2D Signal-Rule: Methods & Applications

• 2D Signal-Rule is a family of rule systems that govern signal detection, propagation, and control in two-dimensional domains across various applications.
• It underpins methodologies ranging from analytically determined thresholds in compressive sensing to optical selection rules in monolayer materials and computational logic in cellular automata.
• These rule systems enable precise signal reconstruction, universality in computational models, and controlled molecular behavior in quorum sensing and tile assembly.

The expression “2D signal-rule” is not used in the cited literature as a single canonical construct. Instead, it denotes several distinct rule systems for detection, propagation, transformation, and control of signals in two-dimensional settings. In the available arXiv record, these include an analytically determined threshold for sparse 2D-DFT reconstruction in compressive sensing, a valley-exciton optical selection rule in monolayer WS$_2$, an isotropic cellular-automaton rule supporting collision-based logic, a two-dimensional Bedrosian identity for partial Hilbert transforms, a stochastic-geometry rule for molecular signaling in quorum sensing, and signal-passing rules in two-dimensional tile assembly (Stankovic et al., 2014, Xiao et al., 2015, Soto et al., 2021, Zhang, 2012, Fang et al., 2020, Fochtman et al., 2013). This suggests that the term is best understood as a family resemblance across fields rather than a unified formal theory.

1. Threshold separation in sparse 2D compressive sensing

Stanković and Orović study sparse two-dimensional signals $s(x,y)$ of size $N_x\times N_y$ with sparsity in the 2D DFT domain, under random undersampling. Their central rule is an analytically determined threshold $T$ that separates true spectral components from the “spectral noise” induced by missing samples. For a signal containing $K\ll N_xN_y$ complex exponential components with amplitudes $A_i$, the variance at DFT bins not coincident with a true frequency is

$$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$

If additional external white noise of variance $\sigma_e^2$ is present, the variance is replaced by $\sigma^2+\sigma_e^2$. Under a Central Limit Theorem argument, the real and imaginary parts at pure-noise bins are Gaussian, so the magnitude is Rayleigh-distributed. The resulting max-noise threshold is

$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$

with the practical approximation

$s(x,y)$0

The associated one-shot reconstruction procedure computes the partial-DFT observation $s(x,y)$1, detects support by $s(x,y)$2, and then solves a least-squares problem on that support,

$s(x,y)$3

When one-shot detection misses weak components, the paper gives an OMP-like iteration that repeatedly estimates a residual spectrum, recomputes the variance and threshold, augments the support, and refits the coefficients. Convergence is described as typically occurring in 2–4 steps because strong components are peeled off first, lowering the noise floor for later passes.

The significance of this rule is operational rather than purely statistical: it converts missing-sample dispersion into a deterministic reconstruction workflow in the 2D DFT domain. In the ISAR example, the method reconstructs simulated data even when less than 10% of samples are available; one reported case uses $s(x,y)$4, $s(x,y)$5, only 9% random samples, and external noise at SNR $s(x,y)$6 dB, with one-shot detection of all 12 peaks above $s(x,y)$7 and coefficient-domain MSE $s(x,y)$8. The paper also states limitations: estimation of $s(x,y)$9 requires knowledge or estimation of $N_x\times N_y$0, the choice of $N_x\times N_y$1 trades missed detections against false alarms, the Rayleigh approximation assumes random missing positions, and very small $N_x\times N_y$2 can invalidate the Gaussian approximation (Stankovic et al., 2014).

