Omni-Steering Plate for Elastic and EM Waves

• Omni-steering plate is a reconfigurable component that dynamically controls flexural and electromagnetic wave propagation through adjustable boundary conditions.
• It leverages topological valley-Hall effects in elastic media and varactor-tuned LC resonators in ISAC systems to achieve precise mode steering and full 360° coverage.
• Experimental results demonstrate <1 dB insertion loss in flexural wave routing and up to 87% NMSE reduction in ISAC, highlighting its robust performance.

An omni-steering plate is a reconfigurable hardware component—implemented in both elastic wave and electromagnetic domains—that enables arbitrary steering of propagating modes or radiated beams by local, controllable changes in boundary conditions. In elastic wave systems, the omni-steering plate routes flexural waves along complex planar trajectories using valley-Hall topological edge states, while in integrated sensing and communications (ISAC), it enables omnidirectional beamforming by coherently redirecting back-lobe energy via tunable passive array elements. The concept has been experimentally demonstrated for topological flexural waveguides and is central to emerging full-space ISAC antenna systems, where it replaces traditional reflectors to deliver 360° coverage and simultaneous radar/communication capabilities (Tang et al., 2019, Zhu et al., 18 Jan 2026).

1. Principle of Operation and Structural Realization

In elastic media, an omni-steering plate consists of a thin (1.6 mm) aluminum plate perforated by a hexagonal Bravais lattice (lattice constant $a=52$ mm) of six peripheral 3 mm holes per unit cell. The critical symmetry-breaking mechanism involves alternating “clamped” (enforcing $w=0$ at the rim) and “free” holes, with the set of clamped holes forming a triangle rotated by an angle $\theta$ about the cell center. By selecting $\theta=+15^\circ$ or $-15^\circ$, distinct valley-Hall phases (characterized by mass term sign $m$) are realized (Tang et al., 2019).

In electromagnetic applications, the omni-steering plate is realized as a linear array ($M$) of passive elements—each a parallel-LC resonator backed by metallic loops—positioned in the near-field ($D = \lambda$) behind an active antenna array. Each element supports dynamically tunable electric ($Z_E$) and magnetic ($Z_M$) surface impedances, achieved by varactor bias, and can be switched between full-transmit (T-mode, $\beta^t=1$) or full-reflect (R-mode, $\beta^r=1$) (Zhu et al., 18 Jan 2026).

Domain	Structure	Key Control Parameter(s)
Elastic (phononic)	Hexagonal lattice, perforated plate	Rotation angle $\theta$, clamping pattern
Electromagnetic	Passive ULA, parallel-LC resonators	Varactor-controlled $Z_E$, $Z_M$; $\theta_m$

2. Topological and Electromagnetic Mechanisms

For flexural waves, the system is governed by the Kirchhoff–Love plate equation,

$D \nabla^4 w(x, y) - \rho h \omega^2 w(x,y) = 0,$

with Bloch–Floquet periodicity. For $\theta=0$, Dirac cones at K and K′ arise from symmetry; rotation $\theta \neq 0$ gaps the dispersion, with an effective Hamiltonian

$$H_\text{eff}(\mathbf{q}) = v_D (\tau q_x \sigma_x + q_y \sigma_y) + m \sigma_z,$$

where $m \propto \sin 3\theta$, yielding domain-dependent valley Chern number $C_v = (\tau/2)\, \mathrm{sgn}(m)$. Interfaces between domains with opposite $m$ host topologically protected zero-line modes (ZLMs), which enable robust mode steering—even around sharp ($2\pi/3$) and gentle ($\pi/3$) bends—by appropriate alternation of the triangular sub-ensemble orientation at the interface (Tang et al., 2019).

For electromagnetic ISAC, the passive plate intercepts the backward lobe of an active array and imposes spatially varying phase shifts $\varphi_m^{r/t}$, such that the plate’s transmission/reflection coefficients $\theta_m$ reshape radiation patterns into arbitrary directions. The plate’s operation leverages array-to-array coupling via a spherical LOS channel and boundary enforcement based on Huygens’ principle. The joint action of active beamforming $\mathbf{W}$ and passive control $$\boldsymbol{\Theta} = \mathrm{diag}(\boldsymbol{\theta})$$ encodes the functional beampattern (Zhu et al., 18 Jan 2026).

3. Mathematical Models and Signal Processing Framework

In the elastic plate, numerical band structure calculations (e.g., COMSOL) determine spectral position of ZLMs and transmissive properties of various interface geometries. The reconfiguration logic follows discrete changes in $\theta$ and local clamping assignment, permitting waveguides of arbitrary topology to be assembled piecewise from six available $\Gamma$–K directions.

