A Phenomenological Background-Excitation Model of Light Propagation: The Impact-Triggered Flash Cascade (ITFC)

1. Introduction

Scope notice. This paper focuses on phenomenology: we formalize optical consequences in terms of \(\lambda,\, \tau_e,\, \kappa\) without committing to a full microscopic model. A concrete definition of the \(U\)-background degrees of freedom, their excitation/interaction rules, a background-state dynamics, and a mapping to standard field theory will be presented in subsequent papers.

Research question. This paper asks whether visible light can be modeled phenomenologically as a cascade of localized background excitations triggered by contact with a high-speed driver, and whether such a model can reproduce, within a single transfer-based framework, the effective propagation speed, refractive index, rectilinear propagation, refraction, scattering, and interference. The goal is not to present a complete microscopic quantum theory, but to test whether a compact phenomenological model based on background excitation transfer can account for a broad set of optical consequences in a unified way.

Historically, ether-like ideas failed in part because the background medium was assumed to possess classical mechanical properties that were difficult to reconcile with observed invariances. In the present work, the background is instead interpreted phenomenologically in quantum-like terms: not as a directly observable mechanical substance, but as an unobservable excitation-supporting substrate whose local transfer dynamics produce the effective propagation of light.

Baseline assumption (vacuum): If pair production is absent in vacuum, we adopt a massless‑U baseline. No indirect‑mass term is required; all ITFC observables are captured by the pair (\(\lambda\), \(\tau_e\)). Cosmology‑level extensions may model \(\Omega\)‑driven modulation as slow changes in \(\tau_e/\lambda\) rather than a rest mass.

Conventional optics treats light via Maxwell's equations and quantum electrodynamics [1, 5, 6, 9]. Here we examine whether a quantum-interpreted background-excitation picture can reproduce macroscopic optical laws while remaining compatible with relativistic constraints [2, 3, 6]. The central idea is that space is filled with unobservable background degrees of freedom (\(U\)); contact with a high-speed driver (\(P\)) excites a localized background state (\(U\)*), which transfers excitation to neighboring \(U\) and emits a flash. The sequence of such flashes along the P track constitutes what we observe as light.

In the present work, "rotation" is used phenomenologically to denote an internal or phase-like degree of freedom of the background, not literal rigid-body motion.

2. Background-Excitation Interpretation — Linkage Summary

Earlier work postulated that superposed rotation-like variables modulate particle mobility and the distribution of momentum. In the present paper, these variables are interpreted not as literal rigid-body motion, but as effective internal or phase-like degrees of freedom of an unobservable background. ITFC instantiates this picture by introducing a transient excitation variable and a dwell time, so that an impact imparts linear impulse to \(P\) and a short-lived localized excitation to \(U\).

3. Model Components

3.1 U-background degrees of freedom (unobservable background)

\(U\) is introduced phenomenologically as an unobservable background degree of freedom rather than as a directly observable mechanical constituent. Its excited state \(U*\) is interpreted as a localized background excitation, and the observable optical signal arises from the transfer and decay of such excitations.

• States. Stable \(U_0\) (non‑emissive) and excited \(U^*\) (flash‑emissive).

Transitions:

$$ U_0 \xrightarrow{impact} U^* \xrightarrow{decay} U_0 $$

U-particles are introduced phenomenologically as unobservable background degrees of freedom rather than directly observable mechanical constituents of a classical medium. Their excited state, denoted by \(U*\), is interpreted as a localized excitation of the background. In this interpretation, the observable optical signal does not arise from the direct visibility of \(P\) itself, but from the transfer and decay of such localized background excitations.

• Local dynamics. Let \(\omega\) denote a transient internal-state variable characterizing the local excited background configuration.

$$ \frac{d\omega}{dt} = -\frac{\omega}{\tau} + \sum_{j\in\aleph} \kappa\omega_j\, \Theta(r_c - \|r - r_j\|) $$

With decay time \(\tau\), coupling \(\kappa\), interaction radius \(r_c\). Emission intensity (phenomenology):

$$ I_U(t) = \eta \left|\frac{d\omega}{dt}\right|^p, \qquad p \geq 1,\ \eta > 0. $$

3.2 P‑particles (high‑speed drivers)

Properties. Effectively massless energy carriers, not directly observable. Average speed \(\upsilon_P\) with a lower bound \(c\) (i.e \(\upsilon_P \geq c\)). Here \(c\) is the critical U‑transfer speed (the light constant).

