Dynamics of a Spin-Wave Active Ring Resonator Driven by Harmonic-Null Square-Wave and Unipolar 8-bit Walsh Code Modulations

Anil Prabhakar; Anirban Mukhopadhyay; Kaustubh Narayan

arxiv: 2604.04612 · v2 · submitted 2026-04-06 · ⚛️ physics.app-ph

Dynamics of a Spin-Wave Active Ring Resonator Driven by Harmonic-Null Square-Wave and Unipolar 8-bit Walsh Code Modulations

Anirban Mukhopadhyay , Kaustubh Narayan , Anil Prabhakar This is my paper

Pith reviewed 2026-05-10 19:19 UTC · model grok-4.3

classification ⚛️ physics.app-ph

keywords spin-wave active ring resonatorreservoir computingnonlinear dynamicsWalsh codesharmonic eliminationyttrium iron garnetshort-term memory

0 comments

The pith

Harmonic-null square-wave modulation and Walsh decomposition map nonlinearity and memory in spin-wave ring resonators.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces two straightforward experimental techniques to probe the nonlinear dynamics and short-term memory of a yttrium iron garnet based spin-wave active ring resonator. Driving the device with a square wave engineered to lack energy at three times its fundamental period produces output peaks at fd plus or minus 3/T, marking five distinct intervals of strong nonlinearity between 2.15 and 2.2 GHz. An 8-bit unipolar Walsh code sequence then measures retention of past inputs, yielding an estimated short-term memory duration of about 300 nanoseconds, while decomposing the output waveform onto the Walsh basis quantifies the nonlinear transformation performed by the resonator. These characterizations together supply a practical route to select operating points and tune such devices for reservoir computing.

Core claim

The authors demonstrate that a third-harmonic-null square-wave drive applied to the SWARR produces clear spectral peaks at fd ± 3/T in five frequency sub-bands, confirming nonlinear response regions, while sequency-ordered 8-bit unipolar Walsh modulation isolates a short-term memory window of approximately 300 ns and allows the temporal output to be expressed as a linear combination of the input Walsh codewords, thereby quantifying the device's nonlinear mapping.

What carries the argument

Harmonic-null square-wave modulation combined with sequency-ordered 8-bit unipolar Walsh code decomposition, which isolates nonlinear sidebands and projects the resonator output onto an orthogonal basis to extract memory duration and response coefficients.

If this is right

The five identified nonlinear frequency intervals can be selected as operating points for reservoir computing tasks.
The 300 ns memory estimate obtained from Walsh modulation directly informs the choice of input sequence length and reservoir size.
Decomposition of the output into Walsh codewords provides coefficients that quantify the strength of the nonlinear transformation.
Varying the modulation pattern supplies a systematic way to optimize the reservoir's memory and nonlinearity for different tasks.
The same harmonic-elimination and basis-decomposition approach can be applied to other nonlinear delay-line resonators.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If memory duration scales predictably with bias field or film thickness, the Walsh method could be used to forecast performance of scaled devices without new fabrication runs.
The harmonic-null technique may be extended to suppress selected higher-order nonlinear terms in optical or acoustic delay-line reservoirs.
Training a linear readout on the Walsh-decomposed coefficients could produce a compact, interpretable model of the reservoir dynamics.

Load-bearing premise

The peaks at fd ± 3/T arise specifically from the resonator's nonlinearity rather than from linear propagation or measurement artifacts, and the Walsh pattern measures memory duration without significant interference from other device dynamics.

What would settle it

Drive an identical linear delay line (YIG film removed or replaced by a passive cable) with the same harmonic-null square-wave input; absence of the fd ± 3/T peaks would confirm that those features indicate nonlinearity in the actual spin-wave device.

Figures

Figures reproduced from arXiv: 2604.04612 by Anil Prabhakar, Anirban Mukhopadhyay, Kaustubh Narayan.

