Design of a planar gradient-index dielectric lens for arbitrary point-to-point beam focusing using critical-angle theory Authors: Mohammad Mahdi Taskhiri, Mohammad Moradi, Saeed Fakhte Published: 2026 ABSTRACT AbstractThis paper presents a novel analytical design method for a planar gradient-index (GRIN) dielectric lens based on the critical-angle theorem, capable of focusing electromagnetic waves between two arbitrarily located points. Unlike conventional curved lenses, the proposed flat lens structure enables flexible beam pointing and precise wavefront control without mechanical movement. The refractive index profile is derived analytically to ensure phase matching and minimal internal reflection. The design is validated through full-wave numerical simulations in COMSOL Multiphysics, demonstrating effective beam focusing, and improved impedance matching. The proposed lens, validated here in the 4–7 THz band, exhibits significant potential for frequency-scalable applications ranging from Terahertz imaging to 5G/6G communications, satellite links, and radar systems where versatile beam alignment is required.Keywords: Ray trajectories in inhomogeneous media, Lens, Beam concentratorSubject terms: Engineering, Optics and photonics, Physics Introduction The rapid advancement of satellite communications1, automotive radar2, fifth-generation (5G) and beyond networks3, and medical hyperthermia systems4 has driven an increasing demand for high-performance, beam-steerable antenna systems. Among various antenna technologies, dielectric lens antennas have attracted significant attention due to their ability to provide wide bandwidth, high gain, and low propagation losses at millimeter-wave frequencies5,6. By concentrating electromagnetic waves in a desired direction, dielectric lenses effectively enhance signal strength while minimizing spatial spreading and sidelobe levels7. Dielectric lenses are generally categorized into homogeneous and inhomogeneous (gradient-index, GRIN) types. Homogeneous lenses offer simplicity and uniform wavefront distribution, making them suitable for many optical systems8. However, they suffer from inherent limitations such as low gain, substantial thickness, and limited design flexibility due to their uniform refractive index distribution9. In contrast, GRIN lenses allow the refractive index to vary continuously or discretely within the lens volume, enabling precise control over the ray paths and wavefront phase. This capability facilitates reduced internal reflection, lower sidelobes, and improved focusing efficiency10. Despite these advantages, most reported GRIN lens designs are either curved (e.g., hemispherical or Luneburg-type) or restricted to fixed focal points11,12. Curved lenses introduce fabrication complexity and integration challenges in planar systems, while fixed-focus designs lack the adaptability required for dynamic communication links where transmitter and receiver positions may vary arbitrarily. In many practical scenarios–such as conformal antennas on vehicles, low-profile satellite terminals, or integrated photonic circuits–a flat lens architecture is essential to meet spatial constraints and ensure seamless integration with planar platforms. Consequently, there is a pressing need for a planar dielectric lens that can dynamically focus waves between two arbitrarily located points without mechanical reorientation or complex beamforming networks. Recent efforts in GRIN lens modeling have also focused on simplified ray-tracing approaches for miniaturized probes13, underscoring the need for accessible design tools in the research community. Parallel to GRIN lens development, analytical design methods for flat optical components have also advanced. For instance, recent work on metasurface-based flat lenses has employed analytical formulations derived from Fermat’s principle and explicit ray mapping to achieve aberration-free focusing with planar phase profiles14. While powerful for discrete-phase or homogeneous substrates, such approaches highlight the growing importance of analytical wavefront engineering in flat optics. In contrast to phase-only metasurfaces or fixed-index lenses, the present work introduces a continuous gradient-index profile for a planar dielectric lens using the critical-angle theorem and Ray-Specific Coordinate Mapping. This method enables arbitrary point-to-point focusing without curved surfaces, discrete phase arrays, or mechanical steering, thereby bridging a gap between traditional GRIN lens design and reconfigurable flat-beamforming devices. In this paper, we propose a novel analytical design method for a