Pith. sign in

REVIEW 4 major objections 5 minor 44 references

A Novel Discrete Memristor-Coupled Heterogeneous Dual-Neuron Model and Its Application in Multi-Scenario Image Encryption

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper constructs a discrete memristor-coupled heterogeneous dual-neuron network (MHDNN) that shows hyperchaotic behavior and supports a real-time image encryption platform for police IoT.

desk verdict New map, real dynamics, broken encryption: Algorithm 2 uses sorted real values as swap indices, so the cipher as printed cannot run and the security metrics are unconnected to a defined algorithm. read the letter →

arxiv 2505.24294 v1 pith:DSUQNKMQ submitted 2025-05-30 cs.IR

Yi Zou , Mengjiao Wang , Xinan Zhang , Herbert Ho-Ching Iu This is my paper
classification cs.IR MSC 37D4594A60
keywords discretememristorheterogeneousdual-neuronnetworkhyperchaosfiringpatternssynchronizationimageencryptionSTM32pseudorandomnumbergenerator
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes a discrete memristor-coupled heterogeneous dual-neuron network (MHDNN): one Chialvo neuron and one Rulkov neuron joined by a discrete tanh memristor acting as a synapse, described by a three-dimensional difference equation. The paper argues that this map exhibits hyperchaotic, chaotic, periodic, and quasi-periodic regimes depending on parameters and initial conditions, and that it produces a variety of spiking and bursting firing patterns as well as synchronization between the two neurons under different coupling strengths. It further claims that a pseudorandom number generator built from the map's orbit passes the NIST SP800-22 test suite, that the map can be implemented on an STM32F407VET6 microcontroller, and that an encryption algorithm combining chaotic permutation, bidirectional diffusion, and a data-dependent S-box securely encrypts color images. If these claims hold, the map offers a low-cost, hardware-friendly source of hyperchaos for real-time image encryption in police Internet of Things systems.

What carries the argument

The load-bearing object is the MHDNN difference equation itself, built from the discrete tanh memristor $V_n = \tanh(\varphi_n) I_n$, $\varphi_{n+1}=\varphi_n+I_n$. The coupling term $m\tanh(z_n)(y_n-x_n)$ appears symmetrically in both neuron equations and the flux update uses the potential difference $y_n-x_n$, giving the system the feedback needed for hyperchaos. This single map plays a double role: it supplies the orbits analyzed for fixed points, firing patterns, and synchronization, and it generates the keystream ($X$, $X_1$, $X_2$, $SBT$) and the dynamic $16\times 16$ S-box for the encryption algorithm.

What would settle it

Run the MHDNN-based PRNG on an STM32F407VET6, extract bits 25 to 32 of IEEE 754 floats from the resulting sequences, and submit the bitstream to the NIST SP800-22 suite; if any P-value falls below the 0.01 significance level or the orbit settles into a short periodic cycle, then the claimed real-time hardware encryption security is contradicted.

Watch

Extended reading notes

Core claim

The central discovery is the MHDNN map $x_{n+1}=a/(1+x_n^2)+m\tanh(z_n)(y_n-x_n)+h$, $y_{n+1}=cy_n^2 e^{b-y_n}+m\tanh(z_n)(y_n-x_n)+3$, $z_{n+1}=z_n+y_n-x_n$, where $x$ and $y$ are the membrane potentials of the Chialvo and Rulkov neurons and $z$ is the memristor flux. With parameters $(a,h,b,c,m)=(-3.4,1.3,1.5,-1.5,0.2)$ and initial state $(0.1,0.1,0.1)$, numerical analysis shows hyperchaotic orbits, and sweeping $a\in[-3,-1]$ or $m\in[0,0.35]$ produces transitions among hyperchaotic, chaotic, periodic, and quasi-periodic states. The paper also establishes that the fixed points of the map form an infinite family whose stability depends on the parameters, that the two neurons can synchronize with Pearson correlation approaching 1 in one configuration as the coupling strength decreases, and that extracting bits 25 to 32 of IEEE 754 representations of consecutive iterates yields pseudorandom sequences that pass all 15 NIST subtests. On top of this, the authors construct an image encryption algorithm and demonstrate it on two hardware platforms for multi-scenario police image encryption.

