Concepts of MIMO Radar and Their

Advantages and Disadvantages over

Phased Array
Major Amit Srivastava, Mtech (RF and Microwave Engineering), IIT Kharagpur

Abstract-Multiple-input multiple-output (MIMO) concept for MIMO radar, e.g., the transmission of
radar is an advanced type of phased array radar multiple orthogonal waveforms from different antennas,
employing digital receivers and waveform generators is usually referred to as the waveform diversity [3],[4].
distributed across the aperture. MIMO radar signals Consequently, the waveform design and optimization
propagate in a fashion similar to multistatic radar. has been the main focus of the research in MIMO radar.
However, instead of distributing the radar elements Many approaches to MIMO radar have been developed
throughout the surveillance area, antennas are that revolve around the main idea of exploiting the
closely located to obtain better spatial resolution, waveform diversity. Based on the array configurations
Doppler resolution, and dynamic range. MIMO used, MIMO radars can be classified into two main
radar may also be used to obtain low-probability-of- types. The first type uses widely separated
intercept radar properties. In a traditional phased transmit/receive antennas to capture the spatial diversity
array system, additional antennas and related of the target’s radar cross section (RCS) [4]. This type
hardware are needed to improve spatial resolution. assumes an extended target model and, therefore, takes
MIMO radar systems transmit mutually orthogonal advantages of the properties of the associated spatially-
signals from multiple transmit antennas, and these distributed signal model. In this case, the waveform
waveforms can be extracted from each of the receive diversity is similar to the multi-path diversity concept in
antennas by a set of matched filters. wireless communications over fading channels.
According to this concept, signals transmitted over
I. INTRODUCTION multiple fading links/channels can be decoded reliably at
the receiver due to the fact that it is unlikely that all

R ADAR technology has been continuously

developing over the last 70 years starting from the
late 1930s when radar was first invented for defence
links/channels undergo unfavorable fading conditions
simultaneously. Similarly, the waveform diversity in
radars enables reliable detection due to the fact that it is
unlikely that all waveforms suffer from scintillation
applications. The desire for new more advanced radar
effects simultaneously. The second MIMO radar type
technologies has been driven and is dictated by radar’s
employs arrays of closely spaced transmit/receive
ubiquitous applicability ranging from micro-scale radars
antennas to cohere a beam towards a certain direction in
applied in biomedical engineering to macro-scale radars
space [5]. In this case, the target is usually assumed to be
used in radio astronomy. To date, previous
in the far-field. Also, it is often assumed that the
developments in radar were based on the idea that the
transmitted signals are narrowband and, therefore, the
signal can be processed coherently at the
point source signal model is commonly presumed. The
transmit/receive antenna arrays if the signal coherency is
waveform diversity, in this case, boils down to
preserved. The corresponding radar technique is well
increasing the virtual aperture of the receive array due to
known under the name phased-array radar [1], [2]. In the
the fact that multiple independent waveforms are
last decade, the development of a new radar paradigm
received by the same receive array.
that is best known under the title multiple-input
multiple-output (MIMO) radar has become the focus of
intensive research. The essence of the MIMO radar II. PRINCIPLES OF MIMO RADAR
concept is to employ multiple antennas for emitting
several orthogonal waveforms and multiple antennas for What is MIMO Radar? MIMO refers to radar with
receiving the echoes reflected by the target. The enabling multiple TX and multiple RX antennas. We describe
three aspects to the understanding of within aperture
MIMO radar. First is the concept of the synthetic, or
“phantom”, element. We then show how filled apertures
can be synthesised using this notion, and how multiple
beams are formed Some of the essential features of the
radar waveforms and associated signal processing
required for MIMO radar are then described.

A. THE “PHANTOM” ANTENNA ELEMENT

Consider the simple bistatic antenna pair shown in
Fig. 1. The geometry shows how, for far-field targets,
signals from the bistatic pair would be identical to
signals received from a monostatic TRx element Fig. 2. One Dimensional Synthesised Linear Array of
positioned mid point between the transmit and receive Phantom Elements.
elements of the bistatic pair. It is clear that a reflected
signal arriving with a specific delay must have arisen
from some point on the surface of an ellipsoid with foci
at the Tx and Rx element positions. It is also clear that
for very distant targets the ellipsoid converges to a
sphere centred on the mid-point between the elements,
forming the phantom TRx element.

