PHYSICAL REVIEW LETTERS 131, 053803 (2023)

Editors' Suggestion Featured in Physics

Super Interferometric Range Resolution

John C. Howell ,1,2,* Andrew N. Jordan ,1,2,4 Barbara Šoda ,5 and Achim Kempf 5
1
Institute for Quantum Studies, Chapman University, Orange, California 92866, USA
2
Racah Institute of Physics, The Hebrew University of Jerusalem, Jerusalem, Israel, 91904
3
Center for Coherence and Quantum Optics, University of Rochester, Rochester, New York 14627, USA
4
Department of Physics and Astronomy, University of Rochester, Rochester, New York 14627, USA
5
Department of Applied Mathematics and Department of Physics, University of Waterloo, and
Perimeter Institute for Theoretical Physics, Waterloo, Ontario, Canada
(Received 7 March 2023; accepted 13 June 2023; published 2 August 2023)

We probe the fundamental underpinnings of range resolution in coherent remote sensing. We use a novel
class of self-referential interference functions to show that we can greatly improve upon currently accepted
bounds for range resolution. We consider the range resolution problem from the perspective of single-
parameter estimation of amplitude versus the traditional temporally resolved paradigm. We define two
figures of merit: (i) the minimum resolvable distance between two depths and (ii) for temporally
subresolved peaks, the depth resolution between the objects. We experimentally demonstrate that our
system can resolve two depths greater than 100× the inverse bandwidth and measure the distance between
two objects to approximately 20 μm (35 000 times smaller than the Rayleigh-resolved limit) for temporally
subresolved objects using frequencies less than 120 MHz radio waves.

DOI: 10.1103/PhysRevLett.131.053803

Range resolution [1–6] is the ability to determine the cτ

dr ≥ ; ð1Þ
distance between two objects along the same line-of-sight 2
when performing remote sensing. The prevailing thought is
that radar range resolution is inextricably linked to the where c is the speed of light, τ is the pulse width, and 2
inverse bandwidth of a pulse [5,6] or to the wavelength of comes from the round-trip of the pulse [5,6]. Phase or
the electromagnetic wave owing to the coherent nature of frequency encoding are commonly employed to realize
the interfering wavefronts. We quote, “Wave theory indi- high time-bandwidth product pulses, which when
cates that the best vertical resolution that can be achieved is
one-quarter of the dominant wavelength [7]. Within that (a)
vertical distance any reflections will interfere in a con-
structive manner and result in a single, observed reflection”
[8]. The desire for better range resolution has driven
scientists and engineers to ever-higher frequencies for radar
and lidar [9–12]. However, this comes at a severe cost
because transmission through and reflection from various (b)
material media is critically tied to frequency [8,13–16]. We
dramatically improve upon these widely accepted limits of
range resolution using a novel class of self-referenced
functions to demonstrate several orders of magnitude
improvement in range resolution beyond known limits.
For transform-limited pulses, two radar targets are (c)
considered range resolved when the range resolution
distance dr obeys the inequality

Published by the American Physical Society under the terms of FIG. 1. Free-space (a) and guided-wave (b) and (c) ranging
the Creative Commons Attribution 4.0 International license. experimental schematics. Using parameter estimation techniques
Further distribution of this work must maintain attribution to on interference-class waveforms, we estimate the distance be-
the author(s) and the published article’s title, journal citation, tween two scattering depths from a target which would otherwise
and DOI. be temporally subresolved.

0031-9007=23=131(5)=053803(6) 053803-1 Published by the American Physical Society

 PHYSICAL REVIEW LETTERS 131, 053803 (2023)