**Figure 1.** Figure 1: Circuit diagram of a spin-wave active ring resonator (SWARR). [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 3.** Figure 3: Power spectra extracted from Tap-2 of the SWARR driven at (a-c) 2.142 GHz and (d-f) 2.1588 GHz. In (a-c), the spectral components at [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

**Figure 5.** Figure 5: RMSE E between actual demodulated signals and their reconstruction for all Walsh functions across different drive frequencies. with faster temporal variations than W7(t) within the linear approximation [PITH_FULL_IMAGE:figures/full_fig_p003_5.png] view at source ↗

**Figure 4.** Figure 4: Output power Pout at fd ± 3 T as function of fd at G = 39 dB. The dashed and dotted vertical lines indicate fd = 2.142 and 2.1588 GHz. B. Spectral output from Tap-2 of SWARR [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗

**Figure 6.** Figure 6: Responses of SWARR to the drive signal modulated by [PITH_FULL_IMAGE:figures/full_fig_p003_6.png] view at source ↗

**Figure 7.** Figure 7: Responses of SWARR to the drive signal modulated by [PITH_FULL_IMAGE:figures/full_fig_p004_7.png] view at source ↗

**Figure 8.** Figure 8: L2 -norm ratio at fd = 2.1588 GHz and G2 = 33 dB, plotted as a function of τ. The green region indicates the 5% band. The SWARR is probed using GHz carrier signals modulated by W6(t) and W7(t) with code durations τ0 = 1, 1.2 and 1.6 µs. The STM is quantified by examining the L2 -norm ratio η between the SWARR output in response to the 0 th bit and the next bit with a value of 1, e.g., b (6) 5 in the Walsh … view at source ↗

read the original abstract

Spin-wave active ring resonators (SWARRs) based on yttrium iron garnet (YIG) films exhibit rich nonlinear dynamics that make them promising platforms for physical reservoir computing. We present systematic and experimentally simple methods to characterize a SWARR's nonlinear behavior and memory. We first use a third harmonic elimination method to probe the nonlinear response. A drive frequency $f_\mathrm{d}$ is modulated by a square-wave pattern engineered to have a spectral null at $3/T$, which is then applied as input to the SWARR. The power spectra at the output of the YIG delay line allow us to identify five distinct regions within a drive frequency range of $2.15 < f_\text{d} < 2.2\ \text{GHz}$ where nonlinearity was observed as frequency peaks at $f_\mathrm{d} \pm \frac{3}{T}$. The STM duration of the SWARR was estimated to be approximately 300 ns using a modulation pattern derived from the sequency-ordered 8-bit unipolar Walsh family. The nonlinear dynamics of the SWARR were further quantified by decomposing its temporal response to analog Walsh pulses in terms of the input Walsh codewords. The proposed methods of harmonic elimination and Walsh-function decomposition together provide a practical and general framework for the design and optimization of tunable spin-wave reservoir computers.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a practical way to map nonlinearity and short-term memory in a YIG spin-wave resonator using a null-at-third-harmonic square wave plus Walsh-code probing, and the observed peaks and 300 ns estimate line up cleanly with the inputs.

read the letter

The main point to take away is that the authors have developed two related experimental techniques for characterizing the nonlinear dynamics and short-term memory of a spin-wave active ring resonator made from YIG. One uses a square-wave modulation chosen so its spectrum has a null at three times the modulation period, letting them see output peaks at the drive frequency plus or minus that value as a sign of nonlinearity. The other applies sequency-ordered 8-bit Walsh codes to probe the memory duration, coming up with roughly 300 nanoseconds. What stands out is how straightforward these methods are and how directly the data supports the interpretation. They map out five separate intervals in the 2.15 to 2.2 GHz drive range where the nonlinearity shows up clearly in the power spectra. The Walsh decomposition of the output response also lets them break down how the device reacts to different input patterns without needing complex modeling. This kind of practical tool could help people building physical reservoirs from spin waves, since knowing where the nonlinearity lives and how long the memory lasts is key to using them for computation. The soft spots are mostly about scope rather than errors in the logic. The work is based on one particular resonator and a narrow frequency band, so it is not yet clear how well the same nulling and decomposition approach would transfer to other geometries or materials. The memory estimate relies on the Walsh pattern response, but the paper does not appear to include extensive controls for linear filtering effects or ring circulation time that might overlap with the 300 ns figure. Those are the kinds of things a referee would want clarified, but they do not break the central observations. The stress test confirms that the peaks at fd plus or minus 3/T are a solid signature of nonlinearity and the sequency probing gives a falsifiable bound on memory. This is the sort of paper that experimental groups in magnonics or unconventional computing would find useful for their lab work. It is not a theoretical breakthrough, but the methods are described in enough detail that others could try them on their own setups. I think it deserves a serious referee because the experimental evidence is presented cleanly and the claims stay close to what the data shows.

