pith. sign in

arxiv: 2605.05884 · v1 · submitted 2026-05-07 · 📡 eess.SP

From Multi-Port Models to Cascade Structures: Optimization of Active Unilateral Stacked Intelligent Metasurfaces

Andrea Abrardo , Giulio Bartoli , Alberto Toccafondi , Marco Di Renzo This is my paper

Pith reviewed 2026-05-08 07:14 UTC · model grok-4.3

classification 📡 eess.SP
keywords stacked intelligent metasurfacesS-parameter modelingnon-reciprocal networkscascade structuresgradient-based optimizationspectral efficiencyelectromagnetic accuracyunilateral interconnections
0
0 comments X

The pith

Modeling each unit cell as a non-reciprocal two-port network gives stacked intelligent metasurfaces a recursive cascade transfer function.

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

The paper establishes that representing active interconnections in stacked intelligent metasurfaces as strictly unilateral non-reciprocal two-port networks produces a feed-forward structure. This structure supports an exact recursive computation of the overall transfer function without losing electromagnetic detail. The resulting model yields a gradient-based optimizer whose complexity scales better than methods for reciprocal architectures. Full-wave simulations then map how spacing, gain, and array size trade off against channel diagonalization and spectral efficiency. If the mapping holds, designers gain a practical way to tune these surfaces for wireless performance.

Core claim

By modeling each unit cell as a non-reciprocal two-port network, the resulting SIM exhibits a feed-forward structure that enables a recursive, cascade-like representation of the end-to-end transfer function while preserving electromagnetic accuracy; this representation directly supports an efficient gradient-based optimization algorithm with reduced computational complexity compared to conventional reciprocal SIM architectures.

What carries the argument

The non-reciprocal two-port S-parameter model for each unit cell, which enforces unidirectional signal flow and produces the recursive cascade form of the composite transfer function.

If this is right

  • The end-to-end transfer function can be evaluated recursively by multiplying two-port matrices layer by layer.
  • Gradient computation for optimization reduces to a backward pass through the same cascade structure.
  • Design parameters such as inter-layer spacing and active gain can be optimized directly for channel diagonalization.
  • Spectral-efficiency gains are achieved at lower computational cost than reciprocal multi-port solvers.
  • Larger SIM apertures become tractable because complexity grows linearly rather than combinatorially with the number of layers.

Where Pith is reading between the lines

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

  • The same unilateral cascade reduction could be applied to other active metasurface topologies that already use non-reciprocal elements.
  • Real-time adaptation algorithms might exploit the recursion to update phases or gains without full re-simulation.
  • If the unilateral assumption holds only above a minimum inter-layer distance, the model supplies an explicit lower bound on spacing for valid optimization.

Load-bearing premise

Treating the active interconnections as strictly unilateral non-reciprocal two-port networks fully captures electromagnetic behavior and introduces no hidden coupling that would break the cascade recursion.

What would settle it

A full-wave simulation or measurement of a fabricated SIM in which measured end-to-end S-parameters deviate measurably from the cascade prediction at the operating frequency would falsify the model.

Figures

Figures reproduced from arXiv: 2605.05884 by Alberto Toccafondi, Andrea Abrardo, Giulio Bartoli, Marco Di Renzo.

Figure 1
Figure 1. Figure 1: , which is characterized by a multi-antenna transmitter with L ports, a multi-antenna receiver with M ports, along with N ports corresponding to the N elements of a SIM that receives, processes, and retransmits electromagnetic waves. Let us suppose the SIM composed of Q stacked T-RISs, referred in the following to as layers. Each layer is composed of a receive array and a transmit array separated by one or… view at source ↗
Figure 2
Figure 2. Figure 2: FIGURE 2 view at source ↗
Figure 3
Figure 3. Figure 3: FIGURE 3 view at source ↗
Figure 4
Figure 4. Figure 4: FIGURE 4 view at source ↗
Figure 5
Figure 5. Figure 5: FIGURE 5 view at source ↗
Figure 6
Figure 6. Figure 6: FIGURE 6 view at source ↗
Figure 7
Figure 7. Figure 7: FIGURE 7 view at source ↗
Figure 12
Figure 12. Figure 12: FIGURE 12 view at source ↗
Figure 11
Figure 11. Figure 11: FIGURE 11 view at source ↗
read the original abstract

This paper develops a multi-port S-parameter framework for the analysis and optimization of stacked intelligent metasurfaces (SIMs) with unilateral active interconnections. By modeling each unit cell as a non-reciprocal two-port network, the resulting SIM exhibits a feed-forward structure that enables a recursive, cascade-like representation of the end-to-end transfer function while preserving electromagnetic accuracy. Based on this model, we derive an efficient gradient-based optimization algorithm with reduced computational complexity compared to conventional reciprocal SIM architectures. Numerical results, obtained from full-wave simulations, illustrate the trade-offs among inter-layer spacing, active gain, and SIM size in terms of channel diagonalization and achievable spectral efficiency.