Fig. 3. Two Dimensional Synthesised Planar Array of

Phantom Elements.

Fig. 1. Round Trip Delays are the same for the Tx/Rx Pair and
the Phantom Element.

B. PHANTOM ANTENNA ARRAYS

Here we will see how M transmitters and N
receivers can be used to synthesise apertures of various Fig. 4. Three Dimensional Synthesised Cylindrical Array of
Phantom Elements.
geometries comprising MXN phantom elements,
allowing the formation of MXN directional beams. The
physical elements may be configured to synthesise a C. WAVEFORMS AND SIGNAL PROCESSING
one-dimensional (Fig. 2), two-dimensional (Fig. 3), or FOR MIMO RADAR
even three-dimensional (Fig. 4) phantom element array, In order to synthesise a MIMO aperture, the
with the phantom elements being synthesised midway transmitted signals from each of the M transmitter
between each transmitter-receiver pair. With there being antenna elements must be separable in some way on
MXN pairwise combinations, we show MXN phantom reception. This separation may be achieved through
elements can be synthesised. temporal, frequency, or code multiplexing, ideally in an
orthogonal form to minimize leakage of transmitter to protect the receiver front ends. The physical
channels into each other on reception. Then, each of the separations of the transmitter and receiver pairs (as
N receiver elements must separately capture the radar shown in Fig. 2, 3, and 4) make this quite practical.
returns and decode or otherwise separate out the returns
into the M components from each transmitter. In this
way MXN data channels are generated for processing.
Conventional beam-forming techniques can then be
applied, such as by delaying the data to give time co-
incidence in a given direction and then summing to form
MXN beams. Alternatively spectral estimation
techniques can be used to estimate angles of arrival or
null-steering techniques to cancel out jamming signals
CW Waveforms Orthogonal frequency division
multiplexing (OFDM) offers a waveform set suitable for
MIMO radar and is a special case of the conventional
CW step-frequency radar technique [6]. For this, each
antenna must sequentially transmit each frequency in
turn, but each element must transmit a different
frequency by staggering the sequences. For example,
consider the transmitter configuration shown in Fig. 5. Fig. 6. An Alternative Step Frequency Sequence
Here a step frequency sequence of 256 steps is used with for MIMO Radar.
4 antennas. A suitable sequence might be for element 1
to transmit f1, f2, f3…f256; element 2 to transmit f2, f3,
f4,….f1; etc. in parallel. Each receiver must be able to separate the signals
from all of the transmitters in order to permit beam
forming. The instantaneous receiver bandwidth must be
sufficiently wide to capture the reflections from all of the
pulses still within the desired range extent. For example,
with the waveform shown in Fig. 5, this could be
achieved with a single step-frequency receiver of
instantaneous bandwidth equal to four times the
frequency step size, and with signal processing used to
separate out the four channels post reception. In general
however the approach shown in Figure 6 would be used,
and with more elements a suitable receiver would cover
the entire RF band and rely on signal processing to
separate out all the individual 256 channels.
Fig. 5. A Frequency-Step Sequence for MIMO Radar.
Signal processing first requires the entire sequence
of transmitter waveforms to be received and separated
This particular sequence causes the range ambiguity to into M channels by de-multiplexing. Each channel is
be determined by duration between each frequency step, then range compressed into L range samples, where L is
this is because the returns from one element must clear the number of frequency steps. This process typically
the two-way return path before the same frequency can requires an L to L point FFT. The above process delivers
be used by the next element. This can lead to the a total of MXN complex data points in each of the L
complete burst of 256 steps being unacceptably long for range cells, M from the de-multiplexed transmitters and
some applications. N from the number of receiver elements. Beam forming
An alternative sequence is shown in Fig 6. Here each may then commence. For the case of the one
antenna transmits all 256 frequency steps, but offset dimensional aperture of Fig. 2, a target’s angle relative
from the neighbouring element by 64 steps. This results to the aperture presents as a spatial frequency in the
in a range ambiguity every 64 steps, rather than every corresponding range bin. Taking the FFT of a given
single step. We note the step frequency sequence can be range sample, using all of the data channels, results in all
implemented as a CW waveform if sufficient isolation MXN beams being calculated simultaneously as shown
can be achieved between transmit and receive elements in Fig. 7. The angular resolution can be defined as the
angle observed by the aperture over the number of FFT Signal Processing Implications Each step of
bins (total number of receiver channels). In the case of a the signal processing required to form the output beams
two dimensional array as shown in Fig. 3 an M point can be separated into separate processes. This allows the
FFT will give the elevation angle of a return and an N signal processing to be organised as a pipeline,
point FFT the azimuth angle. permitting continuous operation. Further signal
processing gains are possible by integrating the output
from multiple step frequency sequences.
Additional signal processing resources can allow a
new output to be generated every frequency step interval
(or every n step intervals) once the sequence has been
completed once. This would not be necessary for most
applications as there is only a small amount of new
information each time; however it may be useful where
high update rates are required.

