDRRF :
peak radar signal level at the antenna output or receiver input that causes
receiver front-end overload (dBm)
input threshold at which receiver front-end overload occurs (dBm)
frequency dependent rejection of the radar fundamental from any RF
selectivity that may be ahead of the receiver RF amplifier (LNA) or that may
be inherent in the RF amplifier (LNA) itself.
Equation (3) can be used to determine whether receiver front-end overload is likely when radars
operate within particular distances of other stations and are separated in frequency by certain
amounts:
I = PT + GT + GR – LT – LR – LP
(3)
where:
I:
PT :
GT :
GR :
LT :
LR :
LP :
I:\NAV\50\10-4.doc
peak power of radar pulses, at the radar’s fundamental frequency, at the
receiving antenna output or receiver input (dBm)
peak power of the radar transmitter (dBm)
main beam antenna gain of the radar (see Note 1) (dBi)
receiver antenna gain in the direction of the radar station under analysis
(dBi)
insertion loss in the radar station transmitter (dB) (2 dB assumed)
insertion loss in the victim receiver (dB)
propagation path loss between transmitting and receiving antennas (dB).
NAV 50/10/4
ANNEX 2
Page 4
In determining the propagation path loss, appropriate propagation models and possible indirect
coupling should be used taking into consideration antenna heights and terrain when appropriate.
If the calculated peak power of the radar pulses, at the fundamental frequency, I, exceed the
threshold at which receiver front-end overload occurs, IT, necessary steps to ensure compatibility
need to be taken.
NOTE 1 – Interference cases of radar transmitter emissions causing receiver front-end overload for radar
mainbeam coupling have been documented. Therefore, it is recommended that the radar mainbeam gain
be used in assessing the maximum potential for interference caused by receiver front-end overload.
2.2
Radar transmitter emission coupling
This interference mechanism occurs when energy emitted from the radar transmitter falls within
the IF passband of the receiver. This energy then passes through the receiver chain with little or
no attenuation. When the radar emission levels in the receiver passband are high relative to the
desired signal level, performance degradation to the receiver can occur.
2.2.1
Assessing the potential for radar transmitter emission interference
The initial step in assessing compatibility is the determination of the signal level at which the
receiver performance starts to degrade, IT.
IT = I/N + N
(4)
where:
I/N :
N:
interference-to-noise ratio at the detector input (IF output) necessary to
maintain acceptable performance criteria (dB)
receiver inherent noise level (dBm)
(N = –144 dBm + 10 log BIF (kHz) + NF
or
N = –168.6 dBm + 10 log BIF (kHz) + 10 log T)
where:
BIF :
NF :
T:
receiver IF bandwidth (kHz)
receiver noise figure (dB)
system noise temperature (K).
Also, the signal level at which a receiver starts to degrade, IT, can be calculated using
equation (5):
IT = C – (C/I)
(5)
where:
C:
C/I :
desired carrier signal level at the antenna output (receiver input) (dBm)
carrier-to-interference ratio at the predetector input (IF output) necessary to
maintain acceptable performance criteria (dB).
Equation (6) can be used to determine whether radar transmitter emission interference is likely
when radars operate within particular distances of other stations and are separated in frequency
by certain amounts.
I = PT + GT + GR – LT – LR – LP – FDRIF
I:\NAV\50\10-4.doc
(6)
NAV 50/10/4
ANNEX 2
Page 5
where:
I:
PT :
GT :
GR :
LT :
LR :
LP :
FDRIF :
peak power of the radar pulses at the receiver (dBm)
peak power of the radar transmitter under analysis (dBm)
main beam antenna gain of the radar under analysis (see Note 2) (dBi)
receiver antenna gain in the direction of the radar station under analysis
(dBi)
insertion loss in the radar station transmitter (dB)
insertion loss in the victim receiver (dB)
propagation path loss between transmitting and receiving antennas (dB)
frequency-dependent rejection produced by the receiver IF selectivity curve
on an unwanted transmitter emission spectra (dB).
NOTE 2 – Interference cases of radar transmitter emissions causing receiver degradation for radar
mainbeam coupling have been documented. Therefore, it is recommended that the radar mainbeam gain
be used in assessing the maximum potential for interference caused by radar transmitter emissions in the
receiver IF passband.