Load-bearing premise

The entire application story depends on the assumption that iterating the MHDNN on the STM32F407VET6 in finite precision still produces the same hyperchaotic behavior and random-looking bitstreams as the ideal double-precision map, since the hardware validation shown is only an oscilloscope phase portrait and no randomness test is run on the device's output.

Editorial extensions

If this is right

  • The MHDNN gives a three-dimensional discrete system whose hyperchaotic region is tunable through the coupling strength $m$ and the parameter $a$, so other chaos-based applications can reuse the same map with different parameter settings.
  • The proposed encryption algorithm with a $10^{128}$ key space, uniform histograms, entropy close to 8, and NPCR/UACI near the ideal values would resist brute-force, statistical, and differential attacks if the underlying chaos is sound.
  • The hardware demonstration on STM32F407VET6 and the ESP8266-based WiFi transmission platform indicates that the encryption can run on widely available, low-cost microcontrollers rather than requiring high-end computing infrastructure.
  • The synchronization finding that weaker coupling can increase the Pearson correlation in one neuron configuration suggests that coupling strength can be used as a control knob for synchronization in memristive neural networks.
  • A PRNG based on map iterates passing NIST SP800-22 supports use of the MHDNN as a random source in secure communication and other IoT security protocols.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper tests the PRNG only on the ideal floating-point map, not on the STM32 itself; a natural next check would be to run the NIST suite on bits extracted from the hardware, since finite-precision arithmetic can shorten chaotic orbits.
  • The counterintuitive synchronization result (lower coupling giving correlation closer to 1) is shown for one parameter set; testing other heterogeneous neuron pairs could show whether it is a general property of tanh-memristor coupling or specific to this map.
  • The dynamic S-box construction depends on sorting chaotic subsequences and swapping entries; the same scheme could be adapted to video frames by re-keying the map per frame, which the paper only mentions as future work for audio and text.
  • The infinite fixed-point set with one eigenvalue always on the unit circle suggests the MHDNN sits at a critical boundary; if that boundary is what generates the rich firing patterns, then tweaking the memristor nonlinearity (e.g., replacing tanh with other sigmoids) would produce a family of related maps with different dynamics.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper introduces a discrete memristor-coupled heterogeneous dual-neuron network (MHDNN) formed by coupling one-dimensional Chialvo and Rulkov maps through a discrete tanh memristor. It analyzes fixed-point stability, reports bifurcation diagrams, Lyapunov exponents, basins of attraction, firing patterns, and synchronization, and describes an STM32 hardware implementation. The second half proposes a chaotic image-encryption algorithm based on MHDNN-generated sequences, a dynamic S-box, permutation, and diffusion, and reports security metrics plus two police-IoT hardware scenarios. The central claims are that the map exhibits hyperchaotic/chaotic/periodic behavior and that the proposed encryption platform is secure and practical.

Significance. If the numerical and hardware claims are reproducible, the MHDNN could be a useful addition to the discrete memristive-neuron literature, and the hardware implementation would be a practical demonstration for chaos-based IoT encryption. The paper reports extensive numerical evidence (bifurcations, Lyapunov exponents, basins, firing patterns), NIST randomness results, and hardware phase portraits, all of which are valuable if they correspond to a well-defined procedure. However, the manuscript does not provide code or data, and the encryption contribution is undermined by a serious internal inconsistency in Algorithm 2: the S-box construction cannot execute as printed, so the reported security metrics are not connected to a definable cipher. The fixed-point and eigenvalue analysis also contains algebraic inconsistencies. These issues are load-bearing for the stated contributions and must be corrected before the paper can be considered reliable.