III. MIMO RADAR CONFIGURATIONS

Known MIMO radars may be divided into two classes:

1) MIMO radars with colocated antennas and coded
signals and 2) radars with widely separated antennas, the
so-called “Statistical MIMO radars”. Let us consider
briefly both classes of MIMO radars.

A. MIMO RADARS WITH COLOCATED

ANTENNAS AND CODED SIGNALS
MIMO radar with colocated antennas, employs
transmit and receive antenna arrays containing elements
which are closely spaced relatively to the working
Fig. 7. Signal Processing Flow for a Linear MIMO Array. wavelength (e.g. spaced by half the wavelength). In a
receiver array of colocated antennas, the signals
The system described above is based on the notion of reflected by the targets have similar amplitude at each
each transmitter element providing a CW step frequency, this receive antenna element and the targets are usually
being efficient from a power generation perspective. Clearly, modeled as punctual. While this configuration does not
other interrupted CW waveforms are possible where the provide spatial diversity, spatial resolution can be
overall sum presents a CW waveform in the far field for good increased by combining the information from all of the
ECM/ECCM purposes. transmitting and receiving paths. This is done by
Pulsed Waveforms If CW operation is not a coherent processing: By exploiting the different time
requirement then more options become available for delays and/or phase shifts, the received signals are
waveform design. A simple example of pulsed MIMO coherently combined to form multiple beams. This
operation is where M transmitters simultaneously emit a configuration also has other benefits such as a good
short pulse, each at one of M different frequencies. The interference rejection and a good flexibility for
frequencies are cycled in sequence so that each antenna transmitting a desired beampattern. MIMO radars with
element transmits all frequencies after M pulses. This colocated antennas can be further classified into bistatic
allows for pulse compression using the MIMO MIMO radars if the transmitter array is widely separated
waveforms and the time between each burst can be long from the receiver array (Fig. 8).
enough to suit the maximum range requirement. This
gives the opportunity to sample the far-field at a rate B. MIMO RADARS WITH WIDELY
equal to M times the round-trip delay at maximum SEPARATED ANTENNAS: “STATISTICAL
range. This is much faster than the CW case, allowing MIMO RADARS”
Doppler processing at higher closing velocities. Clearly
It is well known that if a target is much greater than
many other waveform combinations are to be found
the wavelength of an illuminated signal, the received
which might suit particular applications. signal is random at the input of a radar receiver and
fluctuates in time. Signal fluctuations worsen detection
performance. It is proposed to use multiple widely
separated transmitting and/or receiving antennas. If
distances between antennas are large enough, signal
fluctuations at the receiver inputs become mutually
uncorrelated. Joint incoherent processing of such signals
can smooth signal fluctuations and reduce energy loss.
This configuration allows exploiting the spatial diversity
of the targets' Radar Cross Section (RCS) to improve the
radar performance by addressing the problem similarly
to a MIMO communications problem. Actually, by
combining the different target echoes coming from
different directions (Fig. 9) by non-coherent (or
statistical) processing, a diversity gain is achieved,
similarly to the diversity gain obtained in MIMO Fig. 9. MIMO Radar with Widely Separated Antennas.
communications when data is transmitted over
independent channels.