combined with match-filtered pulse compression, lead to

high temporal resolution. Therefore, a more general range
resolution for a pulse is set by the inverse bandwidth [5,6].
Going beyond these limits has been historically difficult.
Wong and Elefthiardes used superoscillations to reduce
ranging uncertainty by 36% [17]. Recently, Komissarov
et al. used partially coherent radar in an attempt to decouple
range resolution from the signal bandwidth [18] achieving a
factor of 10 improvement. We classify the current state of
the art as a “temporally resolved” paradigm.
Here, we introduce a new “amplitude resolved” para-
digm using self-referenced interference-class functions as
shown in Fig. 1. An interference-class pulse is sent to a
remote object. Multiple scattering depths along the same
line of sight result in the interference of temporally shifted
versions of the waveform. The resultant waveform is
measured by the receiver. In theory, by using parameter
estimation, we can determine range, distance between
objects, relative scattering amplitudes, etc. We expect that
amplitude-resolution ranging methods have been unex-
plored because of the inability to distinguish between loss
and subresolved interference peaks.
Our system relies on parameter estimation from the
interference between coherent pulses. A fundamentally
different parameter estimation has also been used to over- FIG. 2. Pulses and fast Fourier transforms. Panel (a) shows the
come the spatial Rayleigh resolution limit of incoherent functions used in the Letter. Panel (b) shows their Fourier
transforms. Panel (c) shows the interfered waveforms from the
sources using mode sorting [19–23], allowing for funda-
coherent interference of two equal-amplitude pulses.
mental definitions of spatial resolution [24–26]. Of note,
Ansari et al used two incoherent optical pulses and mode
decomposition to achieve supertemporal resolution [27]. with many different types bandlimited functions, we use
In designing our functions, we employ two behaviors. the formalism of Šoda and Kempf [32]. We generate
First, a region of the function must be very sensitive to our tailored and bandlimited functions by exploiting the
interference requiring extended and steep temporal gra- product of a bandlimited “canvas” cðtÞ function and arbi-
dients. Second, a zero-gradient region within the function, trary (e.g., Taylor) polynomials f n ðtÞ:
which is insensitive to the interference, is used as an
amplitude reference. In this manner, as long as all portions gðtÞ ≔ f n ðtÞcm ðtÞ; ð2Þ
of the pulse experience the same attenuation (or amplifi-
cation) and there is a flat spectral response in a medium or where we use
upon reflection, the range resolution properties of the pulse
m
are preserved. Ωt
cm ðtÞ ¼ sinc ð3Þ
We consider three types of pulse functions as shown in m
Fig. 2(a). The dashed orange line shows a specially
designed interference-class, bandlimited pulse. The second and
interference class function is shown in solid green line,
which we denoted as a “triangle” function. The dotted blue X
n−1
f n ðtÞ ¼ ak tk : ð4Þ
line shows a standard sinc2 pulse, which is not an
k¼0
interference-class function, but is only used as a means
of comparison of a Rayleigh criterion for temporal pulse For m > n this is a square-integrable function with band-
resolution. The bandlimit for the sinc2 and the custom limit set by Ω. See Supplemental Material for more
function, in the image is the same. The corresponding fast information [33]. We note that the specific polynomial
Fourier transforms are shown in Fig. 2(b). function used herein is not unique—more generalized,
For the first type of self-referenced pulses, we use optimized functions that have the features we describe
bandlimited function theory (e.g., used in superoscillations here will be pursued in future work.
[28–32]) to generate our specially designed functions. For the second type of self-referenced functions, we
While it is likely that similar behavior could be achieved introduce an idealized line-segment function we call a

053803-2
 PHYSICAL REVIEW LETTERS 131, 053803 (2023)

“triangle pulse” TrðtÞ, shown in solid green in Fig. 2(a). of the Michelson interferometer. In Fig. 1(c), the delay
The triangle pulse is not bandlimited, so we consider cable is connected to both channels 1 and 2. Channel 1
approximations to the bandwidth based on the Fourier measures the interfered waveform and channel 2 measures
transform. Nevertheless, this function is better for meas- the input (non-interfered) waveform. Both input channels
uring the minimum distance between two objects (range on the scope are set at 1 MΩ to achieve the desired re-
resolved objects) owing to the linearity of S over the full flections and measurement. The experiments in Figs. 2(b)
interference range. and 2(c) utilize high precision frequency tunability to
Our third type of function is not an interference-class achieve ultrasmall relative pulse shifts rs .
function, but is used simply to define a temporally resolved The functions are generated numerically, consisting of
function. Similar to the Rayleigh criterion [34], when the 4000 points and a duration of 40 tR units. These signals
peak of one pulse is separated by a distance greater than were uploaded to an arbitrary waveform generator (AWG).
first minimum of a second pulse, the pulses are considered The bandlimit of the system is set by the repetition rate of
to be resolved. A classic example is the sinc2 , the arbitrary waveform generator. For example, a repetition