major comments (4)
  1. [§IV-C, Algorithm 2] The dynamic S-box construction is not executable as written. In lines 4–6, `sort(list)` returns sorted values, but the sort indices are discarded; therefore `L1(i)` and `L2(i)` are floating-point chaotic values, not integer positions in `[1,256]` (or `[1,128]` for the swap loop). The instruction `swap(s1, L1(i), L2(i))` is consequently invalid in MATLAB, Python, or any conventional language, and a 16×16 S-box cannot be produced. Since the ciphertexts, histograms, correlation coefficients, entropy, NPCR/UACI, and robustness results in §IV-D all depend on this S-box, those metrics correspond to a procedure that is not actually specified. The authors must provide a complete, executable version of the S-box algorithm (ideally with reference code), rerun the security evaluations, and either confirm or correct the reported values.
  2. [§II-D, Eqs. (3)–(4)] The fixed-point equation for x* is inconsistent with the stated parameters. Setting y*=x* and z*=d in Eq. (3) gives `a/(1+x*^2) + h - c x*^2 e^{b-x*} - 3 = 0`; Eq. (4) follows only for h=3, while the text immediately below states (b,c,m,h)=(2,1,0.2,1). This discrepancy affects the classification of fixed points in Fig. 4 and the subsequent stability discussion. Please correct Eq. (4) or the assigned parameter values and redo the related analysis if necessary.
  3. [§II-D, Eqs. (8)–(10)] The eigenvalue and instability formulas contain algebraic and typographical errors. Eqs. (8) and (9) are both labelled λ2, Eq. (9) is missing the minus sign before `4(m tanh(d))^2`, and Eq. (10) uses an expression `(J11^2 - J22^2)^2` that does not follow from Eq. (6). Because the instability criterion (10) is the basis for the stability conclusions, these formulas must be re-derived and stated correctly.
  4. [§IV-B and §IV-A] The hardware validation is too weak to support the claim that the MHDNN is 'implemented and validated on the STM32 platform.' Fig. 12 only shows oscilloscope phase portraits, while the NIST SP800-22 results in Table V are reported for software-generated sequences. No statistical tests are reported on the finite-precision bitstream extracted on the device, despite the known risk that fixed-point iterations can collapse into short cycles. The authors should either provide a statistical evaluation of the hardware-generated PRNG output or explicitly weaken the hardware-validation claim.
minor comments (5)
  1. [§III-A] The text refers to 'Fig. 3(b), (c), (e), and (f)' when describing attractors for Case 1; the relevant panels are in Fig. 5.
  2. [§III-B] The caption of Fig. 8 lists parameters (a,b,h)=(−3.4,−1.5,1.3), but the text immediately following it states parameters a=−3.4, b=3, h=1.3 for the firing-pattern analysis; please reconcile this discrepancy.
  3. [§IV-C] The description of Algorithm 1 calls the MHDNN a 'two-dimensional Lorenz-inspired recurrent neural map,' which is factually incorrect; the model is a memristor-coupled Chialvo–Rulkov system, not a Lorenz system.
  4. [§IV-C] Algorithm 1 says the system is initialized with a secret key K but does not specify how K is mapped to the initial conditions or the eight system parameters. Without this mapping, the key-space analysis in §IV-D.2 and any key-sensitivity claims are not fully defined.
  5. [Eq. (12)] The cross-correlation formula is missing parentheses in the denominator; it should read `√(Σ x(n)^2) √(Σ y(n)^2)` or an equivalent explicitly grouped form.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the MHDNN construction, fixed-point stability, PRNG, and encryption metrics are all computed from the model after it is defined; the only potential self-citation is not load-bearing.

