Radar Spectrum

CONTRIBUTED
P A P E R
Radar Spectrum Engineering

and Management: Technical
and Regulatory Issues
This paper presents the spectrum congestion problem from a radar perspective and
describes a number of possible approaches to its solution from a technical and
regulatory point of view.
By Hugh Griffiths, Fellow IEEE , Lawrence Cohen, Senior Member IEEE ,
Simon Watts, Fellow IEEE , Eric Mokole, Fellow IEEE , Chris Baker, Senior Member IEEE ,
Mike Wicks, Fellow IEEE , and Shannon Blunt, Senior Member IEEE
ABSTRACT | The radio-frequency (RF) electromagnetic spec- KEYWORDS | Radar; radar transmitters; radio communication;
trum, extending from below 1 MHz to above 100 GHz, radio broadcast transmitters; interference
represents a precious resource. It is used for a wide range of
purposes, including communications, radio and television
I. INTRODUCTION
broadcasting, radionavigation, and sensing. Radar represents
a fundamentally important use of the electromagnetic (EM) The radio-frequency (RF) electromagnetic (EM) spectrum,
spectrum, in applications which include air traffic control, extending from below 1 MHz to above 100 GHz,
geophysical monitoring of Earth resources from space, auto- represents a precious resource. It is used for a wide range
motive safety, severe weather tracking, and surveillance for of applications, including communications, radio and
defense and security. Nearly all services have a need for television broadcasting, radionavigation, and sensing.
greater bandwidth, which means that there will be ever-greater These applications are strongly influenced by the propa-
competition for this finite resource. The paper explains the gation characteristics of the environment and the direc-
nature of the spectrum congestion problem from a radar tivity achievable by antennas, both of which are dependent
perspective, and describes a number of possible approaches to upon the choice of frequency. The allocation of
its solution both from technical and regulatory points of view. spectrum is regulated by the International Telecommu-
These include improved transmitter spectral purity, passive nication Union (ITU) and continually reviewed at an
radar, and intelligent, cognitive approaches that dynamically international level by the World Radiocommunication
optimize spectrum use. Conference (WRC), with some bands assigned to
services on an exclusive basis while other bands are
Manuscript received November 1, 2013; revised June 27, 2014 and October 5, 2014;
shared between a number of services.
accepted October 15, 2014. Date of publication November 26, 2014; date of current Radar represents a fundamentally important use of the
version December 18, 2014.
H. Griffiths is with University College London, London, WC1E 6BT, U.K.
EM spectrum. It is used for a variety of purposes, including
(e-mail: [email protected]). air traffic control, geophysical monitoring of Earth
L. Cohen is with the Radar Division, Naval Research Laboratory, Washington,
DC 20375 USA (e-mail: [email protected]).
resources from space, automotive safety, severe weather
S. Watts was with Thales, Crawley RH10 9PZ, U.K. He is now with University College tracking, and surveillance for defense and security. As a
London, London, WC1E 6BT, U.K. (e-mail: [email protected]).
E. Mokole, retired, was with the Radar Division, Naval Research Laboratory,
sensor, it has the merits of allowing day or night and all-
Washington, DC 20375 USA (e-mail: [email protected]). weather operation (at frequencies below about 10 GHz)
C. Baker is with the Ohio State University, Columbus, OH 43210 USA
(e-mail: [email protected]).
and providing information such as target range and
M. Wicks is with the University of Dayton, Dayton, OH 45469 USA bearing, atmospheric measurements, onboard altimetry,
(e-mail: [email protected]).
S. Blunt is with the University of Kansas, Kansas City, KS 66160 USA
long-range imaging capabilities, and collision avoidance.
(e-mail: [email protected]). Nearly all services have a need for greater bandwidth.
Digital Object Identifier: 10.1109/JPROC.2014.2365517 In the case of communications and broadcasting, greater
This work is licensed under a Creative Commons Attribution 3.0 License. For more information, see http://creativecommons.org/licenses/by/3.0/
Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 85

Griffiths et al.: Radar Spectrum Engineering and Management Technical and Regulatory Issues
bandwidth is needed to satisfy the growing consumer applications of each. Whereas communications and broad-
demand for higher data rates, particularly to mobile casting use spectrum as a channel through which to convey
devices (e.g., streaming movies to a smartphone or a tablet information, radar derives information from the environ-
PC [1]). In contrast, higher bandwidth for radar translates ment itself with the particular frequency band having a
into finer range resolution, which directly relates to significant impact upon the exact nature of that information.
sensing capability (e.g., to detect an inbound hostile Inspection of a typical frequency allocation chart [3]
target). As the demand continues to grow for more access shows that the allocation plan is certainly complicated.
to spectrum by all these interested parties, there will be The spectrum is allocated to different services (broadcast-
ever-greater competition for this finite resource. ing, radiolocation, land mobile, aeronautical mobile, etc.).
The problem of meeting this demand may be addressed In the United States, some portions are further allocated
both by improved technology and by more intelligent on a government exclusive basis, some government/
frequency regulation [2]. A major impact of increased nongovernment shared, and some nongovernment exclusive.
demand is that many users will be forced to coexist within a Table 1 lists the frequency bands where radars operate
finite spectrum allocation, which in turn yields an increased and highlights the various sensing modes that are
likelihood of mutual interference. The ways in which performed in each band. By convention, some radar bands
different services may interfere with one another is not are designated by letters, for example L-band (1–2 GHz),
always well understood, and this misinformed perspective S-band (2–4 GHz), C-band (4–8 GHz), X-band (8–12 GHz),
may cause over/undercautious decisions to be made with and so on [4].
regard to spectrum allocation. Better appreciation of the An individual radar will not usually occupy the total
nature and effect of interference from/to various users may frequency allocation in any particular radar band. The
allow different services to operate within the same or nearby bandwidth of a radar determines its ability to resolve targets
band with a tolerable level of disruption. Improved in range. The bandwidth that an individual radar will occupy
technology may also help to alleviate this problem, such as will depend not only on the range resolution required, but
through adaptive control of emissionsVin the frequency, also on the need to reduce the potential for interference with
time, space, polarization and coding domainsVthat could other radars in the same band, by operating on different
potentially allow for more efficient use of the spectrum for frequencies. In some scenarios, there may be very many
the coexistence of different services. radars operating in close vicinity. For example, in a busy
Just as digital technology has enabled the telecommuni- shipping lane within, say, a 20-nmi radius, there may be
cation industry to make a quantum leap in terms of capacity many tens of large ships, fishing boats, and recreational craft,
and quality of service in recent years; it is likewise facilitating each with a radar operating in S-band or X-band. There may
the potential for tremendous improvement in radar receiver be a civil aviation flight path above, with each aircraft
performance and the control of radar emissions. For exam- operating a weather radar in X-band. It is easy to imagine that
ple, precise wide-bandwidth radar waveforms can now be more than 100 emitters might be simultaneously detected by
generated and varied dynamically, potentially on a pulse-by- a wideband receiver.
pulse basis. In addition, ongoing improvements in power The higher frequency bands provide some advantages
amplifier and filter capabilities can provide better spectral to radar. For a fixed fractional bandwidth, increasing the
purity to allow closer channel spacing. operating frequency subsequently increases the achievable
The purpose of this paper is to explain the nature of the bandwidth, thus providing finer range resolution. In
spectrum congestion problem from a radar perspective, addition, for a fixed angular beamwidth, the antenna size
and to describe a number of possible approaches to its decreases as the wavelength is reduced (antenna beam-
solution, both from technical and regulatory points of width in one dimension is proportional to =D, where is
view. It is written on behalf of the radar community, but is the RF wavelength and D is the antenna aperture width).
aimed at all users of the EM spectrum, making the case for However, in these higher bands, long-range operation
radar’s use of the spectrum and explaining what is becomes more strongly affected by attenuation due to the
presently being done to contend with spectral congestion atmosphere, especially in the presence of rain or clouds. As
and what may be done in the future. The authors come such, radar sensing via these bands is limited to short-range
from different countries and have participated in a number applications like automotive collision avoidance, police
of different national and international studies on the radar, airport surveillance, and scientific remote sensing.
subject of radar spectrum usage. Furthermore, the lower bands offer some unique capabilities
such as ionospheric propagation for over-the-horizon
surveillance [at high frequency (HF)], foliage and ground
II. RADAR S PECTRUM E NVIRONMENT penetration [at very high frequency (VHF) and ultrahigh
The spectrum environment in which radar operates is frequency (UHF)], and long-range surveillance, tracking,
absolutely critical to the particular sensing operation, of air traffic control, and weather monitoring (at L-, S-, and
which there are numerous forms, including surveillance, C-bands). Because the world is so complex, the task of
imaging, and tracking along with the myriad different sensing clearly does not have a ‘‘one size fits all’’ solution.
86 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Table 1. Standard Radar Frequency Letter Bands and Radar Operating Modes [4]
Many airborne radars, such as airborne interception (AI) tions and broadcasting. Very long-range communications
radars in fast jets, or those used for surveillance of the land are possible at these frequencies, depending in the
and sea, will use X-band. This is a good compromise between propagation conditions which can change considerably
an acceptable antenna size for an airborne platform and the over both short and long timescales. There have previously
ability to operate over long ranges in poor weather. been no designated frequency allocations for HF radars,
The radar allocations are interleaved, or in some cases which may operate anywhere from about 3 to 50 MHz.
shared, with the equivalent communications and broadcast Such operation has been done as a secondary user, having
bands. In the HF bands, the primary uses are communica- to avoid interference with primary users. However, the