2 rate of 1 MHz results in a 40 MHz bandlimit for the
Ωt bandlimited pulses.
c2 ðtÞ ¼ sinc : ð5Þ
2 To perform the ranging experiments as shown in
Fig. 1(c), the arbitrary waveform generator sends the signal
For a bandlimited pulse, the minimally Rayleigh-resolved down the cable. The gðtÞ pulse is sent to channel 1 of the
temporal shift tR (the time analog of resolvability in space) oscilloscope and is teed to another cable which then adds a
is given by tR ¼ 2π=Ω. How deeply we can superresolve one-way temporal measured delay of tc ¼ 4.4 ns (the
the targets is quantified by the ratio measured delay is 3.8 ns when not connected to channel
2 and the measured cable length is lc ¼ 72 cm) implying
rs ¼ td =tR ; ð6Þ that the pulse delay is td ¼ 2tc ¼ 8.8 ns. The reflected
signal from channel 2 interferes in channel 1 with the
where td is the temporal delay between the two returning original displaced signal the gðtÞ function. When using the
pulses. Since rs is both a function of the delay td and the bandlimited function gðtÞ, for example, channel 1 then
bandwidth tR ¼ 2π=Ω, for the work herein it is more measures gðtÞ þ gðt þ 2tc Þ and similarly for the other
precise to change the bandlimit to test the fundamental functions.
properties of the relative shift in the system rather than Figure 2(c) shows the resultant interference waveform in
changing the relative pulse delay. channel 1 for the three different types of pulses used herein
We define a signal S akin to balanced interferometric for ts ¼ 0.5tR. There are several important features of this
detection used to measure transverse deflections [35], graph that should be noted. First, the c2 ðtÞ interference
namely, pattern is not resolved, as expected, since ts ¼ 0.5tR .
Second, the interference regions in the center of the gðtÞ
Acmax − Acmin (peaks have changed value) and triangle functions TrðtÞ
S¼ ; ð7Þ
Almax − Almin (interference plateaus) have changed dramatically. Third,
the heights of the side lobes are still roughly constant for
where Acmax , Acmin are the maximum and minimum both interference-class functions.
amplitude of the function in the steep center region, We consider two important figures of merit for this range
respectively and Almax , Almin are the maximum and mini- resolution system: (i) the minimum distance to amplitude-
mum of the flat temporal lobes. resolve two objects and (ii) the distance resolution between
Consider the ranging experiments shown in Fig. 1. In objects when the objects are amplitude resolved but still
Fig. 1(a), waveforms are sent to a two-depth target. Upon temporally subresolved (i.e., rs < 1).
reflection, the two reflected waveforms interfere and are The triangle function is designed to amplitude resolve
sent to the receiver. The resultant waveform is measured two objects with depths that are closely spaced along the
and processed to estimate the distance between the two same line of sight. Typically, in radar, the spectral band-
depths of the target. width is given by the width of the spectrum at the 3 dB
To better test the limits of this technique, we use low- down point. However, owing to the irregular spectrum of
noise guided-wave experiments as shown in panels 1(b) the triangle pulse we use a conservative estimate based on
and 1(c). Figure 1(b) shows a guided-wave equivalent to the where most of the power is found. Using the bandwidth
free-space radar. However, there are unequal amplitudes from the Fourier transform as shown in Fig. 2(b), we can
from the scattered waves measured at the receiver based see that most of the power lies below 20% of the bandlimit
on the splitting ratios of the tee junction. To create for the bandlimited functions using the same pulse repeti-
equal-amplitude interference, we used the experiment tion rate. As noted above, the bandlimited frequency was
shown in Fig. 2(c), which is the radio wave equivalent 40× the repetition rate. From this observation, we define a

053803-3
 PHYSICAL REVIEW LETTERS 131, 053803 (2023)

