Complementary Split-Ring Resonators

• Complementary Split-Ring Resonators are electrically small metamaterial inclusions that create negative effective permittivity for frequency-selective multi-band operation.
• They are integrated into microstrip patch antennas for IoT applications, where precise control of geometric parameters optimizes resonance frequency and return loss.
• Advanced multi-objective genetic algorithms, such as NSGA-II and NSGA-III, optimize CSRR-loaded designs by balancing gain, S11 reduction, and manufacturability.

Complementary Split-Ring Resonators (CSRRs) are electrically small metamaterial inclusions used to engineer the electromagnetic response of antenna substrates and microwave components within the context of advanced multi-objective optimization frameworks. CSRRs enable compact multi-band and bandstop characteristics by producing negative effective permittivity, which is exploited in microstrip patch antenna design for Internet of Things (IoT) and related applications. The recent integration of CSRR-loaded patches with multi-objective genetic algorithms (MOGAs) provides a rigorous methodology for optimizing electromagnetic properties across multiple criteria simultaneously, including resonance frequency, return loss ($S_{11}$), and antenna gain (Boulaich et al., 24 Jan 2026).

1. Structure, Physical Principles, and Electromagnetic Properties

CSRRs consist of concentric metallic rings etched on the ground plane of microwave substrates, forming the geometric complement of conventional split-ring resonators (SRRs). The canonical CSRR geometry involves two or more split rings with gaps situated opposite one another, producing resonant behavior via electric dipole coupling with incident fields. For microstrip structures, the inclusion of CSRRs beneath the radiating patch results in strong electromagnetic coupling leading to frequency-selective negative permittivity at predetermined resonance frequencies. This produces compact bandstop or passband responses not achievable with traditional patch layouts (Boulaich et al., 24 Jan 2026).

Key parameters governing CSRR physics include:

• Inner and outer ring radii ($R_k$), inter-ring gap ($g$), and split width.
• Resonance frequency ($f_r$) set by the geometric and substrate parameters; analytical and full-wave numerical models are used to predict $f_r$ for typical patch geometries.
• Substrate properties: Rogers RT5880 ($\epsilon_r=2.2$, $\tan\delta=0.0009$, $h=1.57\,\mathrm{mm}$).

Electrical smallness (with respect to wavelength) enables subwavelength tailoring of effective material response. This ensures multi-band operation within the form factor constraints of modern wireless systems.

2. CSRRs in Multi-Band Patch Antenna Design

CSRR-loaded microstrip antennas are engineered for simultaneous resonance at multiple frequencies, where each CSRR configuration targets a specific band (e.g., 2.4 GHz, 3.6 GHz, 5.2 GHz) by structural and substrate manipulation. The optimization process aims to minimize return loss ($S_{11}<−10$ dB) and maximize gain, subject to geometry and substrate constraints.

The physical design variables for tri-band optimization typically include:

Variable	Range	Role in $f_r$, $S_{11}$
$R_k$ (ring radius)	$[R_k^{min}, R_k^{max}]$	Sets band location, controls coupling
$g$ (gap)	$[g^{min}, g^{max}]$	Dictates Q-factor, split width
$W$, $L$ (patch dims)	$[W^{min}, W^{max}]$	Global resonance tuning

Constraints include $S_{11} < -10$ dB at all target bands and physical bounds on geometric variables. Penalty and repair mechanisms are used in optimization to ensure feasible solutions.

3. Multi-Objective Genetic Algorithm Frameworks for CSRR Optimization

CSRR-based antenna design leverages various forms of MOGAs to simultaneously optimize conflicting objectives (e.g., maximizing multi-band $S_{11}$ reduction and gain). Five major MOGA variants have been systematically compared for this task (Boulaich et al., 24 Jan 2026):

1. Pareto Genetic Algorithm (PGA): Assigns Pareto ranks based on dominance; no diversity preservation.
2. NSGA-I: Non-dominated sorting with niching/fittest sharing.
3. NSGA-II: Fast non-dominated sorting with crowding distance and elitism.
4. NSGA-III: Non-dominated sorting supplemented with reference-point guidance for scalable multi-band coverage.
5. SPEA: Maintains an archive of non-dominated solutions selected by strength; includes density-based truncation.

For practical design, a weighted-sum scalarization is often used to collapse the multi-objective space to a single robust compromise solution:

$$F(x) = w_1\,|S_{11}\text{@2.4GHz}| + w_2\,|S_{11}\text{@3.6GHz}| + w_3\,|S_{11}\text{@5.2GHz}|$$

with equal weights typically assigned to reflect the desired balance among bands.

4. Constraint Handling and Solution Feasibility

Rigorous constraint handling is fundamental to physical realizability:

• Penalty Functions: Large coefficient $\alpha$ enforces $S_{11}$ threshold violations.
• Repair Functions: Gene values (e.g., $R_k$, $g$) are clamped to bounds post-crossover/mutation.
• Fitness Evaluation: $F_{total}(x) = F_{objective}(x) + P(x)$, ensuring infeasible antennas are penalized and removed from the elite set.

Designs must satisfy all electromagnetic and geometric criteria for practical fabrication and deployment.

5. Comparative Performance Analysis

Systematic empirical evaluation on tri-band problems demonstrates:

Algorithm	$S_{11}$(2.4 GHz)	$S_{11}$(3.6 GHz)	$S_{11}$(5.2 GHz)	Gain (dBi) @ bands
PGA	−16.02	−34.23	−1.15	1.7 / 2.1 / 3.2
NSGA-II	−16.34	−33.81	−0.98	1.8 / 2.3 / 3.5
NSGA-III	−32.81	−10.52	−10.72	2.0 / 2.7 / 3.7
SPEA	−24.94	−12.52	−10.41	1.9 / 2.4 / 3.4
Scalarized GA	−21.56	−16.60	−27.69	1.96 / 2.6 / 3.99

Scalarization yields a robust single solution meeting all thresholds, while NSGA-III demonstrates the best generational distance and inverted generational distance metrics (GD ≈ 0.19, IGD ≈ 0.18), signifying reliable Pareto-front coverage for multi-band designs. Computational cost is primarily driven by EM simulation runtime (≈2 min/simulation); co-simulation and surrogate modeling strategies are recommended for scalability (Boulaich et al., 24 Jan 2026).

6. Practical Implications and Best-Practice Recommendations

• Algorithm Selection: NSGA-II/III and SPEA suit full trade-off exploration; scalarized GA is preferred for single compromise solutions meeting all specifications simultaneously.
• Parameter Tuning: Mutation scale, tournament size, penalty weights, and reference-point generation for NSGA-III require sensitivity analysis for robust performance.
• Constraint Integration: Combined use of repair and penalty ensures physically feasible, manufacturable CSRR antenna designs.
• Co-Simulation Overhead: Parallel or surrogate EM evaluation schemes recommended to accelerate convergence in realistic design pipelines.

Integration of CSRR technology with advanced evolutionary multi-objective optimization substantially advances the state-of-the-art in compact multi-band microwave antenna engineering. Optimized CSRR-loaded structures deliver high-performance metrics across all frequency bands of interest, with rigorous constraint satisfaction and superior Pareto surface navigation (Boulaich et al., 24 Jan 2026).