2. Valley-exciton locking as an optical selection rule in 2D materials

In monolayer WS$N_x\times N_y$3, the optical “signal-rule” is a selection rule imposed jointly by valley angular momentum (VAM) and excitonic angular momentum (EAM). Because inversion symmetry is broken and the Berry curvature is strong at the $N_x\times N_y$4 points, Bloch electrons carry VAM $N_x\times N_y$5, where $N_x\times N_y$6 at $N_x\times N_y$7 and $N_x\times N_y$8 at $N_x\times N_y$9. The exciton carries EAM $T$0, through the azimuthal phase winding of its envelope wavefunction $T$1. The governing conservation law states that the total optical angular momentum exchange must conserve photon spin angular momentum, the change in VAM, the change in EAM, and any multiple of three quanta absorbed by the lattice under $T$2 rotational symmetry:

$T$3

For two-photon absorption and single-photon emission, this becomes

$T$4

for SHG, and

$T$5

for the resonant TPL selection rule.

The paper’s examples are explicit. For 1s SHG at the $T$6 valley, $T$7 and $T$8, so the rule requires $T$9 pumping and $K\ll N_xN_y$0 SH emission; the paper summarizes this as “$K\ll N_xN_y$1 fundamental $K\ll N_xN_y$2 $K\ll N_xN_y$3 SHG.” For 2p$K\ll N_xN_y$4 TPL at the $K\ll N_xN_y$5 valley, $K\ll N_xN_y$6 and $K\ll N_xN_y$7, so $K\ll N_xN_y$8 two-photon absorption selectively creates $K\ll N_xN_y$9-valley 2p$A_i$0 excitons, and the relaxed 1s exciton emits $A_i$1 luminescence.

The experimental section reports resonant SHG at the 1s exciton near $A_i$2 eV, with a pump scan locating the resonance at $A_i$3 eV, an SHG intensity enhancement of about $A_i$4 at $A_i$5 K, and helicity $A_i$6 under $A_i$7 pumping. The TPL scan shows a broad 2p peak at $A_i$8 eV pump, linewidth $A_i$9 meV, and $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$0. Time-resolved fits yield a 2p$$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$11s relaxation time $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$2 fs, a 1s recombination time $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$3 ps, intervalley scattering during relaxation $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$4 ps, and during recombination $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$5 ps.

A common misconception would be to reduce this rule to ordinary circular-polarization selection alone. The paper instead makes the dependence on coupled valley and exciton angular momenta explicit, with lattice-mediated $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$6 exchange under $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$7 symmetry as an integral part of the rule. The stated implications include 2D valley-polarized LEDs, ultrafast optical switches, and coherent quantum control using $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$8 exciton manifolds (Xiao et al., 2015).

3. The Ameyalli-rule as a 2D cellular-automaton signal rule

In the cellular-automaton literature, the Ameyalli-rule is a two-dimensional “signal-rule” defined on the binary Moore neighborhood. Each cell $$\sigma^2 = \frac{M_xM_y\bigl(N_xN_y - M_xM_y\bigr)}{N_xN_y -1} \sum_{i=1}^K A_i^2.$$9 has eight neighbors plus itself, with synchronous deterministic update

$\sigma_e^2$0

Rather than using survival/birth logic, the rule is isotropic: the 512 neighborhood patterns are partitioned under the dihedral group $\sigma_e^2$1 into exactly 102 iso-groups. The rule is therefore represented by a 102-bit iso-rule $\sigma_e^2$2, where $\sigma_e^2$3 means that any neighborhood in iso-group $\sigma_e^2$4 maps to state 1. The given hexadecimal encoding is

$\sigma_e^2$5

Its signal carriers are gliders. The Ameyalli glider is an orthogonal 2-phase object whose shape repeats every 2 time steps while shifting by one cell in its direction of motion, giving speed $\sigma_e^2$6. The glider-gun is a period-22 oscillator that emits one glider every 22 steps in each of the four cardinal directions, with 11-cell spacing along each ray. Confining the core by two stationary eaters of type B at $\sigma_e^2$7 and two 2-phase eaters of type A at $\sigma_e^2$8 yields a periodic attractor of length 22.