2012 WRC proposed a number of primary allocations for energy in each pulse Pt . The power averaged over time is
the radiolocation service between 3 and 50 MHz to then given by Pav ¼ ðPt f r Þ, where the value of f r is
support HF oceanographic radar operations [6]. known as the duty cycle (i.e., the fraction time available to
At VHF and UHF, the spectrum is also very crowded, listen for returns). At a given range, the minimum
being particularly used for communications and broadcast- detectable target size will be proportional to the average
ing. Again, radar operation in these bands is generally on a power. However, it should also be noted that the minimum
secondary basis. At microwave frequencies, the radar L-band range that can be measured by the radar will also be
and above, there are radar allocations as a primary user, but determined by the pulse length, since the receiver is
there is a growing number of applications putting pressure on blanked during transmission. So a pulse length of 10 s
the available bandwidth. Typical applications around the will give a minimum range of about 1.5 km. Radars also
radar L- and S-bands include mobile telecommunications, in measure the Doppler frequency shift imparted by relative
particular, and also wireless local area networks (LANs), radial motion between the radar and the target. A relative
Bluetooth, and the Global Positioning System (GPS). At radial velocity v will induce a Doppler frequency shift of
higher frequencies, up to about 30 GHz, uses will include 2v= Hz. So a radial velocity of 1 m/s will give a Doppler
radioastronomy, microwave communications links, satellite shift of 66.7 Hz in X-band ( ¼ 0.03 m). The Doppler
television broadcasting, and communications satellites. frequency shift is usually very small compared to the pulse
bandwidth and has to be measured by observing the
successive phase shifts of returns over burst of pulses. In
III . RADAR EMISSIONS these circumstances, the maximum unambiguous Doppler
Radar systems transmit RF signals, known generally as the shift that can be measured is f r =2. This is another
radar waveform, modulated in such a way as to enable constraint on waveform design.
measurement of range and the Doppler shift due to relative The other class of radar emissions is known as continuous
motion between radar and target, and to resolve distinct wave (CW). These radars transmit the interrogating
scatterers. The specification of the radar signal is dictated waveform and receive reflected signals simultaneously,
by sensing requirements such as range resolution (which is thus requiring separate transmit and receive antennas with
bandwidth dependent), Doppler resolution, maximum very good isolation to prevent saturation of the receiver.
ambiguous range, and radar sensitivity. The selection of Once again, the ability of these radars to resolve targets in
a particular waveform or class of waveforms is made range is determined by the bandwidth of the transmissions.
according to the various performance tradeoffs conveyed In a CW radar, bandwidth is achieved by sweeping the
by these requirements. frequency over the required bandwidth, usually with a linear
The most prominent class of radar emissions is based on rate of frequency change over time. A typical frequency-
the successive transmission of pulses, where phase or modulated continuous wave (FMCW) radar might sweep
frequency may in turn be modulated during the pulse. These over a bandwidth of 100 MHz in a repetition period of 2 ms.
pulses may be as short as 100 ns to longer than 100 s, This can be compared with a pulsed radar with the same
depending on the application and possess a pulse repetition bandwidth, which might sweep the frequency over the same
frequency (PRF) from around 300 Hz up to 100 kHz. The bandwidth within the time duration of a pulse, which might
intervals between pulses are used to receive the signals be, for example, 5 s.
reflected from objects (conventionally known as radar For both classes of radars, it is standard for the emissions
targets) at various distances from the radar, and such to possess a constant envelope (i.e., no amplitude modulation
intervals may typically be at least 1000 times longer than the effects) so as to maximize the energy incident upon, and
pulse length. In other words, the radar spends most of its subsequently reflected from, the illuminated objects and to
time ‘‘listening’’ for the faint echoes from distant targets for thereby maximize the achievable sensitivity. The use of
which it requires exclusive access to the requisite part of the constant envelope pulses is also motivated by the need to
spectrum in time, frequency, and space. drive power amplifiers in saturation to obtain the best power
The choice of pulse length and PRF will be dependent efficiency. However, it may also be noted that driving the
on the tasks being performed by the radar. A radar with a amplifiers in this way also causes a high degree of non-
PRF of value f r will be able to listen for a time period of linearity, which can cause a broadening of the transmitted
about 1=f r , before the next pulse is transmitted. The spectrum.
receiver is usually blanked during transmission, to prevent The center frequency of the radar emission is
damage to its very sensitive front–end circuits. This means determined by the mean carrier frequency, which may
that targets out to a range of c=ð2f r Þ can be measured be fixed or variable over time, say from pulse to pulse,
unambiguously, before being potentially confused with according to some preselected basis (the latter is known as
returns from later pulse transmissions. So, for example, a frequency agility). The instantaneous bandwidth of the
radar with a PRF of 1 kHz can measure range unambig- waveform is determined by the modulation that is applied
uously out to 150 km. The pulse length for a given peak to the pulse. A radar bandwidth B can provide a range
power level Pt , from the transmitter, determines the total resolution of c=ð2BÞ, where c is the velocity of light. For

Fig. 1. Power spectral density of a radar using a pulsed magnetron (Furono Model 1953C X-band maritime surface radar), with a pulse length of
100 ns giving a nominal bandwidth of 10 MHz.
example, a radar pulse might be characterized as having a from pulse to pulse. The artifacts resulting from this
center frequency of 9.6 GHz, a pulse length of 5 s, and a process are rather steep asymmetrical sidebands on either
bandwidth of 100 MHz, with the radar being frequency side of the spectral mainlobe (Fig. 1). These frequency
agile over a total band of 500 MHz or more. The sidebands can cause adjacent channel interference to other
bandwidth of a radar pulse determines its ability to resolve occupants of the spectrum. Bandpass filters have been
targets at different ranges. For example, a bandwidth of employed on magnetron-type transmitters as a means of
100 MHz implies a maximum range resolution of 1.5 m. reducing this out-of-band (OOB) interference, though the
Some radars, especially those used to image the ground cost of this improved spectral purity is a significant loss of
using synthetic aperture radar techniques, may have effective transmitter power. Note in this example that the
bandwidths of more than 1 GHz, implying a potential half power bandwidth is about 10 MHz, commensurate
range resolution better than 15 cm. Such wide bandwidths with a pulse length of 100 ns. However, at the level 40 dB
may be needed to classify targets and accurately measure below the peak, often used in defining spectrum occupan-
their position in a complex scene. cy, the spread of frequencies is of the order of 100 MHz. If
this magnetron had a peak power level of 1 MW, which is
A. Radar Transmitters quite feasible, then even these OOB signals at a level 40 dB
The radar transmitter is the component responsible for below the peak will be equivalent to a transmitter with a
the generation and amplification of the radar waveform, 100-W peak power transmitting over a bandwidth of more
thus providing the energy required to detect objects at long than 100 MHz. This may interfere with other radars or
ranges. Depending upon the particular sensing application, services operating in adjacent frequencies.
the generated peak power can be anywhere from milliwatts In contrast to the magnetron, all other types of radar
to megawatts. The transmitter may be based on either transmitters rely on separate amplifier and waveform
vacuum tube or solid-state technology. The simplest and generation stages to enable better control of the waveform
most widespread method of generating high-power RF characteristics. In many modern radar systems, the
energy is with a magnetron tube, which is a high-powered waveform generator is a digital synthesizer operating
oscillator that can be pulsed. The cross field amplifier with very stringent frequency tolerances and extremely
(CFA) or Amplitron is used to further boost the output low levels of sideband energy. The master clock in the
power from magnetrons. While inexpensive, the magne- digital synthesizer is used to derive all timing for the radar,
tron suffers serious drawbacks in terms of spectral purity. including the PRF. The digitally synthesized waveforms
A modulating pulse initiates the magnetron; as the buildup are converted to analog format and passed to a power
of RF energy grows from noise to a critical point, the amplifier, before radiation by the antenna. Commonly
magnetron begins to oscillate. These oscillations differ used radar power amplifiers based on tube technology

include the klystron, traveling wave tube (TWT), and CFA. are under development for higher power transmitter
Klystrons can generate megawatts of peak power but are modules. GaN has a greater power density than GaAs and
limited in bandwidth due to the restrictions of their resonant can operate at higher voltages. GaN devices have the
cavities. For example, the Bendix AN/FPS-20 air surveillance potential to develop higher powers and to operate at higher
radar, which used a klystron-based transmitter of 1950s frequencies (i.e., > 30 GHz). However, a more important
vintage, had a peak power of 2 MW, a pulse length of 60 s, benefit may be the potential for improved power efficiency
and operated between 1.25 and 1.35 GHz. Traveling wave compared to GaAs devices.
tubes provide peak powers of the order of 0.1–50 kW, and Solid-state-based radars have the advantage of being
typically have much broader bandwidths than klystrons (up amenable to techniques for controlling OOB spectral
to two or three octaves). While ongoing work is seeking to emissions such as bandpass filtering and amplifier linear-
improve the spectral purity of these tube devices, the reality ization [13]. As a general rule, solid-state amplifiers cannot
is that legacy systems, particularly for defense applications, provide the high peak power of tubes but they are usually
will be in abundance for the next 50 years due to the long able to sustain a much higher duty cycle (the product of
acquisition cycle for such systems and the enormous costs pulse length and PRF), which causes solid-state radars to
involved with building large modern radar systems. rely on waveforms with large duty cycles to provide
Solid-state power amplifiers have been employed for commensurate ‘‘energy on target.’’ It must be kept in mind
several years in radar applications, as standalone ampli- that, currently, solid-state amplifiers represent a small
fiers, as replacements for amplifiers using vacuum tube minority of the total number of operating systems. They do,
devices, and in distributed modules as part of active however, offer much scope for improved spectral control.
electronically scanned arrays (AESAs) [6]. At lower There are many different types of radar with widely
frequencies, say below 3 GHz, silicon bipolar transistors varying power levels, bandwidths, and spectral characteristics.
may be used with duty cycles of G 10% to generate peak A small number of generic examples are listed in Table 2.
powers of the order of 100 W. At higher frequencies
(usually up to about 30 GHz) GaAs devices are used. A
transmit/receive module with two or more devices in I V. I SS U ES W I T H S PE C TRU M
parallel in an AESA radar at X-band might generate a peak ALLOCATION FOR RADAR
power of 10 W. A small AESA may have 1000 transmit The electromagnetic spectrum is becoming increasingly
elements giving a total peak transmitter power of 10 kW, more congested as a result of rapid expansion by the
with duty cycles up to 20%. Newer materials such as GaN commercial wireless industry and other RF applications. This
TABLE 2 Examples of Generic Radar Types and Their Transmitters (See Also [6])