The FDR value to be used in equation (6) can be determined from Recommendation
ITU-R SM.337. The FDR can be divided into two terms, the on-tune rejection (OTR) and the
off-frequency rejection (OFR), the additional rejection which results from off-tuning the radar
and the receiver.
FDRIF (∆f ) = OTR + OFR(∆f )
(7)
For CW and phase-coded pulsed signals, the OTR factor is given by:
OTR = 0
for BR ≥ BT
(8)
OTR = 20 log (BT / BR)
for BR < BT
(9)
where:
BR :
BT :
receiver 3 dB bandwidth (Hz)
transmitter 3 dB bandwidth (Hz).
For chirped pulsed signals, the OTR factor is given by:
OTR = 0
2
(10)
2
(11)
for BC / (BR T) ≤ 1
2
OTR = 10 log (BC / (BR T))
for BC / (BR T) > 1
where:
T:
BC :
chirped pulse width (s)
transmitter chirped bandwidth during the pulsewidth, T (Hz).
Calculation of the OFR requires the IF response and the emission spectrum characteristics of the
radar transmitter. The ITU-R has provided methods for calculating the emission spectrum
characteristics of CW pulsed and chirped pulsed radars. If information is not available for radar
transmitter rise and fall time characteristics, the radar emission envelops should be calculated for
nominal rise and fall times of 0.1 µs. The spurious emission levels from radar transmitters are a
function of the transmitter output device. Representative spurious emission levels for various
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 6
radar output devices are contained in Recommendation ITU-R M.1314. Since many radars have
high transmitter power and antenna gains, large frequency separations, guardbands, may be
required to ensure compatibility.
In determining the propagation path loss, appropriate propagation models and possible indirect
coupling should be used taking into consideration antenna heights and terrain when appropriate.
If the calculated peak power of the radar pulses, at the receiver input, I, exceed the threshold at
which receiver performance degrades, IT, necessary steps to ensure compatibility need to be
taken.
3
Interference to radars from systems in other services
Introduction
Two primary electromagnetic interference coupling mechanisms between the radar system and
interfering signals from other services exist. The first mechanism is caused by front-end
overload causing saturation, and the generation of intermodulation products. The second is
interfering emissions within the receiver IF passband leading to desensitization and degradation
of performance resulting in an overall lowered quality radar data output.
3.1
Receiver front-end overload
3.1.1
Front-end saturation
This interference mechanism occurs when energy from an undesired signal saturates the LNA of
the radar receiver front-end resulting in gain compression of the desired signal which is sufficient
to degrade receiver performance. The input threshold at which receiver front-end overload
occurs is a function of the 1 dB gain compression (saturation) level and the gain of the receiver
front-end. Given a radar receiver with front-end RF bandwidth, BRF, and 1 dB compression input
power P1 dB (dBm), the total interference power inside BRF entering the radar receiver must not
exceed:
PI, RF max = P1 dB + ksat = C – G + ksat
dBm
(12)
where:
PI, RF
max :
ksat :
P1 dB :
C:
G:
maximum allowed total interference power inside the RF-bandwidth (dBm)
saturation margin (dB), to be determined individually for each radar and
interference type (ksat is generally negative)
defined as the 1 dB-input power compression point (dBm), i.e. when the
gain of the whole receiver chain has decreased by 1 dB
output 1 dB gain compression (saturation) level of the receiver front-end or
LNA (dBm)
gain of the receiver front-end at the fundamental frequency of the potential
interference source (dB).
For example, if the receivers use LNAs with a gain of 60 dB and they have an output 1 dB
compression level of +10 dBm, the value for P1 dB is 10 − 60 = −50 dBm.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 7
Fulfilment of equation (12) is essential in order to avoid driving the receiver close to or into
saturation and thereby to maintain sufficient dynamic range for the radar echo signal itself.
Moreover, the fraction of interference power falling into the radar receiver’s IF bandwidth must
also fulfil the requirements laid down in relevant ITU Recommendations.
A potential for receiver front-end overload from interference will exist whenever:
IT > PI , RF max − FDRRF
(13)
where:
I T:
FDRRF:
interference signal level at the receiver input that causes receiver front-end
overload (dBm)
frequency dependent rejection of the interference source by any RF
selectivity that may be ahead of the receiver RF amplifier (LNA) or that may
be inherent in the RF amplifier (LNA) itself.