full rationale

The paper's derivation chain is not circular. Equation (2) defines the MHDNN, and the fixed-point equations (3)-(9) are derived directly from that definition; the bifurcation diagrams, Lyapunov exponents, basins of attraction, and firing patterns in Section III are numerical observations for hand-selected parameter values, not predictions fitted to a target output. The PRNG in Section IV-A extracts bits from the iterates of Eq. (2) and is validated against the external NIST SP800-22 suite, so the randomness claim is an empirical check rather than an input to the construction. The image-encryption security metrics in Section IV-D (histograms, correlation, entropy, NPCR/UACI, robustness) are measured after the cipher is defined, and the keystream and S-box are generated from chaotic sequences rather than fitted to make the metrics reach their reported values. The only apparent self-citation, reference [23] by M. Wang and Z. Yi, appears in a background list of chaotic image-encryption works and is not used to justify any derivation, uniqueness, or stability claim; even if it is a self-citation, it is not load-bearing. A separate, non-circular correctness issue exists in Algorithm 2, where the sorted values L1(i) and L2(i) are used as swap indices instead of the sort indices; this makes the printed pseudocode non-executable and the reported encryption metrics unreproducible, but it is an internal consistency flaw rather than a reduction of a claimed result to its own inputs.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The MHDNN map is constructed, not derived, so it carries hand-picked parameters and background assumptions. The main unproved premises are the faithfulness of the 1D neuron reductions, the validity of the tanh memristor as a synapse, and the finite-precision hardware implementation preserving chaos. No new physical entities are introduced.

free parameters (2)
  • Neuron and coupling parameters (a, b, c, h, m) = PRNG: a=-3.4, b=1.5, c=4.28, h=1.3, m=0.2; encryption key variant: a=-3.4, b=-1.5, c=-1.5, h=3, m=0.5
    Hand-selected from bifurcation diagrams to produce chaotic and hyperchaotic attractors; no external data fitting. The encryption key also includes these values, so the claimed key space depends on them.
  • Initial conditions (x0, y0, z0) = 0.1, 0.1, 0.1 as default; z0 varied in the firing pattern study
    Chosen by hand and treated as part of the secret key. The map's sensitivity to initial conditions is the basis for the encryption claim, so these values are load-bearing for the application.
assumptions (4)
  • ad hoc to paper The one-dimensional Chialvo and Rulkov maps in Table I adequately represent the original two-dimensional neuron models for the purposes of the MHDNN.
    Section II-B states the reduction 'focusing only on x_n' without a derivation or demonstration that omitted variables do not change qualitative firing behavior. The biological relevance claim depends on this.
  • domain assumption The discrete tanh memristor defined by Eq. (1) exhibits the three memristor fingerprint features and is a valid synapse model.
    Section II-A verifies fingerprints only through numerical examples in Fig. 1, not analytically. The entire coupling mechanism of MHDNN relies on this memristor model.
  • domain assumption The STM32 finite-precision implementation faithfully reproduces the ideal dynamics of Eq. (2).
    Section IV-B shows an oscilloscope attractor purportedly matching MATLAB, but no quantitative comparison or bitstream statistics are given. The hardware encryption platform depends on this assumption.
  • standard math Lyapunov exponents and basins reported in Section III are computed with standard algorithms and reliable numerical accuracy.
    The paper reports LEs and attractor classifications but does not describe the algorithm, iteration lengths, or numerical methods. These computations are the evidence for the claimed chaos and hyperchaos.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Novel Discrete Memristor-Coupled Heterogeneous Dual-Neuron Model and Its Application in Multi-Scenario Image Encryption." pith.science (2026). https://pith.science/paper/DSUQNKMQ

@misc{pith2026250524294,
  author       = {Pith},
  title        = {Pith review of: A Novel Discrete Memristor-Coupled Heterogeneous Dual-Neuron Model and Its Application in Multi-Scenario Image Encryption},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DSUQNKMQ}},
  note         = {Machine review of arXiv:2505.24294}
}
read the original abstract

Simulating brain functions using neural networks is an important area of research. Recently, discrete memristor-coupled neurons have attracted significant attention, as memristors effectively mimic synaptic behavior, which is essential for learning and memory. This highlights the biological relevance of such models. This study introduces a discrete memristive heterogeneous dual-neuron network (MHDNN). The stability of the MHDNN is analyzed with respect to initial conditions and a range of neuronal parameters. Numerical simulations demonstrate complex dynamical behaviors. Various neuronal firing patterns are investigated under different coupling strengths, and synchronization phenomena between neurons are explored. The MHDNN is implemented and validated on the STM32 hardware platform. An image encryption algorithm based on the MHDNN is proposed, along with two hardware platforms tailored for multi-scenario police image encryption. These solutions enable real-time and secure transmission of police data in complex environments, reducing hacking risks and enhancing system security.