has been and continues to apply ever-greater pressure on the radar performance, reducing the maximum detection range
parts of the spectrum formerly reserved exclusively for radar. [7]. However, they also show that a combination of moving
Consequently, as radar frequency allocation dwindles and up the frequency band and improving receiver filter design
assigned bands narrow, spectral crowding and deleterious can mitigate this interference. Thus, although this effectively
effects of OOB leakage further compound this severe and reduces the band allocated to radar, performance can still be
growing problem. With more commercial users occupying maintained even in important safety critical applications
spectrum previously assigned exclusively to the radar such as air traffic management.
community and with guard bands disappearing, spurious
emissions from consumer electronics are causing increased
A. Spectrum Regulation
in-band interference in a multitude of airborne and ground-
An important part of spectrum management is how
based radars. The demand for wireless access (particularly
frequency use and emissions are regulated (see [1] for a broad
wireless video and data services) is increasing at an
perspective on spectrum regulation). For radar, regulation is
accelerating rate, further eroding the spectrum allocation
particularly complex due to the variety of different radar
assigned to radar applications, and is part of a trend that has
modes, their necessary power outputs (which dictates the
roots going back over more than 30 years.
nature of the specific transmitter), and the induced spectral
The issues of spectrum congestion and competition
emissions. Many, but not all, countries adopt ITU emission
with radar by other services currently arise mainly in the
standard. In the United States, emission standards are
frequency bands below 5 GHz (C-band). In the higher
determined by two organizations: the National Telecommu-
frequency bands, the use of bandwidth is still strictly
nications and Information Administration (NTIA), the
regulated to prevent interference, especially to critical
governing body for all U.S. federal government spectrum
services such as airport surface movement radars and radio
use; and the U.S. Federal Communications Commission
navigation systems. Since its inception in the 1950s, HF
(FCC), the regulatory authority on spectrum use by
radar has always competed for spectrum with the primary
nonfederal entities like the commercial broadcasting indus-
users of HF communications and the amateur radio world.
try. The ITU has published their manual of radio regulations
Since the late 1970s, the world’s communication industries
since the dawn of wireless in 1906. Today the regulations
have shown greater interest in the UHF part of the
cover the frequency range from 9 kHz to 1000 GHz for 40
spectrum and, in 1979, the World Administrative Radio
different radio services, including radar (which falls under
Conference (WARC) decided to downgrade the primacy of
the classification of radiodetermination or radiolocation
radar in portions of the UHF band, specifically 420–430
services), in a 1000+ page publication [7]. These regulations
and 440–450 MHz, to secondary status. In the language of
can only be changed by the WRC.
spectrum management, downgrading to secondary status
Within the ITU guidelines, provisions are made for the
means that radars can operate only as long as they do not
computation of emission masks that delineate OOB
interfere with primary users. Over the last ten years,
emissions that emissions from a real transmitter have to
wireless industries have lobbied their member nations
sit within. For example, Fig. 2 shows a typical emission
within the ITU to downgrade radar in the 3.4–3.7-GHz
mask that might be applied to a radar transmission.
band to secondary status as well. Currently, the big
Broadly there is a band over which the radar is designed to
competitor for the 3.4–3.7-GHz band is fourth-generation
(4G) wireless communications [worldwide interoperabil-
ity for microwave access (WiMAX) or long-term evolution
(LTE); though all indications are that the latter will
dominate].
As a more specific example, the UK, coordinated through
the communications regulator, Ofcom, is examining the
allocation of frequencies between 2.62 and 2.69 GHz to
facilitate an expansion of WiFi services. This sits very close to
the radar allocation at S-band that spans 2.7–3.1 GHz. S-band
is predominately used by air traffic management and air
defense radar systems. To determine the viability of such a
move, the U.K. Government departments such as the
Department for Culture, Media and Sport (DCMS), the
Department for Transport (DfT), the Ministry of Defence
(MOD), and the Civil Aviation Authority (CAA) supported
by Ofcom have been working together to help radar operators
make their systems more resilient to interference from
interference due to emitters in the 2.6-GHz band. Studies Fig. 2. Graph of a generic ITU spectral mask showing permissible
have shown that OOB emissions will significantly degrade regions of operation for differing rates of spectral roll-off [9].

transmit that is fixed in frequency and goes down to 40 dB

from the peak. Outside of this, at lower emission levels,
various ‘‘roll-off rates’’ could be applied. Figure three shows
two examples, one at a roll-off rate of 20 dB/decade of
frequency and the other at 30 dB. A roll-off rate of 20 dB/
decade is the current standard while 30 and 40 dB/decade is
only under consideration. The radar emissions have to sit
inside a mask whose power versus frequency shape of the
mask is determined by the regulatory bodies. The shape of
the masks has significant implications for radar design and
performance. While lessening the potential for adjacent band
interference to other services, a 40-dB/decade roll-off rate
poses extreme challenges to the radar designer due to the
intrinsic spectral spreading that results from pulsed opera-
tion combined with the requirement for high transmit power
which tends to drive the use of tube technology and its
subsequent limitations (i.e., nonlinear distortion in the form
of intermodulation products).
In the United States, the NTIA publishes a guide:
‘‘Manual of Regulations and Procedures for Federal Radio
Frequency Management,’’ better known as the Red Book
[10]. Of most relevance here is Chapter 5.5, wherein the
radio spectrum engineering criteria (RSEC) are defined. In
the RSEC, radars are divided into five classes, A through E.
This partition considers such factors as frequency cover-
age, peak power output, type of waveform (pulsed versus
nonpulsed), and functionality (wind profiler, etc.). The
RSEC determines a spectral mask based on a 40-dB
bandwidth with roll-off rates that are calculated with
equations according to the criteria specified in the five
class designations. Fig. 3(a) shows a generic example of an
RSEC mask. The desired in-band radar emissions are
contained within the 40-dB bandwidth as shown, with a
subsequent allowable roll-off at today’s 20 dB/decade
down to a lower limit requirement of 60 dB at all other Fig. 3. (a) RSEC emission mask showing the signal domains and their
permitted relative levels [8]. (b) Example showing a measured
frequencies. The radar transmissions should not exceed
emission of an unmodulated pulse from a weather surveillance
the limits implied by this mask but unwanted emissions radar within the RSEC box. At around 3050 MHz the emission
often occur. Unwanted emissions from a radar transmitter exceeds the allowable limit for the signal domain [8].
are composed of OOB and spurious components. These are
generated by nonlinearities that occur within the trans-
mitter together with the steep rise and fall times of the
radar pulses. Fig. 3(b) shows an example of a radar V. EFFECTS OF RADAR INTERFERENCE
emission relative to the RSEC mask for the current roll-off ON OTHER USERS
rate of 20 dB/decade. The radar operates in S-band with a
transmitted signal that is designed to emit between 2990 A. Radar-to-Radar Interference
and 3000 MHz only. Note that the radar emission is only Most radars currently in service operate in a pulsed
marginally out of compliance in the upper sideband, i.e., it mode, with rotating antennas having narrow transmit and
exceeds the value set by the mask. This requires receive beamwidths and low spatial sidelobes. These
modifications to be made to the transmitter so that it features help to protect against interference from other
can meet the requirements demanded by the mask. There radars. Any significant interference will tend to occur when
are potential methods to better control radar emissions, the two radar beams are aligned, which is usually only for a
and these are discussed in more detail in Section VI. short period of time. The pulsed nature of the systems also
However, the effects of such a transgression on the means that unless the PRFs of the two radars are
performance of, say, a communication system operating in synchronized, the interference will be suppressed in the
an adjacent band are unclear but will be a function of receive signal processing. Given these features and the
waveform and signal processing. relatively small number of radars in a local area, it has in