Trajectory. Ballistic flight over an effective free length \(\ell\) or transfer length \(\lambda\), with direction changes upon contact according to a scattering kernel \(p(\theta)\).

Contact rule. A single contact induces a localized excitation in \(U\), parameterized phenomenologically by \(\omega_0\) and may deflect \(P\) by an angle \(\theta\). Observable propagation is governed by the U‑cascade, not by \(P\) directly.

3.3 Definition of the Cascade (Light)

Along a \(P\) track, the set \(\{U*\}\) forms a time-ordered cascade of localized background excitations. The detector integrates the resulting flashes over a window \(\Delta t\); the integrated signal is the intensity, and its stream defines an effective ray. In this interpretation, the ray is not the trajectory of a directly visible carrier, but the macroscopic record of sequential excitation transfer through the background.

3.4 Observability (Afterimage Hypothesis)

\(P\) is inferred only via the spatio‑temporal pattern of the U‑cascade "afterimage". Identical \(P\) tracks may yield different apparent rays depending on \(\lambda, \tau_e\).

3.5 Chain Dynamics (Summary)

U‑excitation dynamics and emission follow §3.1; cascade formation is determined by \(\lambda, \tau_e, \kappa, r_c\).

3.6 Parameters

\(\lambda\) (effective transfer length), \(\tau_e\) (local dwell/lag), \(c\) (U‑transfer critical speed), \(p(\theta)\) (\(P\to U\) scattering kernel), etc.

3.7 Summary (Causal Direction)

Cause (\(P\)) moves with \(\nu_P \geq c\) → Medium (\(U\)) undergoes transfer at \(c\) → Observation (Light) = cascade of flashes.

4. Effective Propagation Speed and Refractive Index

Although the present paper is focused primarily on light propagation, the same phase-based formalism suggests a possible quantum-interpreted extension to nuclear binding and beta decay. In this extension, phase symmetry within a bound system is treated as an effective equilibrium condition, while beta decay is interpreted as a phase-relaxation process that releases accumulated imbalance through an emitted lepton channel and the accompanying background response.

The observable signal is the U‑cascade. For one step,

$$ t_{step} = \lambda/c + \tau_e, \qquad \nu_{eff} = \frac{\lambda}{\lambda/c+\tau_e} = \frac{1}{c^{-1}+\tau_e/\lambda}. $$

Define the effective refractive index

$$ n \equiv \frac{c}{\upsilon_{eff}} = 1 + \frac{c\tau_e}{\lambda}. $$

medium changes (density \(\rho_U\), structure, coupling) modify \(\lambda\) and \(\tau_e\), thereby changing \(n\) [2, 3, 7].

5. Optical Laws — Consequences

5.1 Rectilinearity

If scattering is forward‑peaked and \(\lambda \gg r_c\), deflection is small and straight‑ray propagation holds.

5.2 Reflection/Refraction (Variational Principle)

Light (the U‑cascade) follows paths minimizing total time \(T = \int \frac{ds}{\nu_{eff}(s)} = \int \frac{n(s)}{c}\,ds\). With

$$ n(s) = 1 + \frac{c\tau_e(s)}{\lambda(s)}, $$

Snell's law follows: \(n_A \sin\theta_A = n_B \sin\theta_B\). Ray bending aligns with \(\nabla n\), i.e., with gradients of \(\tau_e/\lambda\) [2, 7].

5.3 Scattering and Turbidity

Stronger transport scattering shortens \(\lambda\) (e.g., \(\lambda \approx 1/(\rho_U \sigma_{tr})\)) and typically increases \(\tau_e\) by multiple contacts; hence \(n = 1 + c\tau_e/\lambda\) rises and \(\upsilon_{eff}\) falls, transitioning to a transport regime characterized by an effective transport mean free path \(\ell^*\) and diffusion coefficient \(D \sim \upsilon_{eff}\ell^*/3\). Measurables: increased turbidity and delayed pulse tails [3, 12].

5.4 Diffraction/Interference as Time Synchronization

Two cascades arriving within a threshold sync time \(\tau_c\) produce fringes within the detector's integration window \(\Delta t_{int}\). The coherence length is

$$ L_c = \upsilon_{eff}\, \tau_c = \frac{c\,\tau_c}{1+c\,\tau_e/\lambda} $$

Vacuum‑like conditions (small \(\tau_e/\lambda\)) yield \(\upsilon_{eff} \to c\) and longer \(L_c\), enhancing visibility; media with larger \(\tau_e/\lambda\) reduce \(L_c\) and modulate fringe patterns [2, 7].