Referee Report

2 major / 2 minor

Summary. The paper claims to provide systematic experimental methods for characterizing the nonlinear behavior and memory of a spin-wave active ring resonator (SWARR) using a harmonic-null square-wave modulation to identify nonlinearity via output peaks at fd ± 3/T in five regions within 2.15-2.2 GHz, and sequency-ordered 8-bit Walsh codes to estimate a short-term memory (STM) duration of approximately 300 ns through response decomposition. These are proposed as a practical framework for tunable spin-wave reservoir computers.

Significance. This work has potential significance for physical reservoir computing by offering simple, experimentally accessible tools to probe and optimize the nonlinear dynamics and memory properties of YIG-based SWARRs. The direct spectral signatures and falsifiable memory estimation are positive aspects that could aid in the design of such systems, provided the methods are shown to be robust against confounding effects.

major comments (2)

The identification of five distinct nonlinearity regions is central to the harmonic elimination method, but the manuscript does not specify the quantitative criteria (e.g., peak amplitude thresholds or signal-to-noise ratios) used to confirm the peaks at fd ± 3/T, nor does it include control measurements in the linear regime to rule out other sources of spectral features.
The estimation of STM duration as 300 ns using the 8-bit Walsh family assumes that the modulation pattern accurately measures short-term memory without significant interference from other effects such as linear dispersion or external noise; however, no explicit tests or modeling to validate this assumption are described, which is load-bearing for the memory characterization claim.

minor comments (2)

The notation for the drive frequency fd and the period T could be clarified with a consistent definition early in the text to avoid ambiguity in the spectral analysis.
The power spectra figures would benefit from annotations indicating the expected peak positions at fd ± 3/T to improve clarity for readers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thoughtful review and constructive feedback on our manuscript. We address each major comment below and will revise the manuscript to incorporate additional quantitative details and validation steps as outlined.

read point-by-point responses

Referee: The identification of five distinct nonlinearity regions is central to the harmonic elimination method, but the manuscript does not specify the quantitative criteria (e.g., peak amplitude thresholds or signal-to-noise ratios) used to confirm the peaks at fd ± 3/T, nor does it include control measurements in the linear regime to rule out other sources of spectral features.

Authors: We agree that explicit quantitative criteria and control measurements would improve the rigor of the nonlinearity identification. In the revised manuscript, we will specify the criteria used to confirm the peaks (e.g., amplitude exceeding 8 dB above the noise floor with SNR greater than 4) and add control data acquired at reduced input powers to demonstrate linear operation with no observable peaks at fd ± 3/T. revision: yes
Referee: The estimation of STM duration as 300 ns using the 8-bit Walsh family assumes that the modulation pattern accurately measures short-term memory without significant interference from other effects such as linear dispersion or external noise; however, no explicit tests or modeling to validate this assumption are described, which is load-bearing for the memory characterization claim.

Authors: We acknowledge the need for explicit validation of the STM estimate. The revised manuscript will include supporting analysis, such as comparisons to linear dispersion models and measurements at different noise levels, to demonstrate that the 300 ns value is not significantly affected by these confounding factors. revision: yes

Circularity Check

0 steps flagged

No significant circularity; experimental characterization is self-contained

full rationale

The paper describes experimental procedures for probing nonlinear response via harmonic-null square-wave modulation and estimating short-term memory via sequency-ordered Walsh codes. These rely on direct spectral observations (peaks at fd ± 3/T) and temporal decompositions of measured responses, without any derivation chain, fitted parameters renamed as predictions, or load-bearing self-citations. The framework claim follows from the reported measurements rather than reducing to inputs by construction, and the work is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based solely on the abstract, no specific free parameters, axioms, or invented entities are detailed.

pith-pipeline@v0.9.0 · 5557 in / 955 out tokens · 41706 ms · 2026-05-10T19:19:17.995943+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/ArithmeticFromLogic.lean LogicNat (8-tick period forced by distinction) echoes