Fig. 4. Interference to an X-band airborne radar using pulse-to-pulse frequency agility. (a) High duty cycle or CW interference. (b) High and low
duty cycle interferences.
the past usually been possible to minimize radar-to-radar In some circumstances, multiple radars may need to
interference by careful allocation of operating frequen- operate in the same narrow frequency band and in close
cies and control of geographical location. For those radars proximity. One example is civil marine navigation radars
employing traditional plan position indicator (PPI) detec- on ships. For such systems, techniques of ‘‘de-fruiting’’
tion displays, human operators are very good at identifying have been developed that allow for small adjustments to
and addressing interference issues. Examples of interfer- the radar’s PRF and thereby reject signals that are not
ence observed on the PPI display of a noncoherent airborne synchronous. In other systems, such as air traffic
radar are shown in Fig. 4. The green-colored plots on the management radars, interference is avoided by very careful
PPI are an overlay of relatively unprocessed radar returns control of frequency allocations, very good front–end
(the ‘‘raw’’ radar returns), which show the presence of receiver filtering and, if interference is unavoidable,
clutter and interference before the main radar detection techniques such as sector blanking.
processing is applied. The radar was using pulse-to-pulse For some modern radars, where detections are
frequency agility. In Fig. 4(a), some high duty cycle or CW processed without the oversight of a radar operator,
interference can clearly be seen. This interference is interference may be more problematic if not properly
manifested as distinct spokes as only certain frequency addressed during the radar’s design. Modern radars may
channels in the radar receiver are affected. The interfer- also have reduced front–end frequency selectivity as a
ence is also only seen in the main antenna beam and near consequence of the need to operate over a very wide fre-
sidelobes as the radar scans past the interfering signal. quency bandwidth (e.g., >1-GHz bandwidth at Ku -band).
Some similar interference can also been seen in Fig. 4(b). In such cases, mitigation of interference by frequency
Also seen in this image is a spiral of pulse-like returns. This planning is not usually possible within the constraints of
would have been due to another pulsed radar having a low the allocated bands. Where interference is likely,
duty cycle with a PRF similar to that of the radar being sophisticated coordination techniques will be required
interfered with. In this case, the interference must have to operate the radars. In addition to the lack of selectivity
been entering the radar through the antenna sidelobes, when operating over wide bandwidths, radars are
because it is visible at all azimuth angles over the display. increasingly required to detect ever-lower signal levels
For a modern high-performance radar with advanced (i.e., enhanced sensitivity). This requirement has led to
adaptive detection processing, such interference might the development of very low noise level front–end
not result in false detections. This is because the radar will amplifiers, which often lead to a reduced dynamic range
be able to sense the interference and raise its thresholds in the receiver and can result in saturation by large
accordingly. However, this in turn would reduce the signals, including interference from within the band or
sensitivity of the radar in the vicinity of the interference, from adjacent bands. Mitigation of this effect will
making it difficult to detect smaller targets. Such effects predicate very careful front–end design and the develop-
can be quite insidious, with the radar losing performance ment of highly linear components and filters with sharp
without this being obvious to the operator. responses.

Fig. 5. Sixty-four-QAM constellation (a) with no interference; and (b) with interference from an S-band radar.
B. Radar Interference to Other Systems Due to the two-way path loss, achieving sensitivity to the
Various cases of interference by radars with commercial echoes from long-range scatterers necessitates the trans-
electronic equipment have been reported from time to time. mission of much higher energy than other spectral users
Examples include automatic teller machines (ATMs), that are subject only to one-way path loss. Furthermore, to
satellite and terrestrial television receivers, and hearing resolve scatterers that are closely spaced in range, the
aids. Sometimes, this is a problem of poor electromagnetic radar emission must possess a high bandwidth (range
compatibility (EMC) design where improved screening and resolution and bandwidth are inversely proportional). This
filtering within the victim equipment would mitigate the high bandwidth emission could, in principle, be achieved
interference. However, even with the best EMC design, some via an extremely high-power pulse of short duration
problems may be unavoidable and such interference may well (commensurate with the range spacing of the scatterers to
increase in the future without careful spectrum planning. be resolved). However, it is generally far more practical to
emit a longer pulse of lower power (albeit still quite high
C. WiMAX and LTE Communication Systems relative to ‘‘one-way path loss’’ users) that is phase/
Wireless communications, particularly the push for 4G frequency modulated (the waveform, as discussed in
systems, is driving much of the worldwide pressure for Section III). This latter instantiation is known as pulse
spectrum. These 4G systems, based on either the LTE or compression and can be viewed as a form of spread
WiMAX standards, rely on orthogonal frequency-division spectrum operation [the example of a linear frequency
multiplexing (OFDM) and have variable modulation types modulated (LFM) waveform is shown in Fig. 6]. While the
[4, 16, and 64 quadrature amplitude modulation (QAM)] lower peak transmit power enabled by pulse compression
depending on channel conditions and required data rates. is obviously beneficial from the perspective of radar
While the growing number of mobile users plays an hardware requirements and the reduction in interference
important role in this demand for spectrum, it is really the to other spectral users, it also involves a tradeoff with
proliferation of bandwidth-hungry streaming video that is respect to sensitivity for the radar as the ‘‘range sidelobes’’
the driving factor. The LTE and WiMAX systems operate in induced by filtering the transmitted waveform on receive
some spectrum currently used predominantly by radar. generates a form of self-interference with which the radar
With the availability of inexpensive RF and signal must contend (see Fig. 7). As such, a topic of ongoing
processing chip sets in S-band (specifically 3.4–3.7 GHz), scrutiny within the radar community has been the
the wireless community has become a competitor to radar
in this band. The potential for radar to interfere with LTE
and WiMAX is significant. Fig. 5 illustrates a 64-QAM
WiMAX constellation before [Fig. 5(a)] and during
[Fig. 5(b)] interference from an S-band radar, measured
in a bench test [11]. It is clear that, in the presence of radar
interference, the ability of the WiMAX receiver to recover
the transmitted symbol is severely degraded.
VI . CONTROLLING RADAR EMISSIONS

To sense the environment effectively, a radar must extract Fig. 6. LFM radar waveform; in this case, an ‘‘up-chirp’’ since the
the reflected target echoes from noise and interference. frequency sweeps higher during the pulse.

hop to another unoccupied frequency band (as long as such

allocations are available). With some loss in transmitted
power (and thus sensitivity), a bandpass filter could also be
placed after the power amplifier at the output of the
transmitter to suppress the OOB emissions [12]. However,
for radars that employ frequency agility, such a filter would
need to change its bandpass characteristics according to
the operating frequency, thereby significantly increasing
design complexity and cost. Alternatively, interference
subtraction of OOB emissions could conceivably be
realized by generating a replica of the OOB spectral
components produced by the power amplifier and then
subtracting this replica from the actual high-power
emission prior to launching from the antenna [13]. The
limitation of this approach is the calibration accuracy with
which these OOB replica components can be obtained and
subtracted [14]. For radars such as foliage penetration
Fig. 7. Pulse compressed LFM waveform reflected from two targets
systems in which the necessary wide bandwidth and
that have disparate radar cross sections; it is desired that this range
domain filter response to the transmitted waveform provides relatively low operating frequency (VHF/UHF) precludes
sufficient range resolution (narrow mainlobe) with minimal SNR loss complete avoidance of other in-band users, the radar
and low range sidelobe levels to separately identify high dynamic emission must possess spectral gaps of sufficient depth and
range targets in close proximity to one another. These effects are width to minimize the induced interference to other users.
determined by the design of the waveform and the pulse compression
receive filter.
Of course, these modifications to the emission structure
come at the price of significant degradation in sensitivity
due to greatly elevated range sidelobes [15]. Likewise, to
the degree that it is possible to characterize accurately the
development of optimal waveforms and receiver filters nonlinear aspects of the transmitter (primarily the power
that minimize these range sidelobes. Therefore, any amplifier), inversion of this characterization may facilitate
modification of the radar emission to address spectral the use of predistortion techniques for linearization, as
containment must likewise consider the impact such a discussed in [13]. All of these techniques remain topics of
change would have on the pulse compression range continued investigation.
sidelobes, and thus radar sensitivity. It is also important
to note that, for FM-based waveforms which are commonly
used in practice, a prominent attribute inducing spectral VI I. RADAR S USCEPTIBILITY TO
leakage is the pulse shape. A pulse envelope with fast rise INTERFERENCE
and fall times will introduce a wider spread of frequency A typical radar, such as a primary surveillance radar used
sidebands than the one with slow rise and fall times. for air traffic control, may have a peak pulse power from
However, as discussed earlier, operating with significant the transmitter of 20 kW and an antenna with a directive
amplitude modulation on the pulse will reduce the power gain of 30 dB, yielding an effective radiated power (ERP)
efficiency of the amplifier, which will deliver maximum of 73 dBW. However, the power levels detected in the
power when the signal is in saturation. For an LFM radar receiver are expected to be noise limited, while the
waveform with a pulse width of 64 s and a rather modest echo power from a distant target may be of the order of
time-bandwidth product of 64, Fig. 8 illustrates the spectral 110 dBW. In order to achieve the maximum detection
content of the waveform by itself (in blue) and with the performance for a given transmitter power, the radar will
inclusion of the pulse envelope. Clearly the rise/fall time of aim to achieve the best possible receiver sensitivity. It will
the pulse has a significant impact on spectral content, usually be more cost-effective to improve the receiver
though the seemingly easy fix of ‘‘slowing down’’ the rise/ sensitivity rather than attempting to increase the trans-
fall time is easier said than done, given the requirements on mitter power by the same margin. For this reason, long-
the transmitter hardware to achieve the high-power range radars will usually employ ultralow noise amplifiers
necessary to contend with two-way path loss and the at the receiver front–end.
associated loss in power efficiency that would be incurred. For many applications, radar receivers have traditionally
Within the trade space of radar performance and cost, been designed on the assumption that the radar is the
there are multiple existing approaches to limit the exclusive user of its allocated frequency spectrum and that
interference that radar emissions may induce upon other performance will be limited by thermal noise in the receiver.
spectrum users. If the radar observes another in-band (or It was assumed that interference from other radars in the
near-band) user, frequency avoidance can be employed to same band could usually be managed by careful frequency