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.

Referee Report

2 major / 2 minor

Summary. The paper develops a multi-port S-parameter framework for stacked intelligent metasurfaces (SIMs) with unilateral active interconnections. By modeling each unit cell as a non-reciprocal two-port network, the SIM is shown to exhibit a feed-forward structure permitting a recursive cascade representation of the end-to-end transfer function. This enables derivation of an efficient gradient-based optimization algorithm with reduced complexity relative to reciprocal architectures, with full-wave simulations used to illustrate design trade-offs in inter-layer spacing, active gain, and size for channel diagonalization and spectral efficiency.

Significance. If the unilateral non-reciprocal two-port modeling yields an algebraically exact cascade representation that preserves full electromagnetic accuracy without residual couplings, the work would provide a valuable complexity reduction for optimizing large-scale active SIMs, building directly on standard S-parameter network theory applied to a new interconnection assumption. This could advance practical deployment in wireless systems by enabling scalable, gradient-driven designs while maintaining physical fidelity.

major comments (2)
  1. [Model derivation and cascade representation] The central claim that the recursive cascade representation is exact and preserves electromagnetic accuracy (abstract and model section) requires explicit algebraic demonstration that the end-to-end transfer function obtained via the cascade recursion is identical to the product of the full multi-port S-matrices. Any unmodeled parasitic coupling, higher-order modes, or non-ideal port behavior would make the recursion approximate rather than an identity, directly affecting both the complexity-reduction claim and the accuracy guarantee. Please provide the derivation or clarify introduced approximations.
  2. [Numerical results and full-wave simulations] The numerical results section states that full-wave simulations illustrate the trade-offs, but it is unclear whether these simulations include a direct validation of the proposed cascade model against the full multi-port S-matrix computation for the same structures. Without such a comparison, it is difficult to confirm that electromagnetic accuracy is preserved under the unilateral assumption.
minor comments (2)
  1. [Notation and definitions] Clarify notation for S-parameters when transitioning between two-port unit-cell models and the overall multi-port SIM representation to improve readability.
  2. [Introduction and related work] Include a brief comparison to prior reciprocal SIM optimization methods to quantify the claimed complexity reduction.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which help clarify the presentation of our multi-port S-parameter framework. Below we address each major comment directly.

read point-by-point responses

  1. Referee: [Model derivation and cascade representation] The central claim that the recursive cascade representation is exact and preserves electromagnetic accuracy (abstract and model section) requires explicit algebraic demonstration that the end-to-end transfer function obtained via the cascade recursion is identical to the product of the full multi-port S-matrices. Any unmodeled parasitic coupling, higher-order modes, or non-ideal port behavior would make the recursion approximate rather than an identity, directly affecting both the complexity-reduction claim and the accuracy guarantee. Please provide the derivation or clarify introduced approximations.

    Authors: We agree that an explicit algebraic demonstration strengthens the central claim. Under the unilateral active interconnection model (each layer represented by a strictly feed-forward two-port network with S12 = 0), the overall multi-port S-matrix is block lower-triangular. Consequently, the end-to-end transfer function computed by the recursive cascade is algebraically identical to the direct product of the individual layer S-matrices; no residual backward coupling exists by construction of the unilateral assumption. In the revised manuscript we will insert a compact proof of this identity (new Appendix or expanded Section II) that starts from the partitioned S-matrix of the stacked structure and shows the recursion reproduces the exact block-triangular multiplication. The only modeling approximations remain those inherent to any S-parameter representation (e.g., port definitions, truncation of higher-order modes), which are already stated in the original text. revision: yes

  2. Referee: [Numerical results and full-wave simulations] The numerical results section states that full-wave simulations illustrate the trade-offs, but it is unclear whether these simulations include a direct validation of the proposed cascade model against the full multi-port S-matrix computation for the same structures. Without such a comparison, it is difficult to confirm that electromagnetic accuracy is preserved under the unilateral assumption.