Figures

Figures reproduced from arXiv: 2505.24294 by the authors.

Figure 1
Figure 1. Hysteresis loops of a discrete memristor with sinusoidal [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. D. Fix point Fixed point analysis is a tool used to analyze dynamic systems, and the stability of the MHDNN can be determined [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 5
Figure 5. Bifurcation and phase diagrams with varying parameter [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Different types of fixed points on the a − x ∗ plane. III. DYNAMICAL BEHAVIOR A. Dynamic behavior regarding parameters Bifurcation diagrams, LEs, and attractors are essential tools for analyzing the dynamical behavior of chaotic systems. Fixing the initial conditions a…
Figure 6
Figure 6. Figure 6: Bifurcation and phase diagrams with varying parameter [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 8
Figure 8. Figure 8: Memristor dynamics based on 2D bifurcation and LLE [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]
Figure 9
Figure 9. Figure 9: Firing patterns of variable x in the MHDNN under different memristor parameters (z0, m, c): (a) Hyperchaotic bursting (1, -0.3, 0.3); (b) Post-spike hyperchaotic oscillation (1, -0.3, 0.2); (c) Transient chaotic bursting (0.1, -0.2, 0.2); (d) Transient quasi-periodic b…
Figure 10
Figure 10. Figure 10: Evolution of membrane potential x and cross￾correlation in Case 1 under different coupling strengths m: (a1–a2) m = 0.2, (b1–b2) m = 0.1, (c1–c2) m = −0.3, (d1–d2) m = −0.4 [PITH_FULL_IMAGE:figures/full_fig_p006_10.png]
Figure 11
Figure 11. Figure 11: Membrane potential x and cross-correlation of two neurons in Case 2 under different coupling strengths m: (a1–a2) m = 0, (b1–b2) m = 0.1, (c1–c2) m = 0.2, (d1–d2) m = −0.5. TABLE IV: Pearson correlation coefficients for Case 1 and Case 2 Case 1 Case 2 m r m r 0.2 0.14…
Figure 12
Figure 12. Figure 12: Phase portrait plots of hyperchaotic attractors acquired [PITH_FULL_IMAGE:figures/full_fig_p007_12.png]
Figure 13
Figure 13. Figure 13: Image encryption and decryption results: (a) Original [PITH_FULL_IMAGE:figures/full_fig_p008_13.png]
Figure 14
Figure 14. Figure 14: Histogram analysis results. (a) Original image; (b) [PITH_FULL_IMAGE:figures/full_fig_p008_14.png]
Figure 15
Figure 15. Figure 15: Histogram analysis results. (a) Original image; (b) [PITH_FULL_IMAGE:figures/full_fig_p009_15.png]
Figure 16
Figure 16. Figure 16: Histogram analysis results. (a) Original image; (b) [PITH_FULL_IMAGE:figures/full_fig_p009_16.png]
Figure 17
Figure 17. Figure 17: Correlation analysis in different directions. [PITH_FULL_IMAGE:figures/full_fig_p009_17.png]
Figure 19
Figure 19. Figure 19: (a) 10% SPN; (b) 30% SPN; (c) GN (mean 0, variance [PITH_FULL_IMAGE:figures/full_fig_p010_19.png]
Figure 18
Figure 18. Figure 18: Cropping of encrypted images with different shapes [PITH_FULL_IMAGE:figures/full_fig_p010_18.png]
Figure 20
Figure 20. Figure 20: The process of encrypting police data. (a) Cloud encryption; (b) WiFi real-time encryption. [PITH_FULL_IMAGE:figures/full_fig_p011_20.png]
Figure 21
Figure 21. Figure 21: Hardware device used to encrypt police data. (a) [PITH_FULL_IMAGE:figures/full_fig_p011_21.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