susceptible to interference. These problems have been

found to occur even though both the radars and WiMAX
base stations are compliant with the appropriate national
EMC standards.
This problem has been widely investigated by national
regulatory authorities. In the United Kingdom, Ofcom has
undertaken its own studies and commissioned indepen-
dent research, including the effects of radar transmissions
on adjacent WiMAX systems. This has resulted in a
proposed S-band remediation plan [17]. One outcome from
this work has been to show that front–end filtering (e.g., a
low-loss passive bandpass filter in the main antenna feed)
in the radar receivers can considerably reduce the effects
of high-power adjacent communications systems, without
discernibly degrading radar performance.
In the United States, studies have been undertaken by
Fig. 8. Spectrum of an LFM waveform without the pulse envelope
the NTIA Institute for Telecommunications Sciences (ITS)
(blue) and with the pulse envelope (red). The pulse rise/fall time, which [18]. These studies investigated sources of electromagnetic
is very fast due to the switching nature of high-power amplifiers, has a interference (EMI), how EMI is manifested within the
significant impact on spectral content. radar receiver, and what technical solutions could be taken
to mitigate the effects. In particular, it was found that the
NEXRAD radars were susceptible to interference from
planning and site selection for ground-based radars (see also WiMAX. The NTIA investigation quantified power levels
Section V). With increasing demand for spectrum, there is of the interference in NEXRAD receivers under different
growing potential for interference from other users, either conditions. It also assessed the amount of decoupling
those sharing the same frequency allocation or from those required to reduce the observed interference levels to a
occupying adjacent frequencies. The effect on radar perfor- noninterference level, together with the amounts of
mance of interference from different sources can be quite decoupling required to mitigate the interference for
complex to define. For example, a study of this problem was various frequency and spatial separations. Because the
reported in [16]. This report concluded that interference at interference is partly cochannel, the application of
low duty cycles, such as from other radars, can often be additional filtering (see discussion above) within the radar
sustained at interference-to-noise (I/N) levels as high as 30– receiver will not totally eliminate the WiMAX EMI.
60 dB without degrading the receiver performance (i.e., its Further mitigating procedures include careful local
ability to detect small targets). An example of the effects of frequency planning (with a local frequency coordinator
such interference on a radar display can be seen in Fig. 5. or manager to maximize the frequency differences
Interference at higher duty cycles (above 1%–3%), such as between WiMAX transmitters and radars), management
from most communications signals, can cause target of the vertical beam angles of WiMAX base stations, and
detection losses to begin at I/N levels between 10 and control of WiMAX antenna locations and heights.
6 dB, dependent on radar type. It was noted in [16] that the Radars will increasingly have to coexist with other
loss in performance may be very insidious. No obvious effects users. This necessity will require improvements to radar
may be visible on the radar display but nevertheless target designs to minimize the effects of interference, together
detections may be lost. with careful frequency and site planning.
An area of current concern to radar users is the
allocation of spectrum to communication systems, such as
WiMAX and LTE, in the 2500–2690-MHz band. This band VIII. EMERGING TECHNOLOGY
is immediately adjacent to the radar allocations for air DEVELOPM ENTS
traffic control (ATC) radar, 2700–2900 MHz, and Beyond those technology developments already discussed,
maritime radars, 2900–3100 MHz. Also, the U.S. National pervasive spectral congestion will force future radar
Weather Service (NWS) Next Generation Weather Radars capabilities to rely increasingly upon emerging technolo-
(NEXRAD) operate in the 2700–3000-MHz band. It has gies in the areas of passive radar, waveform diversity,
been found that OOB signals (i.e., the spectral sidebands) bioinspired design, and cognitive processing. These are
from WiMAX base station transmitters can at times cause areas of current research and, therefore, we can only
cochannel interference to radar users in adjacent bands. speculate as to the impact and role they may have in
Equally, the OOB response of typical ATC and maritime helping to use the spectral resource more efficiently. Here
radar receivers, which had not been designed to cope with we introduce these concepts to illustrate their potential
adjacent band signals of this sort, may also render then utility.

the same as that of active radar; the main difference is that

the former has no control over or prior knowledge of the
sensing waveforms.
Passive radar has attracted a great deal of research
interest over the past decade, and numerous experimental
systems have been built and demonstrated [20], [21].
These include systems based on FM radio, analog and
digital television transmissions, cell phone base stations,
HF shortwave broadcasts, and satellite GPS transmissions.
It has been found that signals with digital modulation
formats are more suitable for passive bistatic radar since
the waveform is more noise-like and the performance does
not depend on the type of modulation (for example, speech
or music). These types of system have demonstrated
detection and tracking of aircraft to ranges beyond 200 km
[22], with Lockheed Martin [23] THALES [24], and Selex
Fig. 9. Passive bistatic radar. having developed commercial systems. However, it is
difficult to guarantee performance levels, and, hence, they
tend to be viewed more as a supplement to existing radar
A. Passive Radar systems. Whether bistatic radar can be developed to take
Passive radar (or passive bistatic radar) is the name on the role of a primary sensor is far from certain.
used to describe radar that exploits other transmissions Taking this further, it may also be desirable in the
(communications, broadcast, or radionavigation) rather future to design the signals of passive bistatic radar
than having its own dedicated radar transmitter. The illuminators so that they not only fulfill their primary
passive receiver is thus located separately from the function but also have favorable waveform properties for
transmitter (see Fig. 9). Since such transmissions are radar purposes. This has been termed ‘‘commensal
often high power and favorably located within geograph- radar’’Vliterally ‘‘at the same table’’Vand is an example
ical regions of interest, the coverage of such passive of the sort of approaches that will be necessary as the
sensing systems can be substantial. Further, the cost of the spectrum problem becomes worse.
transmitter is avoided, making the system relatively
inexpensive, and the receiving system may be completely B. Waveform Diversity
undetectable, which may be advantageous for defense It is possible to use emissions far more efficiently than is
applications. Despite these attractive characteristics, since currently the case. Modern digital technology means that it
the waveforms are not fundamentally designed for radar is now feasible to generate precise, sophisticated wide-
purposes, and are often time varying, performance in a bandwidth radar waveforms that can be varied, potentially
radar context can be far from ideal. In addition, it is even on a pulse-by-pulse basis. This capability forms the
necessary to choose the right waveforms present in the foundation for what has become known as ‘‘waveform
environment and to process them in the correct way [19]. diversity’’ [25], which includes the optimization of wave-
It is generally necessary to have one antenna dedicated to forms based on mission requirements and prior knowledge
acquiring the direct path signal from the selected emitter of the spectral environment; greater exploitation of
and another separate antenna, having a spatial null, or at available degrees of freedom for both transmit waveform
least little gain, in the direction of the direct path emitter, design and receive filtering; and the general convergence of
from which the radar reflections are acquired. The direct electromagnetic, systems engineering, and signal proces-
path signal therefore serves as the reference waveform for sing requirements and capabilities.
pulse compression filtering of the reflected echoes. For example, it is possible to design radar waveforms that
It is important to note that passive radar is fundamen- have spectral nulls at particular frequencies or frequency
tally different from radiometry, though both share the trait bands [26], [27], and potentially even to adapt these
of operating in a passive manner. Radiometry performs waveforms dynamically in response to a changing interfer-
remote sensing of physically separated sources like Earth ence environment (the latter being an example of a cognitive
(from a satellite) or astronomical objects, usually to extract system). Likewise, recent schemes [28] for the implemen-
various parameters regarding scientific phenomena such as tation and optimization of polyphase-coded FM (PCFM)
temperature and moisture. In contrast, passive radar waveforms based on the continuous phase modulation
measures the scattering of signals from various objects of (CPM) framework demonstrate new avenues for the design
interest, albeit by leveraging the emissions from other RF/ of radar emissions with enhanced spectral containment.
microwave sources such as radio, television, other radars, These latter techniques also highlight the prospective
etc. In other words, the purpose of passive radar remains benefits to incorporating system-level effects such as

Fig. 10. Two constant amplitude, continuous waveforms are driven

into separate saturated power amplifiers (needed for high power
efficiency) and then combined in the 180 coupler sum channel.
Their relative phases control the amplitude of the resulting emission
eðtÞ. A difference channel (not shown) collects the power that is not
radiated into a terminated load.
transmitter distortion directly into the waveform design pro- Fig. 11. Pulse envelope of measured (received) and theoretical
reference eðtÞ in the decibel scale (top) and normalized amplitude
cess (as opposed to performing predistortion after the fact). (bottom). In the decibel scale, the impact of imperfect channel
Such hardware-in-the-loop design schemes enable a holistic calibration is revealed at the beginning and end of the measured
view of the radar so as to address jointly the waveform per- pulse, while some amplitude droop is observed in the normalized
formance and its actual emitted spectral footprint. amplitude scale.
A particular open problem with regard to radar
emission control is the spectral spreading induced by the
rapid rise time and fall time of each transmitted pulse. and radar waveform design to address the demands on
High-power transmitters do not presently have a means to radar spectral containment. For the radar application,
‘‘slow down’’ the rise/fall time as they behave more like a additional considerations include ensuring high fidelity (to
switch that is either on or off. Furthermore, any action minimize mismatch-inducing distortion that degrades
taken to alter the spectral content of the radar emissions sensitivity, where the dynamic range may be several tens
also directly impacts the waveform structure and, by of decibels), achieving sufficient energy on target (mini-
extension, radar sensitivity. Within this context of holistic, mal deviation from maximum power to maintain receive
hardware-in-the-loop design through the use of lineariza- SNR), and operating with bandwidths of tens to hundreds
tion techniques [14], it may potentially provide a means to of megahertz or more (many linearization methods require
slow down the rise/fall time. For example, using the 180 switching frequencies that are several times the opera-
coupler approach depicted in Fig. 10, which is a form of tional bandwidth).
outphasing [13] otherwise known as linear amplification
with nonlinear components (LINC), it was recently shown
[29] that a pulse amplitude tapering can be obtained (see
Fig. 11) while still operating the power amplifiers in
saturation. The resulting improvement in spectral con-
tainment (Fig. 12) demonstrates about 15-dB improvement
in spectral containment, though issues such as heat
management (from the power not emitted), calibration,
and signal-to-noise ratio (SNR) losses remain open issues.
Besides the LINC approaches, there are many more
forms of linearizing methods for power-efficient opera-
tion, including the Kahn technique, envelope tracking, the
Doherty technique, feedback and feedforward lineariza-
tion, and others with new techniques continuing to emerge
(see [30] and references therein for a detailed discussion of
linearization methods). Generally speaking, these methods
rely on a combination of two or more amplifiers that focus
Fig. 12. Spectrum of a nonlinear FM (NLFM) waveform optimized for a
on specific attributes of the signal being amplified and
standard transmitter architecture (higher trace) and optimized
digital signal processing (DSP) to separate the signal into specific to the hardware in the LINC architecture (shaded spectrum),
the requisite components. Clearly, joint research is needed where the latter slows down the pulse rise/fall time while using
at the intersection of advanced transmitter configurations saturated power amplifiers. Vertical increments are 10 dB.