Received interference power, aggregated over the full RF bandwidth, must not be higher than the
level which causes output power of that particular element in the receiver chain which first goes
into saturation to retain a sufficient separation below the 1 dB compression point. This is to limit
the reduction in dynamic range and to prevent 3rd-order intermodulation products exceeding the
acceptable I/N in the receiver’s IF bandwidth.
Equation (13) can be used to determine the interference signal level at the input of the first
amplifier stage of the receiver chain when interference sources operate within particular distances
of other stations and are separated in frequency by certain amounts, but are within the receiver
RF-bandwidth:
I = PT + GT + GR – Lt – LR – LP
(14)
where:
I:
PT:
GT:
GR:
Lt:
LR:
LP:
peak power of interference signal at the receiver input (dBm)
peak power of the interference source transmitter (dBm)
antenna gain of the interference source in the direction of the radar under
analysis (dBi)
receiver antenna gain in the direction of the interference source (dBi)
insertion loss in the transmitter (dB)
insertion loss in the radar receiver (dB)
propagation path loss between transmitting and receiving antennas (dB).
In determining the propagation path loss, appropriate propagation models and possible indirect
coupling should be used taking into consideration antenna heights and terrain when appropriate
(refer to § 3.3). If the calculated power of the aggregated interference sources exceed the
threshold at which receiver front-end overload occurs, IT, necessary steps to ensure compatibility
need to be taken.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 8
3.1.2
Intermodulation
For a radar receiver with large RF front-end bandwidth BRF (as defined in § 3.1.1) and with
usually much narrower IF bandwidth, the mechanism of 3rd-order intermodulation, with the
effect of transferring signal energy from outside the IF band (but inside the RF bandwidth) to
inside the IF bandwidth, must be taken into account.
Any pair of carriers at frequencies (“tones”) f1 and f2 may produce 3rd-order products of either
f3 = 2*f1 − f2, or f3 = 2*f2 − f1. Where f3 is inside the IF bandwidth, on-tune interference power is
generated which is further amplified and processed by the radar receivers IF and baseband
sections.
Intermodulation effects increase significantly with single tone power level. A 10 dB increase of
single tone power causes 30 dB increase of intermodulation product power. The (theoretical)
point where 3rd-order intermodulation product power starts to exceed amplified single tone
power is called 3rd-order output intercept point (IP3out). For RF LNAs as used in radar
front-ends, IP3out lies 10 to 15 dB above the 1 dB compression output power, P1 dB,out.
Interference power mapped by 3rd-order intermodulation from outside to inside the IF bandwidth
is to be treated like on-tune interference, i.e. it must be in combination with interference power
entering the IF band directly and fulfil the requirements laid down for the I/N ratio in the
appropriate ITU-R Recommendation.
3.2
Degradation of sensitivity
The desensitizing effect, on radiodetermination radars from other services of a CW or noise-like
type modulation is predictably related to its intensity. In any azimuth sectors in which such
interference arrives, its power spectral density can, to within a reasonable approximation, simply
be added to the power spectral density of the radar receiver thermal noise.
The initial step in assessing compatibility is the determination of the signal level at which the
radar receiver performance starts to degrade, IT.
IT = I/N + N
(15)
where:
I/N :
N:
interference-to-noise ratio at the detector input (IF output) necessary to
maintain acceptable performance criteria (dB)
receiver inherent noise level (dBm)
N = –114 dBm + 10 log BIF (MHz) + NF
where:
BIF : receiver IF bandwidth (MHz)
NF: receiver noise figure (dB).
Equation (15) can be used to determine whether systems in other services can operate within
particular distances of radars and are separated in frequency by certain amounts.
I = PT + GT + GR – LT – LR – LP – FDRIF
I:\NAV\50\10-4.doc
(16)
NAV 50/10/4
ANNEX 2
Page 9
where:
I:
PT :
GT :
GR :
LT :
LR :
LP :
FDRIF :
peak power of the undesired signal at the radar receiver input (dBm)
peak power of the undesired transmitter under analysis (dBm)
antenna gain of the undesired system in the direction of the radar under
analysis (dBi)
antenna gain of the radar station in the direction of the system under analysis
(see Note 3) (dBi)
insertion loss in the transmitter (dB)
insertion loss in the radar receiver (dB)
propagation path loss between transmitting and receiving antennas (dB)
frequency-dependent rejection produced by the receiver IF selectivity curve
on an unwanted transmitter emission spectra (dB).
NOTE 3 – Most radiodetermination radar antennas scan 360° in azimuth to substantial elevation angles.