FIG. 3. The signal S is plotted against bandwidth of the tri-

angle function with (interference) and without (no interference)
a delay cable.
FIG. 4. Signal S vs time shift. The signal S from Eq. (7) is
conservative bandwidth of the triangle function to be 8× computed and measured as a function of the relative shift of
the pulse repetition rate. two gðtÞ functions.
The resolving power of the triangle function can be
determined from Fig. 3. Here, we used the 72 cm delay relative time between two shifts rs1 and rs2 . For a fixed
cable disconnected from channel 2 and only connected to cable length,
channel 1 yielding a 7.6 ns round-trip delay time. We
change the bandwidth of the triangle pulse until we can no ΔΩ
Δrs ¼ 2tc ; ð8Þ
longer distinguish between the S curve with and without a 2π
cable. Here, we see that at approximately 1 MHz band-
width, it is possible to distinguish the signal with and where Δrs ¼ rs2 − rs1 and ΔΩ=2π is the smallest resolv-
without the delay cable. This corresponds to rs ≈ 0.008 or able change in the bandwidth of the pulse.
better than 100 times the inverse bandwidth. In the inset of Fig. 4, it shows the change in the signal of
To further demonstrate the power of this technique, we a small incremental frequency shift. Using the 72 cm cable
removed the 72 cm cable and used only the extra path length with tc ¼ 4.4 ns delay and a 2 MHz repetition rate
of a bnc T junction, which has about 1.3 cm path length. We (Ω=2π ¼ 80 MHz bandlimit) resulted in approximately
used a 2.5 GSa=s triangle function with 10 samples per time ts ¼ 0.69 shift. The bandwidth of the pulses was changed
unit tR (401 total points as opposed to 4000 described above) from 79.960 to 80 MHz in increments of 8.0 kHz. Each
and measured the interference signal on an 800 MHz trace was averaged 512 times and each setting was
oscilloscope. The estimated bandwidth of the signal, based measured 10 times with 15 to 30 sec between each
on the Fourier transform, was approximately 100 MHz, but measurement. Assuming a signal to noise ratio of 1, we
with some small amount of frequency content up to can resolve 80 MHz bandwidth changes down to about
500 MHz. With and without the T junction, the value of 3.2 kHz. These data imply Δrs ¼ 2.8 × 10−5. The inferred
S was 0.5573  0.0002 and 0.5511  0.0004, respectively, depth resolution of our system is then Δx ¼ vc tR Δrs =2≈
or a separation of 15 standard deviations. This implies sub- 20 μm. Thus, for a target with two equal-amplitude,
mm resolution. We then added a male-to-male bnc connector temporally subresolved reflected pulses, we can measure
that added another couple centimeters and the value dropped the relative distance between them down to 35 000 times
to 0.5285  0.0001. below the Rayleigh limit and several orders of magnitude
The second important figure of merit for range resolution below the timing resolution of the oscilloscope.
is the ability to determine the distance between two objects While the primary emphasis of this paper concerns the
for temporally subresolved pulses. Figure 4 shows the simplest case of equal-amplitude reflections like those
signal S vs rs (the ratio of the round-trip cable delay to the obtained using the setup in Fig. 1(c), in realistic applications,
Rayleigh criterion) using the bandlimited function gðtÞ. target reflection amplitude will typically be unequal, like
It can be seen in Fig. 4 that the theory and experiment those generated in Fig. 1(b). In the Supplemental Material
agree well. [33], we show a two-parameter signal that involves both a
The slope of S tells us the sensitivity of the system to temporal shift and disparate amplitudes and disparate
changes in the relative temporal shift of the two functions. amplitudes based.
The S function for the bandlimited pulse is roughly We note several important comments. (i) These experi-
quadratic in the region of zero shift and becomes linear ments were done using low frequency radio wavelengths.
after about rs ¼ 0.5 or half the Rayleigh resolved time. However, they are equally valid in all parts of the electro-
We want to know how well this system can determine the magnetic spectrum. (ii) As demonstrated preliminarily, the