The increase in interference encountered by the radar of the human eye [49] (a passive sensor) as means to
may be addressed in part by expanding the purview of achieve spatial modulation for active radar emissions [50].
adaptive interference cancellation to incorporate addition- From a spectrum usage standpoint, the key takeaway
al degrees of freedom. For example, an airborne radar from observing nature is that the animal kingdom can far
performing ground moving target indication (GMTI) will surpass our best technological capabilities and yet do so
use the echoes collected from M pulses at N antenna with what appears to be rather nonoptimal ‘‘equipment’’
elements to construct an adaptive filtering structure that [51], at least when taken componentwise. Nature takes the
operates jointly in the spatial and Doppler domains so as to notion of ‘‘the whole being better than the sum of parts’’ to
suppress interference from ground echoes (i.e., clutter) an extreme we still cannot fully comprehend. So the lesson
and possibly jamming (otherwise known as space–time to be learned is that we too must take a holistic view when
adaptive processing (STAP) [31]). Instead of being additive designing systems so that the individual ‘‘components’’ of
as M þ N, this joint filtering approach provides M N electromagnetic, systems engineering, and signal proces-
adaptive degrees of freedom for interference cancellation. sing work in harmony, both within a given radar system
By extension, ongoing research efforts are exploring how and in its interaction with all other spectrum users. For
the different dimensions of range (the waveform domain), example, codesign of the radar transmitter and waveforms
space, Doppler, polarization, and frequency can be may minimize the amount of spectral leakage while still
combined to facilitate greater design freedom on transmit optimizing mission requirements for search, tracking, and
(e.g., [32]–[37]) and interference suppression capability imaging modes. Furthermore, taking a Baldwinian evolu-
on receive (e.g., [38]–[40]). Furthermore, leveraging the tionary perspective [52], it is also clear that systems must
publicly available knowledge of different RF systems (e.g., be designed specifically for the spectral environment in
wireless communication standards [41]) and their defined which they will operate, which in turn results in further
signal structure across these different dimensions could shaping of that environment. Factoring in the exorbitant
also enhance the radar’s ability to contend with uninten- cost of deploying new radar systems, we also must be
tional spectral encroachment. cognizant of the tradeoff between niche-optimized systems
and being sufficiently flexible to adapt when the environ-
ment or need changes.
C. Bioinspired Design
With many million years head start, it is no surprise D. Cognitive Approaches
that echo-locating animals such as bats and dolphins far The notion of cognitive radar can be viewed from two
surpass our active sensing capabilities [42]. For example, different perspectives: as the evolution of bioinspired
bats use sonar to sense, navigate, and communicate in a control systems to higher level decision making [53], [54]
simultaneous manner through the same transmit aperture or as the natural outgrowth of knowledge-aided sensor
(their mouth) and possess only a two-element antenna signal processing [55]. Regardless of its roots, in the most
array on receive (their ears) and yet can successfully hunt general sense, cognitive radar is essentially the application
for food in a large swarm of other bats while not colliding of Bayesian learning, through the use of prior knowledge
with their environment or each other. Clearly bats leverage and feedback, to facilitate the development of autonomous
a very advanced form of waveform diversity that decision making within the radar.
encompasses a form of simultaneous multimode emission If prior knowledge of the spectral environment exists,
coupled with very sophisticated and highly specialized it can also be exploited. This approach has enabled
receive signal processing [43], [44]. A bat can change the cognitive radio to make great strides in recent years.
nature of this emission according to the particular However, cognitive radio has tended to concentrate on
information being sought (e.g., searching for available radio communication rather than considering the problem
prey versus tracking specific prey) while dolphins exploit a in its entirety. A more comprehensive approach would be
form of pulse-to-pulse waveform diversity to distinguish to map out spectrum usage in terms of spectral, temporal,
linear scattering from nonlinear scattering [45], [46] and spatial occupancy of all emitters and exploit this total
(which proves useful in bubble-rich shallow waters). ‘‘spectral landscape’’ in cognitive-type approaches. This
Likewise, bats may leverage echo location as a form of perspective would enable cognitive approaches to embrace
echoic flow for navigation [47], [48] in a manner similar to all emitters in an intelligent fashion. For example, most
how other creatures use optic flow to navigate by vision. radar systems scan at a rate of less than one rotation per
Further, while radar signal processing is linear and second. Most power is concentrated in the main beam
becoming predominantly digital, biological cognition, to whose width may only be a few degrees. Thus, at any one
the degree that it is actually understood, is clearly analog time, the vast majority of the swept volume (typically 90%)
and considered to be rather nonlinear. Finally, the is not being used by the radar. As this operation is fully
potentially useful lessons to be learned from nature determinable in advance, there is considerable opportunity
regarding radar are also not solely limited to biosonar, as for further improving spectrum usage and possibly
evidenced by recent work to mimic the rapid movements spectrum sharing. This form of approach clearly offers

efficiency gains in spectrum use without sacrificing perfor- monitor and track severe weather; it is a technological
mance, thus making it an attractive topic of future study. cornerstone for defense and homeland security; and in the
Existing radar procedures such as automated frequency not-too-distant future, it may even be integral to the
agility to avoid other spectral users and dynamic time- establishment of networks of driverless automobiles [61].
division resource allocation to enable different sensing Yet, to accomplish all of these and numerous other vital
(and possibly other) modes to share the same antenna [56] tasks, radar requires access to spectrum.
can be considered as early examples of cognitive systems. All users have a need for greater bandwidth, and the
However, ongoing research is also exploring more radical only thing that can be said with certainty is that the
modifications such as by leveraging the burgeoning work problem is only going to get worse. Yet, if spectrum usage
in waveform diversity to enable the radar to design were measured at a given point as a function of frequency,
waveforms ‘‘on the fly’’ according to the observed spectral time, space, and polarization, it would certainly be found
environment and mission requirements (e.g., [57]–[59]) that the spectrum is currently not being used efficiently.
through the use of complex feedback mechanisms and Therefore, there is great potential for approaches aimed at
automated decision making. In other words, viewing active using the spectrum in an efficient and dynamically
sensing as a question-and-answer exercise, how can we controlled manner.
enable the radar to select the best questions (i.e., The regulatory framework has thus far taken a
waveforms) so as to obtain the best answers (given the relatively conservative approach. However, it is important
spectral usage constraints) in real time? to have a proper quantitative understanding of the effect of
Besides the sensor-centric area of waveform diversity, interference of one service upon another in order to adopt
cognitive radar research is also building from previous appropriate regulation measures, rather than taking the
work in cognitive psychology and artificial intelligence to view that no service should ever occupy the same part of
mimic our own attributes of learning, memory, attention, the spectrum as any other.
and intelligence [60], all with the goal of making the radar A number of novel radar technology approaches have
‘‘smarter.’’ Compared to the relative ease with which many been described, including improvements to the spectral
animals can sense and interact with their environments, it purity of transmitters; intelligent, cognitive approaches to
is clear that we are only just beginning to realize the dynamic frequency allocation; passive sensing based on the
potential of artificial cognitive sensor systems, though emissions of other RF applications; and even through
continued research is necessary to quantify the likely gains learning to mimic the behavior of echo-locating animals.
that could be accrued. The same digital technology that has enabled the
tremendous growth in communication capabilities is also
facilitating a new paradigm in radar functionality, both
IX. CONCLUSION through the generation of precise broadband waveforms
As with many of the facets of modern life, radar is such and the development of new receive processing methods.
that most people would only become aware of its As this cohabitation of the RF spectrum continues, even
fundamental importance if it were absent. Radar enables further advances in technology will be needed to contend
the control and management of air traffic; it is used to with the growing congestion. h
REFERENCES http://www.ofcom.org.uk/static/spectrum/ [14] A. Choffrut, B. D. Van Veen, and J. H. Booske,