6. Relativistic Constraints and Local Invariance

Treat \(c\) as the invariant U‑transfer critical speed. Although \(\upsilon_P \geq c\) may hold, observable information propagates with \(\upsilon_{eff} \leq c\). Michelson–Morley‑type nulls are recovered by covariant scaling of U‑parameters (e.g., an invariant \(\tau_e/\lambda\) in a local inertial frame), which removes first‑order anisotropy [4, 6].

A weak dependence of the transfer parameters on large-scale background structure may be considered as a possible extension of the present formalism. Such an extension is not essential to the core ITFC model developed here, which is restricted to local propagation phenomenology.

7. Numerical Simulation Frame

Microscopic (agent‑based): Monte‑Carlo P‑jumps plus U‑lattice reaction‑transfer \(U_0 \leftrightarrow U^*\).

Continuum (PDE):

$$ \partial_t u^*(r,t) = D\nabla^2 u^* - \frac{u^*}{\tau} + S_P(r,t), $$

where \(S_P\) is a source tied to P tracks [3, 12].

Table 1. Model Parameters (Baseline & Sweep)

Table 2. Numerical Settings (Recommended)

Note. Start with dimensionless units and map to physical units in later experimental design. Baseline_A is vacuum‑like; Baseline_B represents strong boundary/scattering conditions [3, 12].

Table (Numerical Settings — Recommended)

Output metrics. \(\Delta T\): time delay across length \(L\) (\(\int ds/\nu_{eff}\)); \(V\): fringe visibility \((I_{max} - I_{min})/(I_{max} + I_{min})\); \(n_{eff}\): path‑averaged refractive index \((1/L)\int n(s)\,ds\).

8. Experimental Predictions and Discriminators

1. Pressure/density‑dependent index. Control residual gas (affecting \(\lambda, \tau_e\)) and measure \(\Delta T \sim L\Delta(\tau_e/\lambda)\) [2, 3, 12].

2. Boundary dynamics. Design interfaces (metasurfaces) to tailor \(p(\theta)\), inducing non‑integer refraction and polarization‑dependent \(n\) [3, 11].

3. Time‑synchronization interferometry. Scan pulse delay and locate the threshold \(\tau_c\) for fringe onset [2, 7].

4. UHV gas‑injection step test. Micro‑step \(\lambda, \tau_e\) by controlled injections and read out femto–picosecond timing shifts via interferometry: \(\Delta n \approx c\Delta(\tau_e/\lambda)\) [2, 3, 4].

5. Kernel engineering at boundaries. Actively vary \(p(\theta)\) and compare with ITFC predictions (cascade‑transfer delays) [3, 11].

6. Annual/azimuthal modulation interferometry. Possible large-scale background modulation may be explored through annual or azimuthal interferometric variation, but such effects are not essential to the present local phenomenology of excitation transfer.

9. Mapping to Topological Terminology (Optional)

U‑background corresponds to a near‑zero phase‑feedback exterior of a quasi‑isolated system; contact transitions are phenomenologically analogous to entropy‑direction inversions. This gives a dictionary between excitation-based and phase-based language without altering predictions.

9.1 Phase Symmetry in Nuclear Binding and Phase Relaxation (Beta Decay)

We interpret nuclear binding as a tendency to maintain phase symmetry (or equilibrium) within a bound system.

Let \(\Phi_i\) and \(\Phi_j\) denote the phases of nucleons i and j, and introduce a short-range envelope \(U(R_{ij}) \sim -J_0\exp(-R_{ij}/\lambda_N)\). A phenomenological binding energy can be written as:

$$ E_{bond} = -\sum_{<i,j>} J_{ij}(R_{ij})\cos(\phi_i - s_{ij}\phi_j - \pi p_{ij}) + \sum_{<i,j>} U(R_{ij}) $$

where \(s_{ij} \in \{\pm 1\}\) represents the coupling sign (in-phase or anti-phase), and \(p_{ij} \in \{0,1\}\) encodes a π-junction (e.g., neutron-mediated anti-phase coupling). Energy minimization enforces local phase equilibrium, naturally producing short-range strong attraction and saturation behavior [8, 9, 10].

In this framework, beta decay is interpreted as a phase relaxation (phase-slip) process that occurs when accumulated phase imbalance exceeds a critical barrier. Introducing an isospin angle α with an effective potential:

$$ U_{iso}(\alpha) = -J_{pn} \cos\alpha + \frac{K}{2}\alpha^2 $$

a phase-slip transition \(\alpha \to \alpha \pm \pi\) occurs when the phase energy exceeds a threshold \(E_b\).