planar GRIN dielectric lens based on the critical-angle theorem, enabling continuous beam pointing between arbitrary transmitter and receiver locations. The refractive index profile is derived to ensure phase matching and minimal reflection at the lens boundaries. The design is validated through full-wave numerical simulations in COMSOL Multiphysics, demonstrating effective focusing performance, and improved impedance matching. The proposed flat lens architecture offers significant advantages in terms of fabrication simplicity, integration compatibility, and reconfigurability for modern wireless systems. It is important to note that the design of such a lens involves significant analytical complexity, particularly due to the need for repeated coordinate transformations from the global Cartesian system to multiple rotated coordinate frames, each aligned with the incident ray direction. This step is crucial for accurately modeling wave propagation and ensuring phase matching across the lens aperture. It is worth noting that while the results are presented in the 4–7 THz band to demonstrate a compact proof-of-concept, the proposed analytical framework is inherently frequency-scalable. Due to the non-resonant nature of the GRIN profile, the design principles remain robust for lower frequency regimes, such as the 24–100 GHz millimeter-wave spectrum essential for 5G/6G and satellite communications, by simply applying a linear scaling factor to the lens dimensions. The remainder of this paper is organized as follows: Section 2 presents the theoretical foundation and design methodology. Section 3 describes the simulation setup and numerical results. Section 4 discusses the performance analysis and comparison with existing approaches. Finally, Section 5 concludes the paper and suggests future research directions. Theory and design Critical-angle theorem and coordinate transformations The proposed lens design is based on the critical-angle theorem, which governs the refraction of electromagnetic waves at the interface between two media with different refractive indices. According to Snell’s law: where and are the permittivities of the incident and lens media, respectively, and and are the angles of refraction and incidence. Following the approach introduced by Nasrollahi et al.8, the critical-angle theorem enables the design of an inhomogeneous dielectric lens by effectively rotating the coordinate system and reshaping the wavefront from a transmitting source. This method allows precise control over the ray directions and phase distribution across the lens aperture. To simplify the analysis, we introduce a local coordinate system (x', y') rotated by an angle with respect to the global Cartesian coordinates (x, y), such that the x'-axis aligns with the direction of the incident ray (Fig. 1). The transformation is given by: where d is the lens width, and R, , and are shown in Fig. 2. Schematic of the coordinate rotation from the global (x, y) system to the local (x', y') frame aligned with the incident ray direction. The rotation angle is determined by Snell’s law and the critical-angle theorem. Ray path schematic illustrating the phase-matching condition. The orange path represents the reference ray, while the blue path denotes an arbitrary ray. All optical path lengths from transmitter T to receiver R are equalized to ensure constructive interference at the focal point. This coordinate rotation is performed for each incident ray angle , resulting in a set of local frames that facilitate the phase-matching condition across the lens aperture. The inverse transformation back to the global coordinates is performed after the ray path is computed within the lens. This analytical procedure, hereafter referred to as Ray-Specific Coordinate Mapping, is utilized to determine the static refractive index profile for a given pair of focal points. While the coordinate transformation process itself is dynamic during the design phase to accommodate arbitrary focal points, the resulting index profile remains static once implemented. It is important to distinguish this design-stage algebraic reparameterization from time-variant or physically reconfigurable methods, as it is used solely to enforce the phase-matching condition prior to the full-wave simulation. Phase matching condition To achieve constructive interference at the desired focal point, the optical path length (OPL) for all rays must be equal. Let the reference path (yellow ray in Fig. 2) be defined between the transmitter T and receiver R through points A and U. For an arbitrary ray (pink path) passing through points I and O, the