053803-4
 PHYSICAL REVIEW LETTERS 131, 053803 (2023)

system can be generalized to account for disparate reflec- [11] Xiaoye Liu, Airborne lidar for dem generation: Some
tion amplitudes and multiple layers by creating more exotic critical issues, Prog. Phys. Geogr. 32, 31 (2008).
functions and signal analysis. (iii) It is straightforward to [12] Behnam Behroozpour, Phillip A. M. Sandborn, Ming C.
convert time resolution to space resolution by transversely Wu, and Bernhard E. Boser, Lidar system architectures and
circuits, IEEE Commun. Mag. 55, 135 (2017).
scanning the receiver in Fig. 1(a) and solving the inverse
[13] Tomohiro Oguchi, Electromagnetic wave propagation and
problem opening up high resolution imaging. scattering in rain and other hydrometeors, Proc. IEEE 71,
In summary, we have demonstrated both theoretically 1029 (1983).
and experimentally that it is possible to obtain range [14] Ahmed I. Al-Shamma’a, Andrew Shaw, and Saher Saman,
resolution far better than the Rayleigh criterion or the Propagation of electromagnetic waves at MHz frequencies
inverse bandwidth. We employed the coherent aspects of through seawater, IEEE Trans. Antennas Propag. 52, 2843
radio wave transmission and detection to measure sensitive (2004).
interference patterns. In the future work, we will explore the [15] Andrew Shaw, A. I. Al-Shamma’a, S. R. Wylie, and D. Toal,
fundamental limits of this technique, as well as apply this Experimental investigations of electromagnetic wave
method to more realistic ranging tasks in the field. propagation in seawater, in 2006 European Microwave
Conference (IEEE, 2006), pp. 572–575, 10.1109/EUMC
J. C. H. and A. N. J. acknowledge support from .2006.281456.
Chapman University and the Bill Hannon Foundation. [16] Shan Jiang and Stavros Georgakopoulos, Electromagnetic
A. K. acknowledges support through a Discovery Grant wave propagation into fresh water, J. Electromagn. Anal.
of the National Science and Engineering Council of Canada Appl. 2011, 261 (2011).
(NSERC), a Discovery Project grant of the Australian [17] Alex M. H. Wong and George V. Eleftheriades, Super-
Research Council (ARC) and a Google Faculty Research oscillatory radar imaging: Improving radar range resolution
beyond fundamental bandwidth limitations, IEEE Micro-
Award. B. S. is supported by the Perimeter Institute, which
wave Wireless Compon. Lett. 22, 147 (2012).
is supported in part by the Government of Canada through [18] Rony Komissarov, Vitali Kozlov, Dmitry Filonov, and Pavel
the Department of Innovation, Science and Economic Ginzburg, Partially coherent radar unties range resolution
Development Canada and by the Province of Ontario from bandwidth limitations, Nat. Commun. 10, 1423
through the Ministry of Economic Development, Job (2019).
Creation and Trade. [19] Mankei Tsang, Quantum Imaging beyond the Diffraction
Limit by Optical Centroid Measurements, Phys. Rev. Lett.
102, 253601 (2009).
[20] H. Shin, Kam Wai Clifford Chan, H. J. Chang, and R. W.
*
[email protected] Boyd, Quantum Spatial Superresolution by Optical Cent-
[1] Donald R. Wehner, High Resolution Radar (Artech House, roid Measurements, Phys. Rev. Lett. 107, 083603 (2011).
Inc., Norwood, MA, 1987), p. 484. [21] Mankei Tsang, Subdiffraction incoherent optical imaging
[2] Jian Li, Guoqing Liu, Nanzhi Jiang, and Petre Stoica, via spatial-mode demultiplexing, New J. Phys. 19, 023054
Moving target feature extraction for airborne high-range (2017).
resolution phased-array radar, IEEE Trans. Signal Process. [22] Weng-Kian Tham, Hugo Ferretti, and Aephraim M.
49, 277 (2001). Steinberg, Beating Rayleigh’s Curse by Imaging Using
[3] K. B. Cooper, R. J. Dengler, G. Chattopadhyay, E. Schlecht, Phase Information, Phys. Rev. Lett. 118, 070801 (2017).
J. Gill, A. Skalare, I. Mehdi, and P. H. Siegel, A high- [23] Yiyu Zhou, Jing Yang, Jeremy D. Hassett, Seyed Mohammad
resolution imaging radar at 580 GHz, IEEE Microwave Hashemi Rafsanjani, Mohammad Mirhosseini, A. Nick
Wireless Compon. Lett. 18, 64 (2008). Vamivakas, Andrew N. Jordan, Zhimin Shi, and Robert
[4] Fangzheng Zhang, Qingshui Guo, and Shilong Pan, W. Boyd, Quantum-limited estimation of the axial separation
Photonics-based real-time ultra-high-range-resolution radar of two incoherent point sources, Optica 6, 534 (2019).
with broadband signal generation and processing, Sci. Rep. [24] Mikael P. Backlund, Yoav Shechtman, and Ronald L.
7, 13848 (2017). Walsworth, Fundamental Precision Bounds for Three-
[5] Merrill Ivan Skolnik, Introduction to Radar Systems Dimensional Optical Localization Microscopy with Poisson
(McGraw Hill Book Co., New York, 1980), p. 590. Statistics, Phys. Rev. Lett. 121, 023904 (2018).
[6] Nadav Levanon and Eli Mozeson, Radar Signals (John [25] Carmine Napoli, Samanta Piano, Richard Leach, Gerardo
Wiley & Sons, New York, 2004). Adesso, and Tommaso Tufarelli, Towards Superresolution
[7] Robert E. Sheriff, Seismic Stratigraphy—Applications to Surface Metrology: Quantum Estimation of Angular
Hydrocarbon Exploration (American Association of and Axial Separations, Phys. Rev. Lett. 122, 140505
Petroleum Geologists, 1977), 10.1306/M26490C1. (2019).
[8] Adrian Neal, Ground-penetrating radar and its use in [26] J. Řeháček, M. Paúr, B. Stoklasa, D. Koutný, Z. Hradil,
sedimentology: Principles, problems and progress, and L. L. Sánchez-Soto, Intensity-Based Axial Localization
Earth-Sci. Rev. 66, 261 (2004). at the Quantum Limit, Phys. Rev. Lett. 123, 193601
[9] R. T. H. Collis, Lidar, Appl. Opt. 9, 1782 (1970). (2019).
[10] Ulla Wandinger, Introduction to Lidar (Springer, New York, [27] V. Ansari, B. Brecht, J. Gil-Lopez, J. M. Donohue, J.
2005). Řeháček, Z. Hradil, L. L. Sánchez-Soto, and C. Silberhorn,