?

echoes
ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

modulation pattern derived from the sequency-ordered 8-bit unipolar Walsh family... STM duration of approximately 300 ns
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel (J(x) = ½(x + x⁻¹) − 1) unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

square-wave pattern engineered to have a spectral null at 3/T... peaks at fd ± 3/T

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages

[1]

Nonlinear spin waves in magnetic film feedback rings,

M. Wu, “Nonlinear spin waves in magnetic film feedback rings,” inSolid State Physics. Academic Press, 2010, vol. 62, pp. 163–224

work page 2010
[2]

Prabhakar and D

A. Prabhakar and D. D. Stancil,Spin waves: Theory and applications. Springer, New York, 2009

work page 2009
[3]

Combinatorial logic devices based on a multi-path active ring circuit,

A. Khitun and M. Balinskiy, “Combinatorial logic devices based on a multi-path active ring circuit,”Scientific Reports, vol. 12, p. 9482, 2022

work page 2022
[4]

Magnonic active ring co-processor,

M. Balynsky, Y . Khivintsev, A. Kozhevnikov, Y . Nikulin, V . Sakharov, Y . Filimonov, and A. Khitun, “Magnonic active ring co-processor,” Journal of Applied Physics, vol. 133, p. 023904, 2023

work page 2023
[5]

Magnonic combinatorial memory,

M. Balinskyy and A. Khitun, “Magnonic combinatorial memory,”npj Spintronics, vol. 2, p. 2, 2024

work page 2024
[6]

Implementing a magnonic reservoir computer model based on time-delay multiplex- ing,

S. Watt, M. Kostylev, A. B. Ustinov, and B. A. Kalinikos, “Implementing a magnonic reservoir computer model based on time-delay multiplex- ing,”Physical Review Applied, vol. 15, p. 064060, 2021

work page 2021
[7]

Enhancing computational per- formance of a spin-wave reservoir computer with input synchronization,

S. Watt, M. Kostylev, and A. B. Ustinov, “Enhancing computational per- formance of a spin-wave reservoir computer with input synchronization,” Journal of Applied Physics, vol. 129, p. 044902, Jan. 2021

work page 2021
[8]

Learning trajectories from spin-wave dynam- ics,

S. Watt and M. Kostylev, “Learning trajectories from spin-wave dynam- ics,”Physical Review Applied, vol. 19, p. 064029, 2023

work page 2023
[9]

A current-controlled magnonic reservoir for physical reservoir computing,

A. B. Ustinov, R. V . Haponchyk, and M. Kostylev, “A current-controlled magnonic reservoir for physical reservoir computing,”Applied Physics Letters, vol. 124, p. 042405, 2024

work page 2024
[10]

Bloembergen problem based magnonic physical reservoir computing,

M. Kostylev, “Bloembergen problem based magnonic physical reservoir computing,”Journal of Applied Physics, vol. 137, p. 213903, 2025

work page 2025
[11]

Macromagnetic simulation for reservoir computing utilizing spin dynamics in magnetic tunnel junctions,

T. Furuta, K. Fujii, K. Nakajima, S. Tsunegi, H. Kubota, Y . Suzuki, and S. Miwa, “Macromagnetic simulation for reservoir computing utilizing spin dynamics in magnetic tunnel junctions,”Physical Review Applied, vol. 10, p. 034063, 2018

work page 2018
[12]

Jaeger,Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the echo state network approach

H. Jaeger,Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the echo state network approach. GMD- Forschungszentrum Informationstechnik Bonn, 2002, vol. 5

work page 2002
[13]

Intrinsic nonlinearity probed by intermodulation distortion microwave measurements on high quality MgB2 thin films,

G. Cifariello, M. Aurino, E. Di Gennaro, G. Lamura, A. Andreone, P. Orgiani, X. X. Xi, and J.-C. Vill ´egier, “Intrinsic nonlinearity probed by intermodulation distortion microwave measurements on high quality MgB2 thin films,”Applied Physics Letters, vol. 88, p. 142510, 2006. 5

work page 2006
[14]

Probing the nonlinearities arising in the microwave response of superconductors by intermodulation distortion,