phase-matching condition is expressed as: where denotes the phase shift along each segment. Transmission line analogy To provide an intuitive physical interpretation of the phase accumulation within the planar GRIN lens, each vertical layer of the 1D flat lens located at a transverse coordinate y' can be modeled as a uniform transmission line aligned along the local propagation direction (the x'-axis). In this analogy, the phase constant β of each transmission line segment is determined by the local refractive index and the incident wave angle, given by: where is the free-space wavenumber, and is the refractive index profile derived from the relative permittivity distribution. The effective electrical length (phase shift) accumulated by a wave propagating through a differential layer of thickness dx' is then: The total phase shift experienced by a ray traversing the entire lens along the x'-direction is obtained by integrating Eq. (5) over the lens thickness d: This transmission-line perspective clearly illustrates how the gradient index directly controls the local phase velocity, thereby enabling precise wavefront shaping. The phase-matching condition expressed in Eq. (3) is equivalent to requiring that the total electrical length be equal for all rays traveling from the transmitter to the receiver. Derivation of the refractive index profile The phase-matching condition is formulated by equating the optical path length (OPL) of any arbitrary trajectory (represented by the pink ray) at a given height y' with that of the reference path (indicated by the yellow ray), ensuring a perfectly co-phasal wavefront at the target focal point. Starting from the geometry shown in Fig. 2, the optical path length for the green ray can be written as: Similarly, the optical path for the arbitrary pink ray is: The reference (yellow) ray is given by: Enforcing yields the governing equation for the refractive index profile . Solving for gives the final expression: In the derivation of Eq. (10), is treated as a critical design degree of freedom. For any specific focal geometry, this parameter is uniquely determined such that the refractive index at the lens edges (y' = ± R) naturally equals the refractive index of the surrounding medium (n=1), thereby satisfying the lateral boundary conditions without the need for additional constraints. Impedance matching and two-dimensional gradient index To minimize reflection at the air–lens interfaces, a two-dimensional gradient-index profile is introduced. The relative permittivity distribution is designed to vary smoothly from unity at the lens edges to a maximum value at the center. This is achieved by scaling the one-dimensional profile with a symmetric tapering function N(x') along the propagation direction (x'-axis): where N(x') satisfies the following physical and geometric conditions: N(x') is symmetric about the mid-plane. N(x') is symmetric about the mid-plane. Its value vanishes at both lens interfaces, i.e., . Its value vanishes at both lens interfaces, i.e., . Its derivative is zero at the lens center, at , ensuring a smooth maximum. Its derivative is zero at the lens center, at , ensuring a smooth maximum. It is normalized such that the average value over the lens thickness equals unity, , thereby preserving the effective electrical length of a uniform lens. It is normalized such that the average value over the lens thickness equals unity, , thereby preserving the effective electrical length of a uniform lens. A family of polynomial functions fulfilling these requirements is: The normalization constant is chosen such that the effective electrical length matches that of a uniform lens: For m = 1, the quadratic taper simplifies to: This two-dimensional tailoring ensures gradual impedance transition, reducing the maximum reflection level and improving overall transmission efficiency. The selection of the tapering order m involves a trade-off between the smoothness of the electromagnetic features such as transmission, reflection, and focusing conditions, and the maximum required permittivity. While higher orders of m (e.g., m \ge 2) mathematically satisfy the normalization conditions, they lead to an exponential increase in the peak permittivity value (exceeding 43 for m=2), which induces strong internal reflections and degrades the focusing quality. Consequently, m=1 is selected as the optimal order to maintain a moderate permittivity range () while ensuring high transmission efficiency and a well-defined focal spot. Design parameters The key geometric and electromagnetic parameters used in the lens design are summarized in Table 1. These