Airport_Deployment_Study.pdf. ‘‘Minimizing spectral leakage of nonideal
[1] D. McQueen, ‘‘The momentum behind LTE LINC transmitters by analysis of component
adoption,’’ IEEE Commun. Mag., vol. 47, no. 2, [8] M. A. Richards, J. A. Scheer, and W. A. Holm,
Eds., Principles of Modern Radar. London, impairments,’’ IEEE Trans. Veh. Technol.,
pp. 44–45, Feb. 2009. vol. 56, no. 2, pp. 445–458, Mar. 2007.
U.K.: IET, 2010.
[2] M. J. Marcus, ‘‘Spectrum policy for radio [15] M. E. Davis, ‘‘Frequency allocation challenges
spectrum access,’’ Proc. IEEE, vol. 100, no. 5, [9] International Telecommunication Union
(ITU), ‘‘Manual of radio regulations,’’ 2012. for ultra-wideband radars,’’ IEEE Aerosp.
pp. 1685–1691, May 2012. Electron. Syst. Mag., vol. 28, no. 7, pp. 12–18,
[Online]. Available: http://www.itu.int/pub/
[3] United States Frequency Allocations. R-REG-RR-2012. Jul. 2013.
[Online]. Available: http://www.ntia.doc.gov/ [16] F. H. Saunders et al., ‘‘Effects of RF
files/ntia/publications/spectrum_wall_ [10] NTIA, ‘‘Manual of Regulations and
Procedures for Federal Radio Frequency interference on radar receivers,’’ NTIA Rep.
chart_aug2011.pdf TR-06-444, Sep. 2006.
Management (Red Book),’’ May 2013.
[4] IEEE Standard Letter Designations for [17] Ofcom Information Update, ‘‘Coexistence of
Radar-Frequency Bands, IEEE Std. 521-2002. [11] L. Cohen, E. Daly, J. DeGraaf, and K. Sheff,
‘‘Mitigation of radar interference with WiMax S band radar systems and adjacent future
[5] International Telecommunication Union systems,’’ in Proc. Int. Waveform Diversity services,’’ Dec. 11, 2009.
(ITU), ‘‘2012 World Radiocommunication Design Conf., Niagara Falls, ON, Canada, [18] F. H. Sanders, R. L. Sole, J. E. Carroll,
Conference: Agenda and references.’’ Aug. 2010, DOI: 10.1109/WDD.2010. G. S. Secrest, and T. L. Allmon, ‘‘Analysis and
[Online] http://www.itu.int/dms_pub/itu-r/ 5592449. resolution of RF interference to radars
oth/04/04/R0C040000070001PDFE.pdf . [12] D. R. Jachowski and C. Rauscher, operating in the band 2700–2900 MHz from
[6] H. D. Griffiths, C. J. Baker, C. J. , and ‘‘Frequency-agile bandstop filter with tunable broadband communication transmitters,’’
D. Adamy, Stimson’s Introduction to Airborne attenuation,’’ in Proc. IEEE Int. Microw. Symp., U.S. Dept. Commerce, Washington, DC, USA,
Radar, 3rd ed. Raleigh, NC, USA: Scitech, Boston, MA, Jun. 2009, pp. 649–652. NTIA Tech. Rep. TR-13-490, Oct. 2012.
2014. [19] H. D. Griffiths and C. J. Baker, ‘‘Passive
[13] F. H. Raab et al., ‘‘Power amplifiers and
[7] Ofcom, ‘‘Airport deployment study,’’ Final transmitters for RF and microwave,’’ IEEE bistatic radar and waveform diversity,’’ in
Rep. Ref. MC/045, 2011. [Online]. Available: Trans. Microw. Theory Tech., vol. 50, no. 3, Waveform Design and Diversity for Advanced
pp. 814–826, Mar. 2002.

Radar Systems, F. Gini, A. De Maio, and Top. Signal Process., vol. 4, no. 1, pp. 66–78, paradigm of ecological acoustics,’’ J. Compar.
L. Patton, Eds. Stevenage, U.K.: IET, 2012. Feb. 2010. Physiol. A, vol. 176, pp. 347–354, 1995.
[20] N. J. Willis and H. D. Griffiths, Advances in [34] P. F. Sammartino, C. J. Baker, and [48] G. E. Smith and C. J. Baker, ‘‘Echoic flow for
Bistatic Radar. Raleigh, NC, USA: Scitech, H. D. Griffiths, ‘‘Frequency diversity MIMO radar and sonar,’’ Electron. Lett., vol. 48,
2007. techniques for radar,’’ IEEE Trans. Aerosp. no. 18, pp. 1160–1161, Aug. 2012.
[21] M. Cherniakov, Bistatic Radar: Emerging Electron. Syst., vol. 49, no. 1, pp. 201–222, [49] M. Rolfs, ‘‘Microsaccades: Small steps on a
Technology. New York, NY, USA: Wiley, Jan. 2013. long way,’’ Vis. Res., vol. 49, pp. 2415–2441,
2008. [35] D. Giuli, ‘‘Polarization diversity in radars,’’ 2009.
[22] P. E. Howland, D. Maksimiuk, and G. Reitsma, Proc. IEEE, vol. 74, no. 2, pp. 245–269, [50] S. D. Blunt, P. McCormick, T. Higgins, and
‘‘FM radio based bistatic radar,’’ IEE Feb. 1986. M. Rangaswamy, ‘‘Spatially-modulated radar
Proc. Radar Sonar Navig., vol. 152, no. 3, [36] S. Gogineni and A. Nehorai, ‘‘Polarimetric waveforms inspired by fixational eye
pp. 107–115, Jun. 2005. MIMO radar with distributed antennas for movement,’’ in Proc. IEEE Radar Conf.,
[23] J. Baniak, G. Baker, A. M. Cunningham, and target detection,’’ IEEE Trans. Signal Process., Cincinnati, OH, USA, May 2014, DOI: 10.
L. Martin, ‘‘Silent sentry passive surveillance,’’ vol. 58, no. 3, pp. 1689–1697, Mar. 2010. 1109/RADAR.2014.6875719.
presented at the Aviation Week and Space [37] Y. I. Abramovich, G. J. Frazer, and [51] W. W. L. Au and S. W. Martin, ‘‘Why dolphin
Technology, Jun. 7, 1999. B. A. Johnson, ‘‘Principles of mode-selective biosonar performs so well in spite of mediocre
[24] Air-Defense.net, ‘‘Radar passif Thales MIMO OTHR,’’ IEEE Trans. Aerosp. Electron. ‘equipment’,’’ IET Radar Sonar Navig., vol. 6,
Homeland Alerter 100.’’ [Online] http:// Syst., vol. 49, no. 3, pp. 1839–1868, Jul. 2013. no. 6, pp. 566–575, Jul. 2012.
www.air-defense.net/forum/index.php? [38] T. Higgins, K. Gerlach, A. K. Shackelford, and [52] D. J. Depew, Evolution and Learning: The
topic=11376.0 . S. D. Blunt, ‘‘Aspects of non-identical multiple Baldwin Effect Reconsidered. Cambridge,
[25] Principles of Waveform Diversity and Design, pulse compression,’’ in Proc. IEEE Radar MA, USA: MIT Press, 2003.
M. Wicks, E. Mokole, S. Blunt, R. Schneible, Conf., Kansas City, MO, USA, May 2011, [53] S. Haykin, ‘‘Cognitive dynamic systems:
and V. Amuso, Eds. Raleigh, NC, USA: DOI: 10.1109/RADAR.2011.5960666. Radar, control, radio,’’ Proc. IEEE, vol. 100,
Scitech, 2010. [39] T. Higgins, S. D. Blunt, and A. K. Shackelford, no. 7, pp. 2095–2013, Jul. 2012.
[26] K. Gerlach, M. R. Frey, M. J. Steiner, and ‘‘Time-range adaptive processing for pulse [54] S. Haykin, Cognitive Dynamic Systems:
A. Shackelford, ‘‘Spectral nulling on transmit agile radar,’’ in Proc. Int. Waveform Diversity Perception-Action Cycle, Radar, Radio.
via nonlinear FM radar waveforms,’’ IEEE Design Conf., Niagara Falls, ON, Canada, Cambridge, U.K.: Cambridge Univ. Press,
Trans. Aerosp. Electron. Syst., vol. 47, no. 2, Aug. 2010, DOI: 10.1109/WDD.2010. 2012.
pp. 1507–1515, Apr. 2011. 5592394.
[55] J. R. Guerci, Cognitive Radar: The
[27] L. K. Patton, S. W. Frost, and B. D. Rigling, [40] W. Roberts, J. Li, P. Stoica, Knowledge-Aided Fully Adaptive Approach.
‘‘Efficient design of radar waveforms for T. Yardibi, and F. A. Sadjadi, ‘‘MIMO Reading, MA, USA: Artech House, 2010.
optimized detection in coloured noise,’’ IET radar angle-range-Doppler imaging,’’ in
[56] G. C. Tavik et al., ‘‘The advanced
Radar Sonar Navig., vol. 6, no. 1, pp. 21–29, Proc. IEEE Radar Conf., Pasadena, CA, USA,
multifunction RF concept,’’ IEEE Trans.
Jan. 2012. May 2009, DOI: 10.1109/RADAR.2009.
Microw. Theory Tech., vol. 53, no. 3,
4976950.
[28] J. Jakabosky, S. D. Blunt, M. R. Cook, J. Stiles, pp. 1009–1020, Mar. 2009.
and S. A. Seguin, ‘‘Transmitter-in-the-loop [41] The 3rd Generation Partnership Project
[57] M. R. Bell, ‘‘Information theory and radar
optimization of physical radar emissions,’’ in (3GPP). [Online]. Available: www.3gpp.org
waveform design,’’ IEEE Trans. Inf. Theory,
Proc. IEEE Radar Conf., Atlanta, GA, USA, [42] W. E. Conner, ‘‘An acoustic arms race,’’ Amer. vol. 39, no. 5, pp. 1578–1597, Sep. 1993.
May 2012, DOI: 10.1109/RADAR.2012. Sci., vol. 101, pp. 202–209, May–Jun. 2013.
[58] S. U. Pillai, H. S. Oh, D. C. Youla, and
6212260. [43] A. Balleri, H. D. Griffiths, C. J. Baker, J. R. Guerci, ‘‘Optimum transmit-receive
[29] L. Ryan, J. Jakabosky, S. D. Blunt, C. Allen, K. Woodbridge, and M. W. Holderied, design in the presence of signal-dependent
and L. Cohen, ‘‘Optimizing polyphase-coded ‘‘Analysis of acoustic echoes from a interference and channel noise,’’ IEEE Trans.
FM waveforms within a LINC transmit bat-pollinated plant species: Insight into Inf. Theory, vol. 46, no. 2, pp. 577–584,
architecture,’’ in Proc. IEEE Radar Conf., strategies for radar and sonar target Mar. 2000.
Cincinnati, OH, USA, May 2014, DOI: 10. classification,’’ IET Radar Sonar Navig., vol. 6,
[59] N. A. Goodman, P. R. Venkata, and
1109/RADAR.2014.6875706. no. 6, pp. 536–544, Jul. 2012.
M. A. Neifeld, ‘‘Adaptive waveform design
[30] M. Eron, B. Kim, F. Raab, R. Caverly, and [44] J. A. Simmons and J. E. Gaudette, ‘‘Biosonar and sequential hypothesis testing for target
J. Staudinger, ‘‘The head of the class,’’ IEEE echo processing by frequency-modulated recognition with active sensors,’’ IEEE J. Sel.
Microw. Mag., vol. 12, no. 7, pp. S16–S33, bats,’’ IET Radar Sonar Navig., vol. 6, no. 6, Top. Signal Process., vol. 1, no. 1, pp. 105–113,
Dec. 2011. pp. 556–565, Jul. 2012. Jun. 2007.
[31] W. L. Melvin, ‘‘A STAP overview,’’ IEEE [45] T. G. Leighton, P. R. White, and D. C. Finfer, [60] S. Haykin, Y. Xue, and P. Setoodeh,
Aerosp. Electron. Syst. Mag., vol. 19, no. 1, ‘‘Hypotheses regarding exploitation of bubble ‘‘Cognitive radar: Step toward bridging the
pp. 19–35, Jan. 2004. acoustics by cetaceans,’’ in Proc. Eur. gap between neuroscience and engineering,’’
[32] T. Higgins, T. Webster, and A. K. Shackelford, Conf. Underwater Acoust., Paris, France, Proc. IEEE, vol. 100, no. 11, pp. 3102–3130,
‘‘Mitigating interference via spatial and Jun./Jul. 2008, pp. 77–82. Nov. 2012.
spectral nulls,’’ in Proc. IET Int. Radar Conf., [46] T. G. Leighton et al., ‘‘Radar clutter [61] R. Stevenson, ‘‘A driver’s sixth sense,’’
Glasgow, U.K., Oct. 2012, DOI: 10.1049/cp. suppression and target discrimination using IEEE Spectrum, vol. 48, no. 10, pp. 50–55,
2012.1562. twin inverted pulses,’’ Proc. Roy. Soc. A, Oct. 2011.
[33] D. Fuhrmann, P. Browning, and vol. 469, no. 2160, Oct. 2013, DOI: 10.1098/
M. Rangaswamy, ‘‘Signaling strategies for the rspa.2013.0512.
hybrid MIMO phased-array radar,’’ IEEE J. Sel. [47] D. N. Lee, J. A. Simmons, P. A. Saillant, and
F. Bouffard, ‘‘Steering by echolocation: A