Logical universality is obtained by encoding bits as streams of gliders and gaps separated by 11 cells. The NOT gate works by annihilation between an input stream and a constant gun stream: input $\sigma_e^2$9 destroys the gun glider and yields output $\sigma^2+\sigma_e^2$0, while input $\sigma^2+\sigma_e^2$1 leaves the gun glider intact and yields output $\sigma^2+\sigma_e^2$2. The paper gives the example input “10001” $\sigma^2+\sigma_e^2$3 output “01110.” The AND gate uses two eastward streams colliding with a northward constant stream and simultaneously produces $\sigma^2+\sigma_e^2$4 on one ray and $\sigma^2+\sigma_e^2$5 (NOR) on the other. The OR gate inverts the NOR output through a second gun. Because $\sigma^2+\sigma_e^2$6 is functionally complete, these collision gadgets establish logical universality.

The rule is further analyzed by the input-frequency histogram (IFH), which counts how often each iso-group is queried by the local update rule over a space-time evolution. Active iso-groups appear as spikes in $\sigma^2+\sigma_e^2$7, while absent columns correspond to neutral inputs. The details state that in an “intensive” pattern using all logical components, 53 of the 102 iso-groups are neutral, and setting all neutral inputs to 0 leaves an idealized iso-rule with only 12 ones that still supports the glider-gun and gates. This makes the Ameyalli-rule notable as a two-dimensional signal-rule in which “signals” are literal mobile localized structures and the rule table can be compressed toward the subset required for computation (Soto et al., 2021).

4. The 2D Bedrosian identity and analytic-signal constructions

In harmonic analysis, the relevant “2D signal-rule” is the Bedrosian identity for partial Hilbert transforms on $\sigma^2+\sigma_e^2$8. Zhang defines, for $\sigma^2+\sigma_e^2$9 and $$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$0, the partial Hilbert transform $$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$1 through both a singular-integral and a Fourier-multiplier representation. The multiplier form is

$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$2

The central identity asks for conditions under which

$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$3

holds almost everywhere. A classical sufficient condition is coordinatewise spectral separation: if

$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$4

then the identity holds for each $$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$5. The general characterization is given by the convolution condition

$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$6

for almost every $$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$7. Under an extremal support hypothesis—$$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$8 equal to the full rectangle $$T = \sigma \sqrt{-2\ln\!\Bigl(1 - P_{\text{all}}^{1/(N_xN_y-K)}\Bigr)},$$9 with all four corners in the support, and $s(x,y)$00 compactly supported—the identity for both $s(x,y)$01 forces

$s(x,y)$02

The same rule underlies the construction of multidimensional analytic signals. With

$s(x,y)$03

the analytic signal is $s(x,y)$04, and its Fourier transform satisfies

$s(x,y)$05

which vanishes outside the first quadrant $s(x,y)$06. The paper then constructs tensor-product basis functions $s(x,y)$07, justified by the separable Bedrosian identities.

This usage differs sharply from rule-based detection or logic. Here the “rule” is an algebraic identity about how a multiplier operator interacts with products under support constraints. The literature therefore suggests that “2D signal-rule” can refer not only to discrete transition systems but also to functional-analytic commutation laws governing multidimensional signal representations (Zhang, 2012).

5. Molecular signaling rules in 2D quorum sensing

In quorum-sensing analysis, the 2D signal-rule is a stochastic-geometry description of how continuously emitted molecules propagate, accumulate, and trigger cooperative behavior in a two-dimensional bacterial population. For a point source at the origin in an unbounded 2D medium with diffusion coefficient $s(x,y)$08 and degradation rate $s(x,y)$09, the reaction-diffusion equation is

$s(x,y)$10

The corresponding impulse response is

$s(x,y)$11

For continuous emission at average rate $s(x,y)$12, the steady-state response is

$s(x,y)$13

where $s(x,y)$14 is the modified Bessel function. If each bacterium is a passive circular receiver of radius $s(x,y)$15, the expected steady-state molecule count under the uniform-concentration assumption is approximated by