values are chosen to represent a typical scenario for wave focusing applications. Table 1Design parameters for the planar GRIN lens.ParameterSymbolValueMaximum refractive index2.3Source–lens distanceD0.4 mmLens thicknessd0.1 mmLens radiusR0.9 mmSource offset0.3 mmSource angle45Receiver angle10Receiver direct offset0.7 mm Design parameters for the planar GRIN lens. Simulation and results Simulation setup The proposed planar GRIN lens was modeled and analyzed using the full-wave electromagnetic simulation software COMSOL Multiphysics (version 6.1). The Electromagnetic Waves, Frequency Domain (emw) and Geometrical Optics modules were employed to solve Maxwell’s equations in the frequency domain. Perfectly Matched Layer (PML) boundaries were applied around the simulation domain to absorb outgoing waves and mimic an open space environment. An adaptive mesh was used, ensuring convergence of the electric field solution with an element size smaller than λ /10 at the highest frequency of interest (7 THz). A point source with a Gaussian beam profile was placed at the transmitter location, and the receiver point was set at the desired focal position as defined in Table 1. Also, the transmitter was modeled as a two-dimensional feed horn operating in the fundamental TE mode, providing a realistic wavefront for evaluating lens performance. For the full-wave numerical validation, the spatially varying refractive index profile derived from the Ray-Specific Coordinate Mapping was implemented in COMSOL Multiphysics as a static inhomogeneous medium. The permittivity distribution, , was defined using the “Analytic Function” and “Variables” nodes, allowing the solver to treat the lens as a deterministic material during the mesh assembly. One-dimensional gradient index profile Initially, a one-dimensional (1D) GRIN slab was simulated, where the refractive index varied only along the y'-direction according to Eq. (10). Figure 3 shows the resulting relative permittivity distribution. As expected, the index increases toward the center of the lens, but the abrupt change at the air–lens interfaces leads to significant impedance mismatch and reflection. (a) Simulation space and (b) relative permittivity distribution for the 1D gradient-index slab structure. The electric field distribution at f = 5 THz for this 1D structure is shown in Fig. 4. While some focusing is observed, the focal spot is broad and exhibits significant sidelobe energy, indicating insufficient focusing capability. Furthermore, the presence of strong reflected waves confirms poor impedance matching, particularly at the input and output boundaries. Electric field distribution at f = 5 THz for the 1D gradient-index slab. Two-dimensional gradient index profile To overcome the limitations of the 1D design, a two-dimensional (2D) GRIN profile was implemented using Eqs. (11) and (14). This profile ensures a smooth transition of permittivity from at all lens edges to a maximum value of at the center, as shown in Fig. 5. (a) Simulation space and (b) relative permittivity distribution for the 2D gradient-index slab structure. The corresponding electric field distributions at f = 5 THz is presented in Figs. 6. The wavefront is clearly focused at the receiver point with significantly reduced sidelobe levels and minimal back-reflection. Electric field distribution at f = 5 THz for the 2D GRIN lens. Electric field intensity analysis To quantify the improvement brought by the two-dimensional grading, we compare the electric field intensity distributions at f = 5 THz for both lens designs. As seen in Fig. 7, the unmatched (1D) lens suffers from strong sidelobes and a broad focal region, whereas the matched (2D) lens produces a clean, concentrated beam at the receiver location. This improvement is directly attributed to the smooth permittivity transition provided by the tapering function N(x'). Electric field intensity (|E|) at f = 5 THz for (a) the unmatched 1D GRIN lens and (b) the impedance-matched 2D GRIN lens. The matched lens shows a sharper focal spot and significantly lower sidelobe levels. Also, the electric field intensity (|E|) along the line connecting the transmitter and receiver was extracted and compared for three scenarios: (i) free-space propagation (no lens), (ii) the unmatched 1D GRIN lens, and (iii) the impedance-matched 2D GRIN lens. Figure 8 shows the normalized field intensity at f = 5 THz. Normalized electric field intensity along the transmitter–receiver axis at f = 5 THz for three cases: free-space propagation (blue), unmatched 1D GRIN lens (red), and matched 2D GRIN lens (green). The horizontal axis represents the