44 extracted references · 43 canonical work pages

  1. [1]

    M. J. Ison, R. Q. Quiroga, I. Fried, Rapid encoding of new memories by individual neurons in the human brain, Neuron 87 (1) (2015) 220–230

    work page 2015

  2. [2]

    A. M. Zyarah, K. Gomez, D. Kudithipudi, Neuromorphic system for spatial and temporal information processing, IEEE Transactions on Computers 69 (8) (2020) 1099–1112

    work page 2020

  3. [3]

    J. Ma, J. Tang, A review for dynamics in neuron and neuronal network, Nonlinear Dynamics 89 (2017) 1569–1578

    work page 2017

  4. [4]

    A. L. Hodgkin, A. F. Huxley, A quantitative description of membrane current and its application to conduction and excitation in nerve, The Journal of physiology 117 (4) (1952) 500

    work page 1952

  5. [5]

    M. Chen, J. Qi, H. Wu, Q. Xu, B. Bao, Bifurcation analyses and hardware experiments for bursting dynamics in non-autonomous mem- ristive fitzhugh-nagumo circuit, Science China Technological Sciences 63 (2020) 1035–1044

    work page 2020

  6. [6]

    A. G. Chesebro, L. R. Mujica-Parodi, C. Weistuch, Ion gradient-driven bifurcations of a multi-scale neuronal model, Chaos, Solitons & Fractals 167 (2023) 113120

    work page 2023

  7. [7]

    Zhang, H

    S. Zhang, H. Zhang, C. Wang, H. Lin, Bionic modeling and dynam- ics analysis of heterogeneous brain regions connected by memristive synaptic crosstalk, Chaos, Solitons & Fractals 179 (2024) 114459

    work page 2024

  8. [8]

    Q. Xu, T. Liu, S. Ding, H. Bao, Z. Li, B. Chen, Extreme multistability and phase synchronization in a heterogeneous bi-neuron rulkov network with memristive electromagnetic induction, Cognitive Neurodynamics 17 (3) (2023) 755–766

    work page 2023

Show all 44 references
  1. [9]

    Z. Wang, F. Parastesh, K. Rajagopal, I. I. Hamarash, I. Hussain, Delay- induced synchronization in two coupled chaotic memristive hopfield neural networks, Chaos, Solitons & Fractals 134 (2020) 109702

    2020

  2. [10]

    Q. Lai, Z. Wan, P. D. K. Kuate, Generating grid multi-scroll attractors in memristive neural networks, IEEE Transactions on Circuits and Systems I: Regular Papers 70 (3) (2022) 1324–1336

    2022

  3. [11]

    Indiveri, B

    G. Indiveri, B. Linares-Barranco, R. Legenstein, G. Deligeorgis, T. Pro- dromakis, Integration of nanoscale memristor synapses in neuromorphic computing architectures, Nanotechnology 24 (38) (2013) 384010

    2013

  4. [12]

    C. D. Schuman, T. E. Potok, R. M. Patton, J. D. Birdwell, M. E. Dean, G. S. Rose, J. S. Plank, A survey of neuromorphic computing and neural networks in hardware, arXiv preprint arXiv:1705.06963 (2017)

    2017 arXiv

  5. [13]

    Majhi, M

    S. Majhi, M. Perc, D. Ghosh, Dynamics on higher-order networks: A review, Journal of the Royal Society Interface 19 (188) (2022) 20220043

    2022

  6. [14]

    Y . Li, C. Li, T. Lei, Y . Yang, G. Chen, Offset boosting-entangled complex dynamics in the memristive rulkov neuron, IEEE Transactions on Industrial Electronics (2023)