ABOUT THE AUTHORS

Hugh Griffiths (Fellow, IEEE) was educated at 1978, respectively, and the Ph.D. degree in mathematics from the Georgia
Hardye’s School, Dorchester, U.K., and Keble Institute of Technology in 1982.
College, Oxford University, Oxford, U.K., where He was Head of the Surveillance Technology Branch of the Radar
he received the M.A. degree in physics in 1978. He Division, Naval Research Laboratory, Washington, DC, USA. He is the U.S.
also received the Ph.D. and D.Sc. (Eng) degrees Navy lead for the Tri-Service Radar Symposia, Chair of NATO’s Sensors
from the University of London, London, U.K., in and Electronic Technology Panel, and a government liaison for USNC
1986 and 2000, respectively. URSI.
He holds the Thales/Royal Academy of Engi- Dr. Mokole is a member of the IEEE Aerospace and Electronic Systems
neering Chair of RF Sensors at the University Society (AESS) Radar Systems Panel.
College London, London, U.K. He chaired a Work-
ing Party for the U.K. Ministry of Defence on Radar Operation in Chris Baker (Senior Member, IEEE) was educated
Spectrally Congested Environments. at Ravenswood School, Kent, U.K. and the Univer-
Dr. Griffiths is a Fellow of the Royal Academy of Engineering, and sity of Hull, Yorkshire, U.K., where he graduated
served as President of the IEEE Aerospace and Electronic Systems Society with a Ph.D. degree in laser physics in 1984.
for 2012–2013. He is the Ohio State Research Scholar and
Professor of Integrated Sensing Systems at the
Ohio State University, Columbus, OH, USA.
Lawrence Cohen (Senior Member, IEEE) received the B.S. degree in
Dr. Baker is a Fellow of the Institution of
electrical engineering from The George Washington University, Washing-
Engineering and Technology (IET) and a member
ton, DC, USA, in 1975 and the M.S. degree in electrical engineering from
of the IEEE Aerospace and Electronic Systems
the Virginia Polytechnic Institute and State University, Blacksburg, VA,
Society (AESS) Radar Systems Panel.
USA, in 1994.
He is an Electronics Engineer at the Radar Division, Naval Research
Laboratory, Washington, DC, USA. In this capacity, he assists various U.S. Mike Wicks (Fellow, IEEE) received the B.S.E.E.
Navy commands with problem solving related to radar and wireless from Rensselaer Polytechnic Institute, Troy, NY,
systems. USA, and the Ph.D. in electrical engineering from
Dr. Cohen is a member of the IEEE Electromagnetic Compatibility Syracuse University, Syracuse, NY, USA, in 1995.
Society’s Board of Directors. He retired from the U.S. Air Force Research
Laboratory Sensors Directorate after a 30 year
career, and joined the University of Dayton, Dayton,
Simon Watts (Fellow, IEEE) graduated from the KS, USA, in 2011.
University of Oxford, Oxford, U.K., in 1971 and Dr. Wicks was the recipient of the 1998 IEEE
received the M.Sc. degree from the University of Fred Nathanson Award, the winner of the 2009
Birmingham, Birmingham, U.K., in 1972 and the IEEE Warren D. White Award, and the winner of the 2013 IEEE Picard
Ph.D. degree from CNAA in 1987. Medal.
He was formerly Deputy Scientific Director of
Thales U.K., Defence Mission Systems, and is a Shannon Blunt (Senior Member, IEEE) received
Visiting Professor in the Department of Electronic the B.S., M.S., and Ph.D. degrees in electrical
and Electrical Engineering, University College engineering from the University of MissouriV
London, London, U.K. He was appointed MBE in Columbia, Columbia, MO, USA, in 1999, 2000,
1996 for services to the U.K. defence industry. and 2002, respectively,
Dr. Watts is a member of the IEEE Aerospace and Electronic Systems He is an Associate Professor in the Electrical
Society (AESS) Radar Systems Panel and a Fellow of the Royal Academy Engineering & Computer Science Department,
of Engineering. University of Kansas, Kansas City, KS, USA, where
he is also Director of the Radar Systems Labora-
Eric Mokole (Fellow, IEEE) received the B.S. degree in applied tory (RSL). He is active in the development of
mathematics from New York University, New York, NY, USA, in 1971, spectrally clean radar modulation schemes.
the M.S. degree in mathematics from Northern Illinois University, DeKalb, Dr. Blunt is a member of the IEEE Aerospace and Electronic Systems
IL, USA, in 1973, dual M.S. degrees in physics and applied mathematics Society (AESS) Radar Systems Panel, and the recipient of the 2012 IEEE
from the Georgia Institute of Technology, Atlanta, GA, USA, in 1976 and Fred Nathanson Memorial Radar Award.

Radar Spectrum

Uploaded by

Copyright:

Available Formats

Radar Spectrum

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Radar Spectrum

Uploaded by

Copyright:

Available Formats

CONTRIBUTED

Radar Spectrum Engineering

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 85

86 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 87

88 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 89

90 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 91

transmit that is fixed in frequency and goes down to 40 dB

92 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 93

VI . CONTROLLING RADAR EMISSIONS

94 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

hop to another unoccupied frequency band (as long as such

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 95

susceptible to interference. These problems have been

96 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

the same as that of active radar; the main difference is that

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 97

Fig. 10. Two constant amplitude, continuous waveforms are driven

98 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 99

REFERENCES http://www.ofcom.org.uk/static/spectrum/ [14] A. Choffrut, B. D. Van Veen, and J. H. Booske,

100 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

Vol. 103, No. 1, January 2015 | Proceedings of the IEEE 101

ABOUT THE AUTHORS

102 Proceedings of the IEEE | Vol. 103, No. 1, January 2015

You might also like