distance from reference point U along the u-axis (see Fig. 2). The matched lens shows a sharp, high-amplitude peak at the receiver location (). In free space (blue curve), the intensity decays monotonically with distance, exhibiting no focusing effect. The unmatched lens (red curve) produces a broad peak near the receiver but suffers from poor energy confinement. In contrast, the matched lens (green curve) exhibits a sharp, high-amplitude peak precisely at the receiver location, confirming effective energy concentration and minimal spreading. The full-width at 1-dB beamwidth of the focal spot is reduced compared to the unmatched case, demonstrating the critical role of impedance matching in achieving high-quality beam focusing. Ray tracing analysis To geometrically validate the focusing capability of the designed planar GRIN lens, ray-tracing simulations were performed using the Geometrical Optics module in COMSOL Multiphysics. A set of rays emanating from the transmitter point at different initial angles were traced through the lens medium. The refractive index profile defined by Eq. 11 ensures that each ray is bent appropriately according to the critical-angle theorem and the phase-matching condition. Figure 9 shows the ray trajectories after 5 ps of propagation. It is clearly observed that rays entering the lens at various incident angles are redirected and converge accurately at the predefined receiver location. This confirms that the analytically derived gradient-index profile successfully implements the required optical path length equalization for arbitrary point-to-point focusing. The smooth curvature of the rays inside the lens also indicates minimal internal reflection and efficient impedance matching, consistent with the full-wave simulations. Ray trajectories through the planar GRIN lens at t = 5 ps. Rays originating from the transmitter are focused toward the receiver point. Bandwidth performance To evaluate the broadband performance of the proposed lens, the electric field intensity along the transmitter–receiver axis was simulated at multiple frequencies: 4, 5, 6, and 7 THz. The results are plotted in Figs. 10 and 11, where the normalized field intensity (|E|) is shown as a function of position for each frequency. All curves exhibit a pronounced peak at the receiver location, confirming that the lens maintains effective focusing across the entire band. Table 2 summarizes the Longitudinal 1-dB Beamwidth (BW) and Transverse 3-dB BW across the operating frequency range of 4–7 THz. As shown in the Table 2, the lens maintains a stable transverse 3-dB beamwidth (ranging from 0.1053 mm at 4 THz to 0.0725 mm at 7 THz). The Side-Lobe Level (SLL) remain below -11.38 dB for all frequencies, peaking at -15.40 dB at 5 THz. These values confirm a well-confined focal spot with minimal interference. Normalized electric field intensity along the transmitter–receiver axis at multiple frequencies (4, 5, 6, and 7 THz). The horizontal axis represents the distance from reference point U along the u-axis. All curves show a sharp peak at the receiver location, confirming consistent focusing performance across the band. Normalized transverse electric field distribution along the v-axis at different frequencies (4, 5, 6, and 7 THz). The horizontal axis shows the relative distance from the focal point in millimeters. The results demonstrate a well-defined main lobe and low side-lobe levels (SLL) across the operational bandwidth, indicating high spatial resolution and effective suppression of interference. Table 2Quantitative focusing metrics of the proposed lens at different frequencies.FrequencyLongitudinal 1-dBTransverse 3-dBTransverse(THz)BW (mm)BW (mm)SLL (dB)4.00.52170.1053−13.375.00.52560.0890−15.406.00.51480.0806−13.097.00.49360.0725−11.38 Quantitative focusing metrics of the proposed lens at different frequencies. This result validates the broadband capability of the planar 2D GRIN lens design, making it suitable for wideband applications in terahertz communications and imaging systems. Discussion The simulation results presented in Section 3 demonstrate the effectiveness of the proposed planar GRIN lens design based on the critical-angle theorem and coordinated coordinate transformations. This section provides a detailed analysis of the key findings, explains the physical mechanisms behind the observed performance, and discusses the implications of the design choices. The most notable improvement in the 2D GRIN lens, compared to its 1D counterpart, is the sharp and well-defined focal spot at the desired receiver location (Figs. 7 and 8). This is a direct consequence of the