A. Andreone, G. Cifariello, E. Di Gennaro, G. Lamura, N. Emery, C. Herold, J. F. Mareche, P. Lagrange, P. Orgiani, X. X. Xi, and J. C. Villegier, “Probing the nonlinearities arising in the microwave response of superconductors by intermodulation distortion,”IEEE Transactions on Applied Superconductivity, vol. 17, p. 3640–3643, 2007

work page 2007
[15]

MgB2 nonlinear properties investigated under localized high rf magnetic field excitation,

T. Tamin, B. G. Ghamsari, T. Tan, C. G. Zhuang, X. X. Xi, and S. M. Anlage, “MgB2 nonlinear properties investigated under localized high rf magnetic field excitation,”Physical Review Special Topics - Accelerators and Beams, vol. 15, p. 122002, 2012

work page 2012
[16]

Higher order harmonic detection for exploring nonlinear interactions with nanoscale resolution,

R. K. Vasudevan, M. B. Okatan, I. Rajapaksa, Y . Kim, D. Marincel, S. Trolier-McKinstry, S. Jesse, N. Valanoor, and S. V . Kalinin, “Higher order harmonic detection for exploring nonlinear interactions with nanoscale resolution,”Scientific Reports, vol. 3, p. 2677, 2013

work page 2013
[17]

Nonlinear transition shift in heat assisted magnetic recording,

B. F. Valcu and M. Alex, “Nonlinear transition shift in heat assisted magnetic recording,”IEEE Transactions on Magnetics, vol. 49, no. 7, pp. 3648–3651, 2013

work page 2013
[18]

A. V . Oppenheim, A. S. Willsky, and S. H. Nawab,Signals & systems (2nd ed.). Prentice-Hall, Inc., 1996

work page 1996
[19]

A closed set of normal orthogonal functions,

J. L. Walsh, “A closed set of normal orthogonal functions,”American Journal of Mathematics, vol. 45, pp. 5–24, 1923

work page 1923
[20]

Walsh functions and their applications,

K. G. Beauchamp and C. Yuen, “Walsh functions and their applications,” IEEE Transactions on Systems, Man, and Cybernetics, pp. 794–795, 1976

work page 1976

[1] [1]

Nonlinear spin waves in magnetic film feedback rings,

M. Wu, “Nonlinear spin waves in magnetic film feedback rings,” inSolid State Physics. Academic Press, 2010, vol. 62, pp. 163–224

work page 2010

[2] [2]

Prabhakar and D

A. Prabhakar and D. D. Stancil,Spin waves: Theory and applications. Springer, New York, 2009

work page 2009

[3] [3]

Combinatorial logic devices based on a multi-path active ring circuit,

A. Khitun and M. Balinskiy, “Combinatorial logic devices based on a multi-path active ring circuit,”Scientific Reports, vol. 12, p. 9482, 2022

work page 2022

[4] [4]

Magnonic active ring co-processor,

M. Balynsky, Y . Khivintsev, A. Kozhevnikov, Y . Nikulin, V . Sakharov, Y . Filimonov, and A. Khitun, “Magnonic active ring co-processor,” Journal of Applied Physics, vol. 133, p. 023904, 2023

work page 2023

[5] [5]

Magnonic combinatorial memory,

M. Balinskyy and A. Khitun, “Magnonic combinatorial memory,”npj Spintronics, vol. 2, p. 2, 2024

work page 2024

[6] [6]

Implementing a magnonic reservoir computer model based on time-delay multiplex- ing,

S. Watt, M. Kostylev, A. B. Ustinov, and B. A. Kalinikos, “Implementing a magnonic reservoir computer model based on time-delay multiplex- ing,”Physical Review Applied, vol. 15, p. 064060, 2021

work page 2021

[7] [7]

Enhancing computational per- formance of a spin-wave reservoir computer with input synchronization,

S. Watt, M. Kostylev, and A. B. Ustinov, “Enhancing computational per- formance of a spin-wave reservoir computer with input synchronization,” Journal of Applied Physics, vol. 129, p. 044902, Jan. 2021

work page 2021

[8] [8]