The released energy is partitioned as:

$$ Q = \Delta E_{phase} - E_b = T_e + T_\nu + E_{U-cascade} \geq 0 $$

where \(T_e\) and \(T_\nu\) denote the kinetic energies of the emitted electron and (anti)neutrino, and the remaining energy is carried by the U-cascade (flash chain) [8, 9, 10].

Charge conservation is ensured via the label-charge continuity relation (Appendix D.5′):

$$ \partial_t\rho_q + \nabla \cdot J_q = S_{slip} $$

which reduces to standard conservation (\(S_{slip} \to 0\)) in the rapid recombination limit. Thus, beta decay is interpreted as a dynamical process restoring phase equilibrium within the bound system.

10. Limitations and Open Problems

• Establish a precise effective field theory mapping to Maxwell/QED [1, 5, 6, 9].
• Quantitatively confront high‑precision null experiments [4].
• Integrate a nonlinear detector model for biological perception.

11. Conclusion

ITFC redefines light as a cascade of localized \(U*\) excitations and their emitted flashes, and interprets the light constant as a U‑transfer critical speed. Thus, while a driver P may satisfy \(\upsilon_P \geq c\), information transfer remains bounded by \(\upsilon_{eff} \leq c\). Medium sensitivity is quantified by \(n = 1 + c\tau_e/\lambda\). The frame unifies rectilinearity, reflection/refraction, scattering/turbidity, and diffraction/interference through the single spatial field \(\tau_e/\lambda\). We propose UHV step tests and boundary‑kernel engineering as near‑term discriminators.

In addition to its optical interpretation, the ITFC framework may be read as a phenomenological background-excitation model. In this view, observable propagation is determined by the transfer of localized excitations through unobservable background degrees of freedom. Possible large-scale background modulation may be explored in future work, but it is not essential to the present formulation, which is primarily concerned with the local phenomenology of light propagation.

Future work. (1) Construct a microscopic background-excitation model and an EFT correspondence; (2) infer the effective parameters from metasurface/nanointerferometry datasets; (3) perform precision comparisons of the main observables; (4) explore possible large-scale background-modulation effects through refractive-index mapping.

Appendix A. Covariant Parameter‑Scaling Hypothesis (Sketch)

In local inertial frames, let \(\lambda \propto f(\Phi)\) and \(\tau_e \propto f(\Phi)\) scale homogeneously with a local scalar \(\Phi\), keeping. Then \(\upsilon_{eff}^{-1} = c^{-1} + const\) is locally invariant and first‑order anisotropy cancels.

Appendix B. Notation

\(\rho_U\): density of background degrees of freedom; \(\sigma\): (transport) cross section; \(\ell\): transport mean free path; \(\lambda\): effective transfer length; \(\tau, \tau_e\): decay/dwell constants; \(\kappa\): transfer coupling; \(r_c\): transfer radius; \(c\): U‑transfer critical speed; \(n\): effective refractive index; \(u^*\): excited U concentration; \(\upsilon_P\): P speed; \(\upsilon_{eff}\): effective cascade speed.

Appendix C. Diagrams (v1)

Display note. If Mermaid is unavailable, use the ASCII alternates. Variables appear in plain text (tau_e, lambda, r_c).

[Fig. 1] U‑cascade propagation (P track and excitations)

flowchart LR
 P0((P)) --> U1((U*))
 U1 --> U2((U*))
 U2 --> U3((U*))
 P0 -. dotted path .-> P1((P))
 P1 --> U4((U*))
 U4 --> U5((U*))
 classDef glow fill:#fffbd6,stroke:#555;
 class U1,U2,U3,U4,U5 glow;

ASCII alt:

P path: . . . . . . . . .
 o* -> o* -> o* (U* : excited U; flash)
 \
 -> o* -> o*
r_c : local transfer radius
lambda : effective transfer length between successive U*

[Fig. 2] Trajectory change at a boundary (refraction/reflection)

flowchart LR
 subgraph A[region A: n_A = 1 + c*tau_eA/lambda_A]
 IA[/incident ray θ_A/]
 end
 subgraph B[region B: n_B = 1 + c*tau_eB/lambda_B]
 RB[/refracted ray θ_B/]
 end
 IA -->|boundary| RB

ASCII alt:

region A (n_A) | region B (n_B)
 \
 \ θ_A | / θ_B
 \____boundary|_/________
Snell: n_A * sin(θ_A) = n_B * sin(θ_B)
with n = 1 + c * tau_e / lambda