    Authors: We acknowledge that a side-by-side numerical validation was not explicitly presented. The full-wave results in the original manuscript were obtained by extracting the S-parameters of each layer and then applying the cascade model for optimization; the physical structures were simulated only to generate those S-parameters. In the revision we will add a dedicated validation subsection that directly compares (i) the end-to-end transfer function obtained from the cascade recursion and (ii) the transfer function computed from the assembled full multi-port S-matrix, for representative 2-layer and 3-layer SIMs at several inter-layer spacings. Both quantities will be derived from the same full-wave data set, thereby confirming numerical equivalence within solver tolerance. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation follows from modeling assumption via standard network theory

full rationale

The paper introduces a modeling choice (each unit cell as a non-reciprocal two-port network with unilateral interconnections) and then derives the feed-forward cascade representation and optimization algorithm from it using standard S-parameter cascade formulas. This is a direct consequence of the chosen model rather than a reduction of any claimed prediction or result back to fitted data or self-citations. No equations or sections in the provided text show a self-definitional loop, a fitted parameter renamed as prediction, or load-bearing self-citation chains. The end-to-end transfer function is algebraically obtained from the assumed unilateral structure, and the accuracy claim is presented as a modeling property, not derived circularly. This is the expected non-circular outcome for a modeling paper.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that non-reciprocal two-port modeling of active unit cells yields an accurate feed-forward cascade while preserving electromagnetic behavior.

axioms (1)
  • domain assumption Modeling each unit cell as a non-reciprocal two-port network with unilateral active interconnections preserves electromagnetic accuracy and produces a feed-forward cascade structure.
    Directly invoked in the abstract as the basis for the recursive representation and optimization algorithm.

pith-pipeline@v0.9.0 · 5417 in / 1207 out tokens · 29367 ms · 2026-05-08T07:14:11.053739+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages

  1. [1]

    State of the art on stacked intelligent metasurfaces: Communication, sensing and computing in the wave domain,

    M. Di Renzo, “State of the art on stacked intelligent metasurfaces: Communication, sensing and computing in the wave domain,”arXiv preprint arXiv:2411.19687, 2024

    work page arXiv 2024

  2. [2]

    Stacked intelligent metasurfaces for efficient holographic MIMO communications in 6g,

    J. An, C. Xu, D. W. K. Ng, G. C. Alexandropoulos, C. Huang, C. Yuen, and L. Hanzo, “Stacked intelligent metasurfaces for efficient holographic MIMO communications in 6g,”IEEE J. Sel. Areas FIGURE 12. Transfer-function diagonality versus inter-layer spacing for K= 64andK= 256. Commun., vol. 41, no. 8, pp. 2380–2396, 2023. [Online]. Available: https://doi.o...

    work page doi:10.1109/jsac.2023.3288261 2023

  3. [3]

    Efficient beamforming and radiation pattern control using stacked intelligent metasurfaces,

    N. U. Hassan, J. An, M. Di Renzo, M. Debbah, and C. Yuen, “Efficient beamforming and radiation pattern control using stacked intelligent metasurfaces,”IEEE Open J. Commun. Soc., vol. 5, pp. 599–611, 2024. [Online]. Available: https://doi.org/10.1109/OJCOMS.2023.3349155

    work page doi:10.1109/ojcoms.2023.3349155 2024

  4. [4]

    Near-field beamforming for stacked intelligent metasurfaces-assisted MIMO networks,

    A. K. Papazafeiropoulos, P. Kourtessis, S. Chatzinotas, D. I. Kaklamani, and I. S. Venieris, “Near-field beamforming for stacked intelligent metasurfaces-assisted MIMO networks,”IEEE Wirel. Commun. Lett., vol. 13, no. 11, pp. 3035–3039, 2024. [Online]. Available: https://doi.org/10.1109/LWC.2024.3438840

    work page doi:10.1109/lwc.2024.3438840 2024

  5. [5]

    Stacked intelligent metasurface performs a 2d DFT in the wave domain for DOA estimation,

    J. An, C. Yuen, Y . L. Guan, M. Di Renzo, M. Debbah, H. V . Poor, and L. Hanzo, “Stacked intelligent metasurface performs a 2d DFT in the wave domain for DOA estimation,” inIEEE International Conference on Communications, ICC 2024, Denver, CO, USA, June 9-13, 2024. IEEE, 2024, pp. 3445–3451. [Online]. Available: https://doi.org/10.1109/ICC51166.2024.10622963

    work page doi:10.1109/icc51166.2024.10622963 2024

  6. [6]

    Two-dimensional direction-of-arrival estimation using stacked intelligent metasurfaces,

    ——, “Two-dimensional direction-of-arrival estimation using stacked intelligent metasurfaces,”IEEE J. Sel. Areas Commun., vol. 42, no. 10, pp. 2786–2802, 2024. [Online]. Available: https://doi.org/10.1109/JSAC.2024.3414613

    work page doi:10.1109/jsac.2024.3414613 2024

  7. [7]