IV. MIMO RADAR : ADVANTAGES AND

DISADVANTAGES OVER PHASED ARRAY

A. THE NON-IDEAL FACTOR IN RADAR

ANTENNA
[9] Because of the instability of the radar transmitter,
the radar signal is not a pure ideal signal. It has many
components which will affect the detector. Main source
of these non-ideal factors includes signal source, the
amplifier, the timer, etc. The important non-ideal factor
which greatly impair the detector are the signal
amplitude instability, the phase instability, the impulse
duration instability and signal frequency instability and
signal spur etc. Especially, the signal spur will
significantly damage the detector performance because
(a) Bistatic MIMO Radar with Colocated Antennas.
the signal spur will produce false target which will be
detected by detector. Radar designers must also be
concerned with spurious (i.e., undesirable) signals that
can arise during the creation and transmission of a
waveform. Amplifiers, for example, produce harmonics
as well as in-band and adjacent-band spurs. Waveform
generators, such as Direct Digital Synthesizers, generate
spurs due to varying mechanisms including phase
truncation and quantization. When transmitted, such
spurs can interfere with radar detection, and can also
cause EMI that is harmful to other RF systems. To make
matters worse, such spurs can be correlated across the
array. Consequently, these spurs may experience array
gain during transmit beamfoming. Reducing beamformer
gain (eg, through beamspoiling) can thus be of some
help. Better still, when the source of the spur is
waveform related, MIMO can offer a great benefit.
Since the MlMO transmit waveforms are orthogonal,
many spurs (e.g, those due to phase truncation and
(b) Monostatic MIMO Radar with Colocated Antennas. quantization in DDSs) will not be coherent across the
Fig. 8.
array. As a result, these spurs will not be subject to array C. LIMITATIONS OF MIMO
gain on transmitting, thereby reducing interference- both The MIMO architecture has several important
to the radar itself and to other RF systems. Supposed the advantages over traditional phased arrays, but
ideal transmitting signal is constraints imposed by the waveform requirement may
limit the suitability of MIMO for some applications.
The increased element spacing in a MIMO antenna
array reduces mutual coupling between the elements,
When the signal include spurs, the signal becomes minimizing errors in the array factor and in some
configurations permitting CW operation. There are also
significantly fewer physical elements than the equivalent
classical phased array; this can lead to substantial cost
At receiver, the echo is processed by matched filter savings. Also the weight and wind-loading of the
(MF), and judged by the detector. The output of the MF antenna structure is significantly reduced and this can be
is a critical issue for ship-borne radars. When using step
frequency waveforms, the step order can be randomised
to provide an ECCM capability. In the far field the
transmit aperture appears as spatially varying band
limited noise. The spatial variance of the instantaneous
field pattern defeats side-lobe target injection ECM
techniques bymaking it impossible to calculate the ghost
target signal without prior knowledge of the transmitter
In this formulation, the first item is the real target echo. step-frequency sequences and physical geometry.
The second item is false target produced by the signal Thus the radar aperture can easily be disguised, both
spur. It can be seen that the position of the false target is visibly and electronically, for covert operations.
∆t=fm/µ, the amplitude of the false target is 0.5bm/E0. Constraints arise with the use of CW MIMO radar for
Where, the fm and bm is the frequency and the amplitude long range and high resolution applications. These
of the signal spur, respectively. In the conventional require a waveform with a large number of frequency
phased array radar, the false is added up correaltely. steps and a suitable dwell time at each frequency,
Then the false target amplitude is so strong that it can be resulting in long intervals between updates and
judge by the detector. But in the MIMO radar, because consequential difficulties in Doppler processing for fast
all the spur of different signal is un-correlate, they moving targets.
cannot be added up. It will impair the detector
performance.