enforced phase-matching condition (Eq. 3) and the carefully derived refractive index profile (Eq. 10). The planar geometry, combined with the gradient index, allows for effective beam bending without the need for a curved surface, which is a significant advantage for integration into flat panel systems. The electric field intensity plots (Figs. 7 and 8) confirm that the energy is concentrated within a small region around the receiver point. This is particularly important for applications such as point-to-point wireless links and radar systems, where interference and multipath reflections must be minimized. The electric field intensity distributions (Figs. 8 and 10) reveal a slight displacement of the peak intensity from the predefined geometric focal point. This focal shift is a known phenomenon in dielectric lens systems, particularly for finite-aperture designs operating in the Fresnel region15,16. The shift originates from the amplitude taper of the incident field: while the optical path length is equalized for all rays to satisfy phase matching, the field amplitude decays as 1/r from the source. Consequently, rays arriving at points closer to the lens contribute with higher amplitudes, leading to a constructive interference peak that is shifted toward the lens relative to the geometric focus. The extent of this shift depends on the frequency and lens dimensions; higher frequencies (or larger electrical apertures) produce a narrower focusing zone (1-dB beamwidth), thereby reducing the apparent focal shift and improving targeting accuracy. This behavior confirms that the lens design maintains phase coherence while exhibiting realistic amplitude-based focusing characteristics typical of practical beam-forming systems. The proposed lens maintains consistent focusing performance across a wide frequency range, as evidenced by the field intensity profiles shown in Figs. 10 and 11. This broadband behavior stems from the non-resonant, smoothly graded index profile, which avoids sharp impedance discontinuities that typically induce narrowband resonances. Furthermore, the ability to focus waves between arbitrarily located points distinguishes this work from fixed-focus designs reported in the literature8,12. This flexibility is achieved through the analytical coordinate transformation procedure described in Section 2, which adapts the phase front for any given transmitter–receiver geometry. Performance comparison with classical luneburg models To validate the practical efficacy of the proposed analytical framework, a comparative study is conducted against the conventional bifocal Morgan-Luneburg lens17. The quantitative metrics, summarized in Table 3, reveal that while the ideal spherical symmetry of the Morgan lens yields a slightly narrower beamwidth, the proposed planar design significantly excels in signal integrity. Specifically, a reduction of approximately 7 dB in transverse SLL is achieved at . More importantly, the proposed method bypasses the inherent ’coaxial constraint’ of traditional Luneburg-type lenses, where focal points must be diametrically aligned. By enabling non-aligned, arbitrary focal placements within a low-profile planar substrate, the design offers a degree of freedom that is crucial for modern integrated beam-pointing systems. Table 3Quantitative Performance Comparison of the Proposed Planar Lens vs. the Analytical Morgan-Luneburg lens.MetricMorgan Lens17Proposed ()Proposed ()GeometryCurved (Bulky)Planar (Flat)Planar (Flat)Max. Focus Intensity (dB)*-8.94-9.171-dB Long. BW (mm)0.35590.44080.52563-dB Trans. BW (mm)0.06930.07880.089Transverse SLL (dB)-10.29Focal AlignmentCoaxial (Fixed Axis)Arbitrary (Non-aligned)Arbitrary (Non-aligned)*Defined as the ratio of maximum focal intensity to source intensity. Quantitative Performance Comparison of the Proposed Planar Lens vs. the Analytical Morgan-Luneburg lens. *Defined as the ratio of maximum focal intensity to source intensity. Sensitivity and robustness analysis To evaluate the practical reliability of the proposed analytical framework, a comprehensive sensitivity analysis was conducted by introducing source displacement errors. In THz experimental setups, alignment tolerances are critical; thus, we investigated the impact of a ± 0.05 mm shift in the source position across eight different directions (axial, lateral, and diagonal). The results, summarized in Table 4, demonstrate that the lens maintains a highly stable focusing performance. The maximum deviation in peak field intensity is less than 0.71 dB, and the focal spot shift remains within a predictable range (łe 0.082 mm). Even in the most sensitive