053803-5
 PHYSICAL REVIEW LETTERS 131, 053803 (2023)

Achieving the ultimate quantum timing resolution, PRX [32] Barbara Šoda and Achim Kempf, Efficient method to create
Quantum 2, 010301 (2021). superoscillations with generic target behavior, Quantum
[28] Y. Aharonov, S. Popescu, and D. Rohrlich, How can an Stud. Math. Found. 7, 347 (2020).
infra-red photon behave as a gamma ray, Tel-Aviv [33] See Supplemental Material at http://link.aps.org/
University Preprint TAUP, 1847–1890 (1990). supplemental/10.1103/PhysRevLett.131.053803 for addi-
[29] Michael V. Berry, Optical currents, J. Opt. A 11, 094001 tional information and calculations about bandlimited
(2009). functions, the Rayleigh criterion, signal analysis and
[30] Sacha Kocsis, Boris Braverman, Sylvain Ravets, Martin J. multi-parameter estimation.
Stevens, Richard P. Mirin, L. Krister Shalm, and Aephraim M. [34] Max Born and Emil Wolf, Principles of Optics: Electro-
Steinberg, Observing the average trajectories of single pho- magnetic Theory of Propagation, Interference and
tons in a two-slit interferometer, Science 332, 1170 (2011). Diffraction of Light (Elsevier, New York, 2013).
[31] Konstantin Y. Bliokh, Aleksandr Y. Bekshaev, Abraham G. [35] P. B. Dixon, D. J. Starling, A. N. Jordan, and J. C. Howell,
Kofman, and Franco Nori, Photon trajectories, anomalous Ultrasensitive Beam Deflection Measurement via Interfero-
velocities and weak measurements: A classical interpreta- metric Weak Value Amplification, Phys. Rev. Lett. 102,
tion, New J. Phys. 15, 073022 (2013). 173601 (2009).

053803-6