    2023

  7. [15]

    X. Zhou, K. Sun, H. Wang, Z. Yao, Coexisting hyperchaos and mul- tistability in a discrete memristor-coupled bi-neuron model, Nonlinear Dynamics (2024) 1–15

    2024

  8. [16]

    M. K. Wouapi, B. H. Fotsin, E. B. M. Ngouonkadi, F. F. Kemwoue, Z. T. Njitacke, Complex bifurcation analysis and synchronization optimal control for hindmarsh–rose neuron model under magnetic flow effect, Cognitive neurodynamics 15 (2021) 315–347

    2021

  9. [17]

    Shang, K

    C. Shang, K. Sun, H. Wang, Z. Yao, S. He, Spatial patterns and chimera states in discrete memristor coupled neural networks, Nonlinear Dynamics 111 (21) (2023) 20347–20360

    2023

  10. [18]

    Q. Lai, G. Hu, U. Erkan, A. Toktas, A novel pixel-split image encryption scheme based on 2d salomon map, Expert Systems with Applications 213 (2023) 118845

    2023

  11. [19]

    X. Li, J. Mou, L. Xiong, Z. Wang, J. Xu, Fractional-order double-ring erbium-doped fiber laser chaotic system and its application on image encryption, Optics & Laser Technology 140 (2021) 107074

    2021

  12. [20]

    W. Feng, Q. Wang, H. Liu, Y . Ren, J. Zhang, S. Zhang, K. Qian, H. Wen, Exploiting newly designed fractional-order 3d lorenz chaotic system and 2d discrete polynomial hyper-chaotic map for high-performance multi- image encryption, Fractal and Fractional 7 (12) (2023) 887

    2023

  13. [21]

    H. Wen, Y . Lin, Cryptanalysis of an image encryption algorithm using quantum chaotic map and dna coding, Expert Systems with Applications 237 (2024) 121514

    2024

  14. [22]

    G. Ye, L. Guo, A visual meaningful encryption and hiding algorithm for multiple images, Nonlinear Dynamics (2024) 1–24

    2024

  15. [23]

    M. Wang, Z. Yi, Z. Li, A memristive ikeda map and its application in image encryption, Chaos, Solitons & Fractals 190 (2025) 115740

    2025

  16. [24]

    J. Liu, J. Zhang, Y . Wang, Secure communication via chaotic synchro- nization based on reservoir computing, IEEE Transactions on Neural Networks and Learning Systems 35 (1) (2022) 285–299

    2022

  17. [25]

    A. A. Kekha Javan, A. Zare, R. Alizadehsani, S. Balochian, Robust multi-mode synchronization of chaotic fractional order systems in the presence of disturbance, time delay and uncertainty with application in secure communications, Big Data and Cognitive Computing 6 (2) (2022) 51

    2022

  18. [26]

    W. Liu, K. Sun, S. He, H. Wang, The parallel chaotification map and its application, IEEE Transactions on Circuits and Systems I: Regular Papers 70 (9) (2023) 3689–3698

    2023

  19. [27]

    H. Bao, K. Li, J. Ma, Z. Hua, Q. Xu, B. Bao, Memristive effects on an improved discrete rulkov neuron model, Science China Technological Sciences 66 (11) (2023) 3153–3163

    2023

  20. [28]

    H. Lin, X. Deng, F. Yu, Y . Sun, Diversified butterfly attractors of memristive hnn with two memristive systems and application in iomt for privacy protection, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (2024)

    2024

  21. [29]

    H. Lin, C. Wang, L. Cui, Y . Sun, C. Xu, F. Yu, Brain-like initial- boosted hyperchaos and application in biomedical image encryption, IEEE Transactions on Industrial Informatics 18 (12) (2022) 8839–8850

    2022

  22. [30]

    D. R. Chialvo, Generic excitable dynamics on a two-dimensional map, Chaos, Solitons & Fractals 5 (3-4) (1995) 461–479

    1995

  23. [31]