case (diagonal displacement), the side-lobe level (SLL) is maintained at an acceptable level. These findings confirm that the design is robust against common assembly misalignments, ensuring consistent performance in real-world THz applications. Table 4Focus performance metrics under various source displacement scenarios (± 0.05 mm).CaseSourceFocusMax. Focus1dB BW3dB BWSLL(x, y)(x, y)Intensity (dB)(mm)(mm)(dB)Reference(0.4, 0.00)(−0.7521, 0.2140)−9.170.52560.0890−15.40Left ()(0.35, 0.00)(−0.8016, 0.2473)−9.830.67370.1049−16.17Right ()(0.45, 0.00)(−0.6896, 0.1946)−8.580.40770.0810−13.74Up ()(0.4, 0.05)(−0.7993, 0.1486)−9.180.58230.0923−15.62Down ()(0.4, −0.05)(−0.6522, 0.2619)−9.290.43010.0815−13.57(0.35, −0.05)(−0.6922, 0.2937)−9.880.54190.0902−13.16(0.45, −0.05)(−0.6452, 0.2442)−8.70.37980.0730−11.86(0.45, 0.05)(−0.7305, 0.1361)−8.680.44930.0815−15.43(0.35, 0.05)(−0.8221, 0.1610)−9.810.73860.1047−9.31 Focus performance metrics under various source displacement scenarios (± 0.05 mm). Fabrication feasibility and tolerance analysis To ensure the practical feasibility of the proposed GRIN lens, two primary fabrication routes are envisioned. For the microwave regime, Fused Filament Fabrication (FFF) using ceramic-loaded thermoplastics (e.g., PREPERM) allows for local permittivity control by varying infill density18,19. For THz applications, High-Resistivity Float Zone Silicon (HRFZ-Si) is identified as the optimal material due to its exceptionally low loss ()20,21. The gradient profile can be realized by creating sub-wavelength air-filled pores using deep reactive ion etching (DRIE), or alternatively, via 3D-printing techniques such as Stereolithography (SLA) and projection micro-lithography, which have been successfully employed to fabricate 3D THz GRIN lenses22,23. The impact of manufacturing imperfections is quantified in Table 5. Numerical results indicate that a deviation in local permittivity –typical for 3D-printed photopolymers24 – results in a minor longitudinal focal shift of ≈ 0.03 mm, while maintaining peak intensity within 0.3 dB of the reference. However, the design is more sensitive to geometric thickness; an error of can lead to a gain reduction of up to 1.38 dB and an increase in sidelobe levels. Therefore, a fabrication tolerance of is recommended to maintain optimal gain and SLL levels. These results confirm that the analytical design remains robust under standard micro-fabrication tolerances. Table 5Quantitative analysis of fabrication tolerances and their impact on focusing performance.TolerancePermissibleImpact on Max.Impact onTransverse3-dB BW Parameter Range Focus IntensityFocal Shift SLL(dB) VariationPermittivity Error ()< 0.3 dB variation mm-12.88 ∼ -15.34< 0.008 mmThickness Error (Δ L)< 0.5 dB drop mm-12.84 ∼ -15.13 mmThickness Error (Δ L)0.95 ∼ 1.38 dB drop± 0.06 mm-13.53 ∼ -13.75< 0.01 mm Quantitative analysis of fabrication tolerances and their impact on focusing performance. Conclusion This paper has presented an analytical design methodology for a planar gradient-index (GRIN) dielectric lens capable of focusing electromagnetic waves between two arbitrarily located points. The design leverages the critical-angle theorem combined with Ray-Specific Coordinate Mapping to control the wavefront phase. The refractive index profile is derived analytically to satisfy the optical path length equality condition, ensuring constructive interference at the desired receiver point. To mitigate reflections and improve impedance matching, a two-dimensional grading function is introduced, which smoothly varies the permittivity from the lens edges to the center. Full-wave simulations in COMSOL Multiphysics validated the design, demonstrating significant performance improvements over a conventional one-dimensional GRIN structure. The electric field distributions confirmed effective beam concentration at the target location with suppressed sidelobes. The planar architecture of the lens offers practical advantages in terms of fabrication simplicity and integration into flat-panel systems, making it suitable for emerging applications in Terahertz imaging, 5G/6G communications, satellite terminals, automotive radar, and conformal antennas. Author contributions MMT: conceptualization, data curation, validation, writing, and project administration. MM: conceptualization. SF:methodology and review. All authors have read and agreed to the published version of the manuscript. Funding This research did receive funding. Mohammad Mahdi Taskhiri received funding from Iran National Science Foundation; Grant ID 4047177. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.