    Stacked intelligent metasurfaces for integrated sensing and communications,

    H. Niu, J. An, A. K. Papazafeiropoulos, L. Gan, S. Chatzinotas, and M. Debbah, “Stacked intelligent metasurfaces for integrated sensing and communications,”IEEE Wirel. Commun. Lett., vol. 13, no. 10, pp. 2807–2811, 2024. [Online]. Available: https://doi.org/10.1109/LWC.2024.3447272

    work page doi:10.1109/lwc.2024.3447272 2024

  8. [8]

    Transmit beamforming design for ISAC with stacked intelligent metasurfaces,

    S. Li, F. Zhang, T. Mao, R. Na, Z. Wang, and G. K. Karagiannidis, “Transmit beamforming design for ISAC with stacked intelligent metasurfaces,”IEEE Transactions on Vehicular Technology, pp. 1–6, 2024

    work page 2024

  9. [9]

    Stacked intelligent metasurfaces assisted integrated-sensing-and-resistance anti jamming,

    C. Pei, K. Huang, L. Jin, X. Xu, Y . Zhou, and Y . Guo, “Stacked intelligent metasurfaces assisted integrated-sensing-and-resistance anti jamming,”IEEE Communications Letters, pp. 1–1, 2024

    work page 2024

  10. [10]

    Multi-user ISAC through stacked intelligent metasurfaces: New algorithms and experiments,

    Z. Wang, H. Liu, J. Zhang, R. Xiong, K. Wan, X. Qian, M. Di Renzo, and R. C. Qiu, “Multi-user ISAC through stacked intelligent metasurfaces: New algorithms and experiments,” in2024 IEEE Global Communications Conference (GLOBECOM): 8–12 December 2024, Cape Town, South Africa, Cape Town, SA, 2024

    work page 2024

  11. [11]

    Physically consistent modeling of stacked intelligent metasurfaces implemented with beyond diagonal ris,

    M. Nerini and B. Clerckx, “Physically consistent modeling of stacked intelligent metasurfaces implemented with beyond diagonal ris,”IEEE Communications Letters, vol. 28, no. 7, pp. 1693–1697, 2024

    work page 2024

  12. [12]

    A novel comprehensive multiport network model for stacked intelligent metasurfaces (sim) characterization and optimization,

    A. Abrardo, G. Bartoli, and A. Toccafondi, “A novel comprehensive multiport network model for stacked intelligent metasurfaces (sim) characterization and optimization,”IEEE Transactions on Communi- cations, vol. 73, no. 11, pp. 11 559–11 573, 2025

    work page 2025

  13. [13]

    Analysis and optimization of unilateral stacked intelligent metasurfaces by using a multi-port network model,

    A. Abrardo, G. Bartoli, A. Toccafondi, and M. Di Renzo, “Analysis and optimization of unilateral stacked intelligent metasurfaces by using a multi-port network model,” in2025 IEEE-APS Topical Conference on Antennas and Propagation in Wireless Communications (APWC), 2025, pp. 295–298. 10 VOLUME ,

    work page 2025

  14. [14]

    Design of reconfig- urable intelligent surfaces by using S-parameter multiport network theory—Optimization and full-wave validation,

    A. Abrardo, A. Toccafondi, and M. Di Renzo, “Design of reconfig- urable intelligent surfaces by using S-parameter multiport network theory—Optimization and full-wave validation,”IEEE Transactions on Wireless Communications, vol. 23, no. 11, pp. 17 084–17 102, 2024

    work page 2024

  15. [15]

    Modeling, optimization and electromagnetic validation of stacked intelligent metasurfaces by using a multiport network model,

    G. Pettanice, A. Abrardo, A. Toccafondi, and M. D. Renzo, “Modeling, optimization and electromagnetic validation of stacked intelligent metasurfaces by using a multiport network model,” 2026. [Online]. Available: https://arxiv.org/abs/2603.14266

    work page arXiv 2026

  16. [16]

    Stacked intelligent metasurfaces for efficient holographic mimo communications in 6g,

    J. An, C. Xu, D. W. K. Ng, G. C. Alexandropoulos, C. Huang, C. Yuen, and L. Hanzo, “Stacked intelligent metasurfaces for efficient holographic mimo communications in 6g,”IEEE Journal on Selected Areas in Communications, vol. 41, no. 8, pp. 2380–2396, 2023. VOLUME , 11

    work page 2023