[Fig. 3] Time‑synchronization interference (coherence length L_c)

P-chain #1: |* * * * *|
P-chain #2: |* * * * *|
 <---- Δt ---->
Detector window Δt_int: [===========]
Sync time τ_c: threshold for fringe formation
L_c = v_eff * τ_c = c * τ_c / (1 + c * tau_e / lambda)

Appendix D. Effective Action (Lagrangian) — Sketch

This appendix presents a refined effective action sketch for ITFC while preserving the phenomenological parameters \(\lambda, \tau_e, \tau_{swap}, \kappa\). The purpose is twofold: (i) encode memory effects and finite transfer range directly in the action, and (ii) to recover, in the long‑wavelength and low‑frequency limit, the ITFC propagation law

$$ t_{step} = \frac{\lambda}{c} + \tau_{eff}, \qquad \upsilon_{eff} = \frac{1}{c^{-1} + \tau_{eff}/\lambda}, \qquad n = 1 + \frac{c\tau_{eff}}{\lambda} $$

Where

$$ \tau_{eff} = \begin{cases} \tau_e, & dilute\ background, \\ \tau_{swap}, & dense\ swap-dominated\ background. \end{cases} $$

The effective action introduced in Appendix D is not presented as a complete microscopic quantum theory, but as a coarse-grained phenomenological model compatible with a quantum-interpreted background picture.

D.1 Fields and External Source

Let \(\psi(x) \equiv \psi(t, r)\) denote an effective scalar field representing the intensity/state of U-transfer, i.e. the observable carrier of the U-cascade. The P-track enters only through an external source

$$ J_p(x) = g \int d\tau\, \delta^{(4)}(x - X(\tau)), $$

where \(X(\tau)\) is the worldline of the P-particle. Thus, P itself is not directly observable; only the induced field \(\psi\) is.

D.2 Effective Action (Sketch)

The refined effective action is written as

$$ S_{eff} = \int d^4x \left[\frac{1}{2}(\partial_t\psi)^2 - \frac{c^2}{2}|\nabla\psi|^2 - V(\psi) + J_P\psi\right] + \frac{1}{2}\int d^4x\,d^4x'\,\psi(x)K(x - x')\psi(x'). $$

Here \(K(x - x')\) is a causal nonlocal kernel satisfying

$$ K(t - t' < 0, r - r') = 0, $$

so that temporal ordering and observable causality are preserved.

A separable illustrative form is

$$ K(x - x') = K_t(t - t')\, K_r(r - r'), $$

with

$$ K_t(\Delta t) = \frac{\alpha}{\tau_{eff}} \exp\left(-\frac{\Delta t}{\tau_{eff}}\right)\Theta(\Delta t), $$

$$ K_r(\Delta r) = -\beta\, \frac{c}{\lambda}\, \frac{\exp(-|\Delta r|/\lambda)}{N_\lambda}, $$

where \(N_\lambda\) is a normalization factor and \(\alpha, \beta\) are matching coefficients.

If a cosmic reversible rotation field \(\Omega = \Omega(R)\) is included, the kernel may be parameterized as

$$ K(x - x';\Omega) = K_t(t - t';\Omega)K_r(r - r';\Omega), $$

with Fourier transform

$$ \tilde{K}(\omega, k;\Omega) \simeq A(\Omega)(i\omega) + B(\Omega)k^2 + \cdots. $$

The matching conditions are then imposed as

$$ \left(\frac{\partial\tilde{K}}{\partial(i\omega)}\right)_{\omega=0,k=0} = \frac{1}{\tau_{swap}(\Omega)}, \qquad \left(\frac{\partial\tilde{K}}{\partial k^2}\right)_{\omega=0,k=0} = -\frac{c}{\lambda(\Omega)} $$

These ensure the long-wavelength relations

$$ \upsilon_{eff}(\Omega) = \frac{1}{c^{-1} + \tau_{swap}(\Omega)/\lambda(\Omega)}, \qquad n(\Omega) = 1 + \frac{c\tau_{swap}(\Omega)}{\lambda(\Omega)} $$

D.3 Equations of Motion and Dispersion (Linearized)

Near a local steady state, take

$$ V(\psi) \simeq \frac{m^2}{2}\psi^2. $$

Then the Euler–Lagrange equation becomes

$$ \partial_t^2\psi - c^2\nabla^2\psi + m^2\psi + \int d^4x'\, K(x - x')\psi(x') = -J_P(x). $$

For a plane-wave ansatz

$$ \psi(x) \sim e^{i(k\cdot r - \omega t)}, $$

one obtains the dispersion relation

$$ -\omega^2 + c^2k^2 + m^2 + \tilde{K}(\omega, k) = 0. $$

In the long‑wavelength/low‑frequency regime,

$$ \tilde{K}(\omega, k) \approx A(i\omega) + Bk^2 + \cdots, $$

And the coefficients \(A\) and \(B\) are fixed by the ITFC matching rules below.