$s(x,y)$16

With bacteria distributed according to a Poisson point process of density $s(x,y)$17, a bacterium at $s(x,y)$18 receives mean aggregate count

$s(x,y)$19

Assuming Poisson arrival counts, the probability of cooperation at $s(x,y)$20, given threshold $s(x,y)$21, is

$s(x,y)$22

For the total number of cooperators $s(x,y)$23, the moment generating function is

$s(x,y)$24

and the stated formulas give

$s(x,y)$25

The details further state that $s(x,y)$26 is exactly Poisson$s(x,y)$27 under the independence implied by thinning of the PPP, while the abstract reports that Poisson and Gaussian approximations were compared and that the Poisson distribution provided the best overall approximation.

This rule is directly tied to control variables. Increasing $s(x,y)$28 or $s(x,y)$29 lowers $s(x,y)$30, and thus decreases $s(x,y)$31 and $s(x,y)$32; increasing $s(x,y)$33 boosts all responses linearly; increasing $s(x,y)$34 or reducing community radius $s(x,y)$35 raises local aggregate signals; and increasing the threshold $s(x,y)$36 suppresses cooperation. The paper identifies biofilm-prevention strategies such as raising $s(x,y)$37, reducing $s(x,y)$38, or increasing $s(x,y)$39, and conversely suggests increasing $s(x,y)$40, decreasing $s(x,y)$41, or confining cells when quorum sensing is desired (Fang et al., 2020).

6. Signal-passing rules in 2D tile assembly

In algorithmic self-assembly, the relevant object is the 2D Signal-passing Tile Assembly Model, $s(x,y)$42. Fochtman et al. define a tile type as

$s(x,y)$43

where $s(x,y)$44 assigns glues and their latent/on states to the four directions $s(x,y)$45, and the signal-transition function

$s(x,y)$46

specifies which latent glues are turned on when a bound glue fires. An $s(x,y)$47 system is a triple $s(x,y)$48, with tile set $s(x,y)$49, initial state $s(x,y)$50, and temperature $s(x,y)$51. Two assemblies bind when the total strength of adjacent matching on-glues is at least $s(x,y)$52, and the corresponding $s(x,y)$53-functions are invoked. Crucially, pending glue activations occur asynchronously, with no assumed ordering between signal firings and tile attachments.

The paper’s signal-rule is therefore a local asynchronous causality rule: a binding event on one side of a tile can turn on latent glues on other sides of the same tile at a later time. The toy example given is a tile $s(x,y)$54 with $s(x,y)$55, so that binding on the west side eventually activates glue $s(x,y)$56 on the east side.

The paper then studies signal-complexity reduction. A tile set is $s(x,y)$57-simplified if every tile has at most $s(x,y)$58 total signals, either fan-out or mutual-activation is globally forbidden, and every tile has fan-in at most 2. Theorem 1 states that every $s(x,y)$59 system at $s(x,y)$60 can be simulated by a 2-simplified $s(x,y)$61 system at scale factor $s(x,y)$62 and tile-complexity $s(x,y)$63, eliminating either fan-out or mutual-activation. Theorem 2 states that for $s(x,y)$64, every $s(x,y)$65 system can be simulated by a 1-simplified system with the same asymptotic scale factor and tile-complexity, and that both fan-out and mutual-activation can be completely removed.

The universality result extends this 2D signal-passing model into passive 3D self-assembly. Theorem 3 states that there is a single 3D tile set $s(x,y)$66, using only two $s(x,y)$67-layers, that is intrinsically universal at temperature 2 for the class of all 2D $s(x,y)$68 systems with $s(x,y)$69. For any such system $s(x,y)$70, an input supertile $s(x,y)$71 configures $s(x,y)$72 to simulate $s(x,y)$73 at scale $s(x,y)$74. The broader consequence is that static 3D 2HAM tiles can simulate asynchronous glue-activation behavior of 2D $s(x,y)$75 systems, even though the simulator itself uses only passive components (Fochtman et al., 2013).