Learning trajectories from spin-wave dynam- ics,

S. Watt and M. Kostylev, “Learning trajectories from spin-wave dynam- ics,”Physical Review Applied, vol. 19, p. 064029, 2023

work page 2023

[9] [9]

A current-controlled magnonic reservoir for physical reservoir computing,

A. B. Ustinov, R. V . Haponchyk, and M. Kostylev, “A current-controlled magnonic reservoir for physical reservoir computing,”Applied Physics Letters, vol. 124, p. 042405, 2024

work page 2024

[10] [10]

Bloembergen problem based magnonic physical reservoir computing,

M. Kostylev, “Bloembergen problem based magnonic physical reservoir computing,”Journal of Applied Physics, vol. 137, p. 213903, 2025

work page 2025

[11] [11]

Macromagnetic simulation for reservoir computing utilizing spin dynamics in magnetic tunnel junctions,

T. Furuta, K. Fujii, K. Nakajima, S. Tsunegi, H. Kubota, Y . Suzuki, and S. Miwa, “Macromagnetic simulation for reservoir computing utilizing spin dynamics in magnetic tunnel junctions,”Physical Review Applied, vol. 10, p. 034063, 2018

work page 2018

[12] [12]

Jaeger,Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the echo state network approach

H. Jaeger,Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the echo state network approach. GMD- Forschungszentrum Informationstechnik Bonn, 2002, vol. 5

work page 2002

[13] [13]

Intrinsic nonlinearity probed by intermodulation distortion microwave measurements on high quality MgB2 thin films,

G. Cifariello, M. Aurino, E. Di Gennaro, G. Lamura, A. Andreone, P. Orgiani, X. X. Xi, and J.-C. Vill ´egier, “Intrinsic nonlinearity probed by intermodulation distortion microwave measurements on high quality MgB2 thin films,”Applied Physics Letters, vol. 88, p. 142510, 2006. 5

work page 2006

[14] [14]

Probing the nonlinearities arising in the microwave response of superconductors by intermodulation distortion,

A. Andreone, G. Cifariello, E. Di Gennaro, G. Lamura, N. Emery, C. Herold, J. F. Mareche, P. Lagrange, P. Orgiani, X. X. Xi, and J. C. Villegier, “Probing the nonlinearities arising in the microwave response of superconductors by intermodulation distortion,”IEEE Transactions on Applied Superconductivity, vol. 17, p. 3640–3643, 2007

work page 2007

[15] [15]

MgB2 nonlinear properties investigated under localized high rf magnetic field excitation,

T. Tamin, B. G. Ghamsari, T. Tan, C. G. Zhuang, X. X. Xi, and S. M. Anlage, “MgB2 nonlinear properties investigated under localized high rf magnetic field excitation,”Physical Review Special Topics - Accelerators and Beams, vol. 15, p. 122002, 2012

work page 2012

[16] [16]

Higher order harmonic detection for exploring nonlinear interactions with nanoscale resolution,

R. K. Vasudevan, M. B. Okatan, I. Rajapaksa, Y . Kim, D. Marincel, S. Trolier-McKinstry, S. Jesse, N. Valanoor, and S. V . Kalinin, “Higher order harmonic detection for exploring nonlinear interactions with nanoscale resolution,”Scientific Reports, vol. 3, p. 2677, 2013

work page 2013

[17] [17]

Nonlinear transition shift in heat assisted magnetic recording,

B. F. Valcu and M. Alex, “Nonlinear transition shift in heat assisted magnetic recording,”IEEE Transactions on Magnetics, vol. 49, no. 7, pp. 3648–3651, 2013

work page 2013

[18] [18]

A. V . Oppenheim, A. S. Willsky, and S. H. Nawab,Signals & systems (2nd ed.). Prentice-Hall, Inc., 1996

work page 1996

[19] [19]

A closed set of normal orthogonal functions,

J. L. Walsh, “A closed set of normal orthogonal functions,”American Journal of Mathematics, vol. 45, pp. 5–24, 1923

work page 1923

[20] [20]

Walsh functions and their applications,

K. G. Beauchamp and C. Yuen, “Walsh functions and their applications,” IEEE Transactions on Systems, Man, and Cybernetics, pp. 794–795, 1976

work page 1976