D.4 Long‑Wavelength Matching (Design Rules)

The refined design rules are:

$$ (M1)\quad t_{step} = \frac{\lambda}{c} + \tau_{eff} \implies \upsilon_{front} = \frac{1}{c^{-1} + \tau_{eff}/\lambda} \equiv \upsilon_{eff}, $$

$$ (M2)\quad n = \frac{c}{\upsilon_{eff}} = 1 + \frac{c\,\tau_{eff}}{\lambda}, $$

$$ (C1)\quad \left(\frac{\partial\tilde{K}}{\partial(i\omega)}\right)_0 = \frac{1}{\tau_{eff}}, $$

$$ (C2)\quad \left(\frac{\partial\tilde{K}}{\partial k^2}\right)_0 = -\frac{c}{\lambda}. $$

Choosing \(\alpha, \beta\), and \(N_\lambda\) so that \((C1) - (C2)\) hold ensures that the front velocity of the causal U-cascade response reproduces the ITFC law.

Remark. In a generic nonlocal medium, phase velocity and group velocity need not coincide with the signal front velocity. In ITFC, the physically relevant quantity is the propagation of the U-cascade front.

D.5 Minimal Potential and Stability

A stable minimal potential may be chosen as

$$ V(\psi) = \frac{m^2}{2}\psi^2 + \frac{\lambda_4}{4!}\psi^4 + \cdots, \qquad m^2 \geq 0,\ \lambda_4 \geq 0. $$

If one prefers a phase interpretation, a periodic form such as

$$ V(\phi) = \Lambda_\phi(1 - \cos\phi) $$

may be used instead.

D.5′ Label-Charge Continuity (Noether Sketch)

Assume a global \(U(1)\)-like symmetry associated with a label charge \(q\). Then the effective current obeys

$$ \partial_t\rho_q + \nabla \cdot J_q = 0. $$

D.6 Summary

The refined effective action preserves:

$$ (1)\ casuality, \quad (2)\ finite\ memory/range\ (\tau_{eff}, \lambda), \quad (3)\ invariant\ c. $$

in the long-wave-length limit. In laboratory vacuum without pair production, any effective-mass construct may be set to zero, and the pair (\(\lambda, \tau_e\)) alone suffices. A full microscopic derivation and EFT correspondence remain future work.

References

[1] Maxwell, J. C., A Treatise on Electricity and Magnetism, Vols. I–II, Clarendon Press, Oxford (1873).

[2] Born, M. and Wolf, E., Principles of Optics: Electromagnetic Theory of Propagation, Interference and Diffraction of Light, 7th expanded ed., Cambridge University Press (1999).

[3] Landau, L. D., Lifshitz, E. M., and Pitaevskii, L. P., Electrodynamics of Continuous Media, 2nd ed., Butterworth-Heinemann / Elsevier, Vol. 8 of Course of Theoretical Physics (1984).

[4] Michelson, A. A. and Morley, E. W., "On the Relative Motion of the Earth and the Luminiferous Ether," American Journal of Science 34(203), 333–345 (1887).

[5] Feynman, R. P., QED: The Strange Theory of Light and Matter, Princeton University Press (1985).

[6] Jackson, J. D., Classical Electrodynamics, 3rd ed., Wiley (1998).

[7] Hecht, E., Optics, 5th ed., Pearson (2016).

[8] Sakurai, J. J. and Napolitano, J., Modern Quantum Mechanics, 2nd ed., Cambridge University Press (2017).

[9] Peskin, M. E. and Schroeder, D. V., An Introduction to Quantum Field Theory, CRC Press / Westview Press (1995).

[10] Weinberg, S., The Quantum Theory of Fields, Vol. I, Cambridge University Press (1995).

[11] Joannopoulos, J. D., Johnson, S. G., Winn, J. N., and Meade, R. D., Photonic Crystals: Molding the Flow of Light, 2nd ed., Princeton University Press (2008).

[12] Ishimaru, A., Wave Propagation and Scattering in Random Media, Vols. I–II, Academic Press (1978).