B. BEAMFORMING V. CONCLUSIONS
MIMO radar can steer a narrow beam employing
multiple antennas of the transmitter and the receiver. The Multi-antenna based radar systems are widely used
beam pattern of full MIMO is consistent to that of in both military and civilian applications. One of the
phased array radar. However, partial MIMO radar has most implemented radar configurations is the phased-
narrower beamwidth with a gain loss and about 13dB array radar system. Phased-arrays employ multiple
peak sidelobe level. For the sparse transmitting antennas, transmitter and multiple receiver antenna elements
partial MIMO can avoid aliasing in angle (Fig. 10). which are usually colocated. The multiple transmitter
elements are capable of cohering and steering the
transmitted energy toward a desired direction by
transmitting scaled and delayed versions of a single
waveform. At the receiver array, the received signals can
be steered in a given direction in order to maximize the
probability of detection or the Signal-to-Noise Ratio
(SNR). This can be done in two different ways: By
performing analog beamforming via the use of phase
shifters in the different receiver architectures, or by
performing digital beamforming via adaptive processing.
Digital beamforming offers several advantages over its
Fig. 10. Beam Patterns analog counterpart, including the capability to steer
multiple simultaneous beams and the possibility to REFERENCES
implement single and multiple sidelobe cancelers.
Another type of multi-antenna radar system is the [1] M. I. Skolnik, Introduction to Radar Systems, 3rd. New York: Mc-
Multiple-Input Multiple-Output (MIMO) radar. A Graw-Hill, 2001.

MIMO radar also employs multiple transmitter and [2] S. Haykin, J. Litva, and T. J. Shepherd, Radar Array Processing.
multiple receiver elements, but unlike the phased-array New York: Springer-Verlag, 1993.
systems, the different waveforms transmitted by a [3] E. Fishler, A. Haimovich, R. Blum, D. Chizhik, L. Cimini, and R.
MIMO radar can be correlated or uncorrelated with each Valenzuela, “MIMO radar: An idea whose time has come,” in Proc. IEEE
other. Compared to phased-array radars, MIMO radars Radar Conf., Honolulu, HI, Apr. 2004, vol. 2, pp. 71–78.
offer more degrees of freedom which lead to improved [4] A. Haimovich, R. Blum, and L. Cimini, “MIMO radar with widely
angular resolution, improved parameter identiability, and separated antennas,” IEEE Signal Process. Mag., vol. 25, pp. 116–
more flexibility for transmit beampattern design. 129,Jan. 2008

Additionally, MIMO radars can synthesize larger virtual [5] J. Li and P. Stoica, “MIMO radar with colocated antennas,” IEEE
arrays which increases resolution and the number of Signal Process. Mag., vol. 24, pp. 106–114, Sep. 2007.
targets that can be detected. [6] D. R. Wehner, High Resolution Radar, 2nd edition, pp. 200-209,
Artech House, Inc., 1995.

[7] B J Donnet and I D Longstaff, “MIMO Radar, Techniques and

Opportunities” September 2006, Manchester UK, Proceedings of the 3rd
European Radar Conference.

[8] V.S. Chernyak Moscow Aviation Institute (State Technical

University) Moscow, Russia “On the Concept of MIMO Radar” IEEE
National Radar Conference - Proceedings · June 2010.

[9] Zeng j iankui, Dong ziming, Electronic information school ,

Chongqing University of science and Technology, “Some MIMO radar
advantages over phased array radar” 2nd International Conference on
Industrial Mechatronics and Automation, 2010.

[10] Y. Qu, G. s. Liao, Member, IEEE, S. Q. Zhu, X.Y.Liu ,H.liao,

“Performance Comparisons ofMIMO and Phased-Array radar”,
MIKON 2008 - 17th International Conference on Microwaves, Radar and
Wireless Communications. IEEE.