For ISAC, communication and sensing are jointly modeled:

• Communication:

$$\mathbf{y}_C = \frac{1}{2}\, \mathbf{H}_C \mathbf{W} \mathrm{diag}(\boldsymbol{\alpha}) \mathbf{s} + \mathbf{z}_C,$$

• Sensing (echo):

$$\mathbf{y}_S = \frac{1}{2} \mathbf{H}_S \boldsymbol{\Theta} \mathbf{G} \mathbf{W} \mathrm{diag}(\boldsymbol{\alpha}) \mathbf{s} + \mathbf{z}_S,$$

where $\mathbf{H}_{C/S}$ are respective channels, $\mathbf{G}$ is array-to-plate coupling, $\mathbf{s}$ is the transmit vector, $\boldsymbol{\alpha}$ is the scheduling vector, and $\mathbf{z}_{C/S}$ are noise (Zhu et al., 18 Jan 2026).

Performance metrics include per-user mutual information (MI), sensing MI in virtual uplink form, and sum-MI. Optimization involves maximizing a convex combination of sum communication and sensing MI over $$(\boldsymbol{\alpha},\mathbf{W},\boldsymbol{\Theta})$$, subject to power and hardware constraints.

4. Reconfigurable Steering and Implementation Methodologies

Arbitrary wavefront routing in elastic plates is attained by forming and reconnecting interfaces between $m > 0$ and $m < 0$ domains using magnetically clamped holes, following design rules:

1. All interface segments follow one of six allowed $\Gamma$–K ($\Gamma$–K′) orientations.
2. At each vertex (bend), select $\pi/3$ or $2\pi/3$ as needed.
3. Ensure operating frequency is within ZLM overlap bandwidth $[f_1, f_2]$.

This approach enables networks, beam splitters, delay loops, and planar circuits that remain passive and reconfigurable (Tang et al., 2019).

In ISAC, the optimization decomposes into:

• Passive coefficient update via Riemannian gradient ascent (over the unit-modulus manifold).
• Joint user scheduling and active beamforming via weighted min-MSE reformulation. Algorithmically, the iterative US-AGO procedure alternates between these steps, reaching convergence in $5\!-\!10$ outer iterations. Complexity depends on array size $M$, number of users $K$, and optimization method (Riemannian vs. SDR) (Zhu et al., 18 Jan 2026).

5. Experimental and Quantitative Performance

Elastic Plate (Topological Routing)

Performance metrics from experiment (Tang et al., 2019):

• Insertion loss (straight): < 1 dB (over $L \approx 500$ mm)
• Steering loss: < 0.5 dB (sharp bend); ~1 dB (gentle mode-conversion)
• Mode purity: > 95%
• ZLM overlap bandwidth: $\Delta f \approx 300$ Hz ($f_1 \approx 14.70$ kHz, $f_2 \approx 15.00$ kHz)

Electromagnetic ISAC

Simulation results (Zhu et al., 18 Jan 2026):

• At SNR 5 dB, sum-MI exceeds MMSE by 13.1%; at SNR 10 dB, outperforms MMSE-RP by 37.4%.
• Sum-NMSE decreases by up to 87.4% compared to random-phase baselines at 10 dB.
• Beampatterns with $M=64$ elements show narrow mainlobes and suppressed sidelobes; near-uniform MI across $0^\circ$–$360^\circ$.

Metric	Elastic Plate (Tang et al., 2019)	ISAC, Electromagnetic (Zhu et al., 18 Jan 2026)
Insertion/steering loss	<1 dB; <0.5–1 dB	—
Mode purity (ZLM)	>95%	—
360° MI coverage	—	Fluctuation bars small; near-uniform
NMSE/Sum-MI gain over baseline	—	13–46% MI gain; up to 87% NMSE red.

6. Scalability, Limitations, and Applications

Elastic omni-steering plates scale across frequencies via proportional adjustment of lattice constant $a$, hole radius $b$, and plate thickness $h$. The method applies broadly where passive, robust control of flexural waves is required and supports rapid in-situ reconfiguration (Tang et al., 2019).

In ISAC, the embedded passive plate provides full-space coverage and mitigates the synchrony, cost, and insertion loss issues associated with external or distributed RIS. Challenges include near-field coupling calibration, varactor non-idealities, finite reconfiguration speed, and bias-network complexity (Zhu et al., 18 Jan 2026).

Key applications are found in low-altitude UAV connectivity, trajectory-aware sensing, urban and environmental monitoring, and any deployment demanding concurrent 360° communication and sensing without additional major hardware.

7. Context and Research Directions

Omni-steering plate technology embodies two foundational innovations: the exploitation of discrete symmetry and local reconfiguration in topological phononic/elastic lattices, and the unification of active-passive electromagnetic manipulation in ISAC systems. By supplanting monolithic or static reflectors with programmable passive interfaces, the approach achieves both robustness (topological protection) and adaptivity (beam/path reconfiguration) unprecedented in traditional architectures.

Future developments may target overcoming nonlinearities, increasing spatial resolution at higher frequencies, and integrating dynamic control for real-time beam/path steering. The demonstrated information-theoretic optimality, reconfigurability, and efficient routing suggest broad utility in emerging intelligent infrastructure and adaptive physical-layer networks (Tang et al., 2019, Zhu et al., 18 Jan 2026).