Ilmu Komputer & AI editorial
Open AccessOA2026
High-Precision Hybrid FA-PSO Based Inversion of Building Material Parameters for Fundamental Wireless Performance Evaluation
A firefly-optimized particle swarm framework for estimating permittivity, conductivity, and thickness of building materials, benchmarked against the Cramér-Rao lower bound
Zhuowei Li; Yalei Zhu; Hanqing Zhang; Sui Li; Meng Chen; Tong Zhang; Zi-Yang Wu; Dan Yang; Songjiang Yang; Jiliang Zhang· 2026· DOI 10.48550/arXiv.2607.12721
The core problem
Accurate knowledge of the electromagnetic properties of building materials is a prerequisite for evaluating wireless performance in indoor and urban environments. Parameters such as permittivity, conductivity, and thickness govern reflection, transmission, and absorption of radio waves, and therefore directly influence coverage prediction, channel modeling, and network planning. The free-space method offers a practical route to extract these parameters from measured transmission or reflection coefficients, but the associated inverse problem is nonlinear and often ill-conditioned. Classical gradient-based inversion schemes are sensitive to initial guesses and can become trapped in local minima, while standard metaheuristics may converge slowly or lack a theoretical accuracy benchmark. This paper addresses both issues by introducing a hybrid inversion method based on the firefly algorithm (FA) and particle swarm optimization (PSO), referred to as FA-PSO. The central idea is to use an adaptive FA to systematically optimize the hyperparameters of PSO, thereby improving convergence efficiency and robustness. In addition, the authors derive the Cramér-Rao lower bound (CRLB) for permittiv
Innovation
The numerical experiments evaluate the FA-PSO inversion method against the derived CRLB. The key finding is that, for relatively thin materials, the estimation accuracy of the proposed method approaches the theoretical lower bound. This indicates that the hybrid optimizer is able to extract permittivity, conductivity, and thickness with near-optimal efficiency in the thin-material regime, where the inverse problem is particularly challenging due to limited interaction between the electromagnetic wave and the material. The convergence behavior is improved by the adaptive FA tuning of PSO hyperparameters, and the optimized Gaussian initialization contributes to progressively better parameter estimates. The comparison with the CRLB confirms the effectiveness of the inversion framework: rather than merely reporting empirical errors, the study demonstrates that the achieved accuracy is close to the fundamental limit imposed by the noise model. For thicker materials, the results suggest that the gap to the CRLB may widen, although the abstract emphasizes the thin-material case as the primary validation. Overall, the results support the claim that FA-PSO provides a robust and accurate mea
Accurate knowledge of the electromagnetic properties of building materials is a prerequisite for evaluating wireless performance in indoor and urban environments. Parameters such as permittivity, conductivity, and thickness govern reflection, transmission, and absorption of radio waves, and therefore directly influence coverage prediction, channel modeling, and network planning. The free-space method offers a practical route to extract these parameters from measured transmission or reflection coefficients, but the associated inverse problem is nonlinear and often ill-conditioned. Classical gradient-based inversion schemes are sensitive to initial guesses and can become trapped in local minima, while standard metaheuristics may converge slowly or lack a theoretical accuracy benchmark. This paper addresses both issues by introducing a hybrid inversion method based on the firefly algorithm (FA) and particle swarm optimization (PSO), referred to as FA-PSO. The central idea is to use an adaptive FA to systematically optimize the hyperparameters of PSO, thereby improving convergence efficiency and robustness. In addition, the authors derive the Cramér-Rao lower bound (CRLB) for permittivity, conductivity, and thickness under a complex Gaussian noise model, providing a theoretical benchmark against which the estimation accuracy of FA-PSO can be assessed. The study thus combines algorithmic innovation with rigorous statistical performance analysis, aiming to deliver reliable material parameter extraction that supports fundamental wireless performance evaluation.
The inversion framework operates on free-space measurements and seeks to recover the vector of unknown material parameters
, where is the permittivity, the conductivity, and the thickness. The forward model relates these parameters to the measured complex transmission coefficient (or reflection coefficient), and the inverse problem is cast as a minimization of the discrepancy between measured and modeled responses.
Why it matters
The significance of this work lies in bridging metaheuristic optimization and statistical estimation theory for material characterization. By deriving the CRLB under a complex Gaussian noise model, the authors establish a rigorous benchmark that goes beyond typical convergence curves or error tables. The fact that FA-PSO approaches this bound for thin materials suggests that the hybrid strategy successfully mitigates the premature convergence and slow exploration often observed in standalone PSO or FA. The adaptive FA component effectively acts as a hyperparameter controller, allowing PSO to adapt its search behavior without manual tuning. The optimized Gaussian initialization further reduces the risk of poor starting conditions, which is critical in ill-posed inverse problems. From an application perspective, accurate extraction of permittivity, conductivity, and thickness directly supports wireless performance evaluation, including indoor propagation modeling, coverage prediction, and the design of communication systems in built environments. The taxonomy candidates—Architecture, Cybersecurity, Network, and Cryptography—reflect the broader relevance of material-aware channel modeling to network planning and secure wireless system design. Limitations include the focus on the free-space method and the thin-material regime; future work could extend the CRLB analysis and FA-PSO framework to multilayer structures, oblique incidence, and frequency-dependent material properties. Nonetheless, the study provides a solid methodological foundation for high-precision material parameter inversion.
Who should read this
CS practitioners and researchers
Opening member content…