    N. F. Rulkov, Regularization of synchronized chaotic bursts, Physical Review Letters 86 (1) (2001) 183

    2001

  24. [32]

    Q. Xu, L. Huang, N. Wang, H. Bao, H. Wu, M. Chen, Initial-offset- boosted coexisting hyperchaos in a 2d memristive chialvo neuron map and its application in image encryption, Nonlinear Dynamics 111 (21) (2023) 20447–20463

    2023

  25. [33]

    T. A. Stowe, E. G. Pitts, A. C. Leach, M. C. Iacino, F. Niere, B. Graul, K. F. Raab-Graham, J. T. Yorgason, M. J. Ferris, Diurnal rhythms in cholinergic modulation of rapid dopamine signals and associative learning in the striatum, Cell reports 39 (1) (2022)

    2022

  26. [34]

    Kapogiannis, M

    D. Kapogiannis, M. P. Mattson, Disrupted energy metabolism and neuronal circuit dysfunction in cognitive impairment and alzheimer’s disease, The Lancet Neurology 10 (2) (2011) 187–198

    2011

  27. [35]

    Esmaeili, K

    V . Esmaeili, K. Tamura, S. P. Muscinelli, A. Modirshanechi, M. Boscaglia, A. B. Lee, A. Oryshchuk, G. Foustoukos, Y . Liu, S. Crochet, et al., Rapid suppression and sustained activation of distinct cortical regions for a delayed sensory-triggered motor response, Neuron 109 (1...

    2021

  28. [36]

    Ye, H.-S

    G.-D. Ye, H.-S. Wu, X.-L. Huang, S.-Y . Tan, Asymmetric image en- cryption algorithm based on a new three-dimensional improved logistic chaotic map, Chinese Physics B 32 (3) (2023) 030504

    2023

  29. [37]

    H. Liu, H. Wan, K. T. Chi, J. L ¨u, An encryption scheme based on synchronization of two-layered complex dynamical networks, IEEE Transactions on Circuits and Systems I: Regular Papers 63 (11) (2016) 2010–2021

    2016

  30. [38]

    R. Chen, X. Li, L. Teng, X. Wang, An image encryption algorithm based on the lscmm chaotic map and bidirectional dynamic diffusion, Multimedia Tools and Applications 83 (2) (2024) 3681–3706

    2024

  31. [39]

    X. Wang, Y . Wang, S. Wang, Y . Zhang, X. Wu, A novel pseudo-random coupled lp spatiotemporal chaos and its application in image encryption, Chinese Physics B 27 (11) (2018) 110502

    2018

  32. [40]

    Liang, Q

    Z. Liang, Q. Qin, C. Zhou, An image encryption algorithm based on fibonacci q-matrix and genetic algorithm, Neural Computing and Applications 34 (21) (2022) 19313–19341

    2022

  33. [41]

    X. Kong, F. Yu, W. Yao, S. Cai, J. Zhang, H. Lin, Memristor-induced hyperchaos, multiscroll and extreme multistability in fractional-order hnn: Image encryption and fpga implementation, Neural Networks 171 (2024) 85–103

    2024

  34. [42]

    J. Sun, Y . Zhao, C. Li, Y . Wang, P. Liu, Dynamical analysis of mem- ristive hnn and medical image encryption via bi-directional permutation and multi-directional diffusion to pacs, IEEE Transactions on Circuits and Systems I: Regular Papers (2024)

    2024

  35. [43]

    L. Teng, X. Wang, F. Yang, Y . Xian, Color image encryption based on cross 2d hyperchaotic map using combined cycle shift scrambling and selecting diffusion, Nonlinear Dynamics 105 (2) (2021) 1859–1876

    2021

  36. [44]

    F. Yu, Y . Lin, W. Yao, S. Cai, H. Lin, Y . Li, Multiscroll hopfield neural network with extreme multistability and its application in video encryption for iiot, Neural Networks (2024) 106904

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.