However some radar system antenna scan in sectors, but the radar platform can ordinarily be oriented in
any azimuth. Interference to radar systems generally occurs when the radar antenna mainbeam points at
the undesired signal. Therefore, the radar mainbeam should ordinarily be used in the analysis. In some
special situations, the radar mainbeam may not illuminate the interacting station (e.g. sector blanking), in
which case the appropriate antenna sidelobe level should be used.
The FDR value to be used in equation (13) can be determined from Recommendation
ITU-R SM.337. Calculation of the FDR, requires the radar receiver IF selectivity response and
the emission spectrum characteristics of the radar transmitter. If the radar receiver IF selectivity
response is not provided, a selectivity fall-off of 80 dB per decade from the 3 dB bandwidth
should be used.
3.3
Protection criteria
In general cases, a signal from another service resulting in an I/N ratio below –6 dB is acceptable
by the radar users for signals from the other service with high-duty cycle (e.g. CW, BPSK,
QPSK, noise-like, etc.). An I/N ratio of –6 dB results in a (I + N)/N of 1.26, or approximately a 1
dB increase in the radar receiver noise power. In some cases, an I/N ratio of –6 dB may not be
appropriate, and further studies or compatibility measurements may be necessary to assess the
interference in terms of the operational impact on the radar’s performance. In the case of radar
systems operating in the band for which ITU-R Recommendation on radar characteristics and
protection criteria exist, then the relevant Recommendation1 should be consulted for specific
guidance regarding the protection criteria.
The effect of pulsed interference is more difficult to quantify and is strongly dependent on
receivers/processor design and mode of operation. In particular, the differential processing gains
for valid-target return, which is synchronously pulsed, and interference pulses, which are
usually asynchronous, often have important effects on the impact of given levels of pulsed
interference.
1
Some examples of Recommendations containing technical characteristics and protection criteria for specific
bands include: ITU-R M.1313, ITU-R M.1460, ITU-R M.1462, ITU-R M.1463, ITU-R M.1464,
ITU-R M.1465 and ITU-R M.1466.
I:\NAV\50\10-4.doc
NAV 50/10/4
ANNEX 2
Page 10
Several different forms of performance degradation can be inflicted by such desensitization.
Assessing it will be an objective for analysis of interactions between specific radar types. In
general, numerous features of radiodetermination radars can be expected to help suppress lowduty cycle pulsed interference, especially from a few isolated sources. Techniques for
suppression of low-duty cycle pulsed interference are contained in Recommendation
ITU-R M.1372 – Efficient use of the radio spectrum by radar stations in the radiodetermination
service.
When multiple interferers are present, the recommended I/N protection criteria remains
unchanged (because it depends on the type of radar receiver and its signal processing
characteristics). However, the total interference level actually arriving at the radar receiver
(which has to be checked against recommended I/N protection criteria) depends on the number of
interferers, their spatial distribution and their signal structure, and needs to be assessed in the
course of an aggregation analysis of a given scenario. The aggregation factor can be very
substantial in the case of certain high-density communications systems. If interference were
received from several azimuth directions, an aggregation analysis has to cumulate simultaneous
contributions from all these directions, being received via the radar antenna’s mainbeam and/or
side lobes, in order to arrive at the actual I/N protection criteria.
4
Choice of propagation model
In determining the propagation path loss, appropriate propagation models and possible indirect
coupling should be used taking into consideration antenna heights and terrain when appropriate.
In the general case, no account should be taken of shielding offered by terrain or man-made
obstructions. In specific sharing cases where accurate data is available, a detailed path analysis
and loss calculation may be carried out, or where the actual path loss between the interferer site
and the victim radar are determined experimentally for all frequencies of interest, this may be
accepted. If the calculated peak power of the aggregate interference sources at the radar receiver
input, I, exceeds the threshold at which receiver performance degrades, IT, necessary steps to
ensure compatibility need to be taken.
In the case of radars used for safety purposes, such as air traffic surveillance, the propagation
model must take into account all enhancing phenomena, which, even for short periods, could
cause the acceptable limit to be exceeded. Due to their operational requirement, radar systems
used for safety of life purposes, for example airport surveillance, require to be protected against
interference both for long-term and for short-term periods. Where the interfering signal can
originate without location identification over an area and be of short duration, special caution
must be observed.
___________
I:\NAV\50\10-4.doc